The Emergence of Tab in Najdi Arabic

Abstract

This study empirically investigates functional and social variation in the use of TAYYIB (‘okay, well, right’) in light of grammaticalisation. Thirty naturally occurring conversations of 60 Najdi Arabic speakers were recorded in Riyadh, Saudi Arabia. The analysis demonstrates that TAYYIB has two realisations: full Tayyib [tˁajjib] and reduced Tab [tˁab]. Drawing on the conversation analytic approach within a variationist framework, TAYYIB was used to perform multiple discourse-pragmatic functions: interpersonal, textual and interpersonal–textual. The statistical analysis reveals that variant choice is significantly conditioned by the pragmatic functions. While Tayyib is employed to perform all three functions, Tab is only used for textual and interpersonal–textual meanings. As for social factors, Tab is significantly more likely to be used by younger speakers than adults and also more likely to be used by females than males. This can be interpreted as an indication of ongoing change driven by young people, primarily females, towards the greater use of the innovative Tab. Given the evidence of linguistic change in Tab including semantic bleaching, pragmatic strengthening and phonological reduction, the study suggests that Tab has undergone advanced grammaticalisation.

1. Introduction

The Arabic TAYYIB is originally an adjective, as defined in Arabic dictionaries, which means ‘good, opposite of bad’ (Ibn Manzur 2003, p. 564). TAYYIB has developed an extensive variety of discourse-pragmatic functions in modern Arabic to imply the meaning of ‘well, all right, OK’ (Badawi and Hinds 1986, p. 553). The pragmatic development of TAYYIB has been reported in many spoken Arabic varieties, such as Egyptian (Ghobrial 1993; Ismail 2015), Jordanian (Al-Harahsheh and Kanakri 2013), Lebanese (Ayash 2016) and Saudi (Aljutaily 2021). Furthermore, as corroborated by Al-Harahsheh and Kanakri (2013), Ayash (2016) and Ismail (2015), TAYYIB has two different phonetic realisations in Egyptian, Jordanian and Lebanese Arabic: (i) the full variant Tayyib [tˁajjib], and (ii) the reduced Tab [tˁab], whereby Tab underwent phonological reduction by the deletion of [jji] in the word-internal position.1 Although the form variability in TAYYIB has been documented in the literature and the diversity in the function of TAYYIB has been extensively qualitatively studied, the variation and change in TAYYIB have not previously been explored in the context of spoken Arabic. In the current Najdi Arabic (NA) data, TAYYIB is frequently used, and it is identified in two variants, as shown in Example (1).2

(1)	Tab yi3zmah fi: baitah↓ Tayyib (..) w £yimtaHnah£	[S044, F, Adult]
	Tab he can invite him to his home↓ Tayyib (..) and £test him£

Thus, this study empirically examines the form, function and social variation in the use of TAYYIB in an NA interaction by combining qualitative and quantitative methods. It investigates discourse-pragmatic functions of TAYYIB variants in the interactional situation and how the variation is constrained by linguistic and social factors. It also aims to examine whether the linguistic change in TAYYIB reflects grammaticalisation. Therefore, the study attempts to answer the following questions:

• What are the discourse-pragmatic functions performed by TAYYIB in NA conversations?
• Are there form-function association pattens in the use of TAYYIB?
• Are there age and/or gender differences among NA speakers regarding the use of TAYYIB?
• Is there any indication of ongoing change in TAYYIB in NA that can suggest grammaticalisation?

1.1. Previous Studies of Arabic TAYYIB

Examining previous research on Arabic Tayyib and its variant Tab reveals that these forms are utilised to accomplish a wide range of discourse-pragmatic meanings, which represent two categories: interpersonal and textual functions. In terms of interpersonal functions, TAYYIB is employed to indicate speakers’ goals, attitudes and awareness of previous contributions and recipients. This includes signalling the speaker’s acceptance, permission, approval and compliance, as found in Cairene Egyptian (Ismail 2015) and Jordanian Arabic (Al-Harahsheh and Kanakri 2013). Tayyib is frequently used by recipients to indicate acknowledgement and for continuation, as observed in Egyptian (Ismail 2015), Jordanian and Saudi Arabic. The other common functions of using TAYYIB are to reduce the force of disagreement and introduce objection (e.g., Al-Harahsheh and Kanakri 2013; Aljutaily 2021; Ghobrial 1993), as well as to minimise the effect of directive acts, such as requests (e.g., Aljutaily 2021; Ismail 2015). As for textual functions, TAYYIB is commonly used to structure the discourse units and organise the exchange of the interaction, for example, using TAYYIB as a turn-taking device, as found by Ghobrial (1993), Ayash (2016) and Aljutaily (2021), and to close the turn and end the conversation, as documented by Al-Harahsheh and Kanakri (2013), Aljutaily (2021) and Ghobrial (1993). It has repeatedly been documented that TAYYIB is employed to indicate the transition to a new topic or activity and to shift the topic of the discourse (e.g., Al-Harahsheh and Kanakri 2013; Aljutaily 2021; Ayash 2016; Ghobrial 1993). As demonstrated in Lebanese (Ayash 2016) and Syrian Arabic (Alkhalil 2005), marking a transition in a conversation by TAYYIB can be signalled by requesting an explanation or raising questions.

While TAYYIB has received considerable attention from scholars seeking to analyse its pragmatic roles, sociolinguistic variation and change have received less attention. The form variability in TAYYIB has been confirmed in Egyptian (Ismail 2015), Jordanian (Al-Harahsheh and Kanakri 2013) and Lebanese Arabic (Ayash 2016). In Egyptian and Jordanian Arabic, the reduced-form Tab is the most common variant, accounting for approximately 70% in both varieties (Al-Harahsheh and Kanakri 2013; Ismail 2015). However, neither of these studies examined the effect of discourse-pragmatic functions or social variables on the selection of TAYYIB variants.

1.2. TAYYIB Equivalents in Other Languages

TAYYIB has pragmatic equivalents in a wide variety of languages, including the English okay (e.g., Hopper 1989; Schegloff and Sacks 1973, inter alia.), right, alright (e.g., Filipi and Wales 2003; Schleef 2005) and well (e.g., Aijmer 2013; Beeching 2016); Chinese hao, which is an adjective meaning ‘good’ (e.g., Miracle 1989, 1991; Wang and Tsai 2005); Hebrew tov, which is an adjective meaning ‘good’ (e.g., Maschler 2004, 2009); and Spanish bien, which is an adverb meaning ‘well’ (e.g., De Fina 1997; Fuentes 1993). It is frequently established that these discourse-pragmatic features signal an extensive number of discourse-pragmatic functions that contribute to the interpersonal and textual meanings, which are approximately equivalent to Arabic TAYYIB. For example, the English okay is frequently observed to perform two main functions: (i) discourse transitional functions (e.g., Beach 1993; Beach 1995; Condon 1986, 2001; Couper-Kuhlen 2021a; Hopper 1989; Kovarsky 1989; Levin and Gray 1983; Merritt 1978; Rendle-Short 1999; Schegloff 1968; Schegloff 1986; Schegloff and Sacks 1973; Schleef 2008), and (ii) interactional functions (e.g., Beach 1993; Condon 1986, 2001; Couper-Kuhlen 2021a; Filipi and Wales 2003; Heisler 1996; Heritage and Sorjonen 1994; Merritt 1978; Schleef 2005, 2008). Marking the transition in discourse is a common function of well (Aijmer 2013; Beeching 2016), the Hebrew tov (e.g., Maschler 2004, 2009), Chinese hao (Miracle 1989, 1991; Wang and Tsai 2005) and Spanish bien (De Fina 1997; Fuentes 1993). Additionally, these items have been shown to perform interpersonal functions such as asking continuation, as seen in bien (Fuentes 1993); signalling agreement and acceptance, as observed in well (Aijmer 2013), hao (Wang and Tsai 2005) and tov (Maschler 2004, 2009); and introducing disagreement, as reported in well (Aijmer 2013; Beeching 2016) and tov (e.g., Maschler 2004, 2009). The influence of social variables on the distribution of the English equivalents has been extensively investigated. With regard to variation in gender, according to Heisler (1996), men employ okay more frequently than women in all pragmatic functions, whereas women utilise well substantially more than men (Beeching 2016). On the other hand, no gender association patterns were found regarding the use of okay, alright or right (Schleef 2008) and this observation is consistent with the findings of Levin and Gray (1983). With regard to the age effect, Schleef reported that younger university lecturers (31 to 50) use okay and alright significantly more frequently than their older counterparts (51+), thereby supporting the results of Heisler (1996) and Levin and Gray (1983). Conversely, Beeching (2016) maintains that older speakers of British English utilise well to a greater extent than younger speakers.

This section has provided a summary of the results of prior research on Arabic TAYYIB and its functional equivalents in other languages. The following sections are structured as follows: Section 2 offers an overview of the corpus and analytic methodologies and a brief description of TAYYIB variants in NA. Section 3 provides the findings of the qualitative and quantitative analysis of TAYYIB, followed by the discussion and conclusion in Section 4 and Section 5, respectively.

2. Data and Methods

2.1. The Data

The speaker sample included 60 native speakers of NA who were born and spent most of their lives in Riyadh (the capital of Saudi Arabia). The participants were evenly stratified by age and gender, as indicated in

Table 1. Speaker sample.

Age/Gender Groups	Young (18–21)	Adults (30–40)	Older Adults (55+)
Females	10	10	10
Males	10	10	10
Total	20	20	20

. The data of the current study are based on self-recorded naturally occurring speech. Speakers engaged in informal dyadic conversations with self-selected partners (a member of their family or friend of the same sex) without the presence of the researcher in the recording situation. A total of 30 conversations were obtained, with each lasting approximately 30–45 min. The conversations were audio-recorded using a Zoom H4n Pro portable voice recorder, which was provided to the participants. The corpus contains approximately 18 h of audio and has been segmented into turns and transcribed using the annotation programme ELAN (2023), yielding approximately 145,000 words (for transcription conventions, see Appendix A).

2.2. Variants of TAYYIB

The variable TAYYIB in the current dataset is found in two distinctive phonetic realisations and the differentiation between the variants is dependent on the phonological characteristics of TAYYIB. The variants are classified into full and reduced forms based on the form variability. The full form corresponds to standard Tayyib [tˁajjib], which has two syllables tˁay and yib, as illustrated in (2a). The reduced-form Tab is pronounced as [tˁab] (see (2b) for an example) and it has undergone phonetic reduction by the deletion of [jji] in the word-internal position. Palva (1968, p. 8) suggests that Tab is ‘doubly reduced from Tayyib > Tayb > Tab […where there is] probably no intermediate form with a monophthongised diphthong existing in either instance.’ This suggests that Tab has undergone phonetic reduction, whereby it becomes monosyllabic [tˁab] from the original disyllabic adjective [tˁajjib] and it becomes unaccented. The gradual loss of phonological substance is called phonological attrition by Lehmann (2015, p. 134). It is also defined as phonetic erosion by which there is the loss of phonetic segments, including the loss of full syllables (Heine and Kuteva 2007, p. 43). During the transcription processes, orthographic representations of the two variants were established (i.e., the full-form Tayyib is coded as [طيب] and the reduced-form Tab as [طب]).

(2)	a	Tayyib, wish Sa:r 3ali:h?	[S018, M, Adult]
		Tayyib (okay), what’s happened to him?
	b	Tab, 2intabhi: 2ntabhi:↓ lah	[S033, F, Young]
		Tab (well then), take care, take care↓ of it

2.3. Extraction and Coding

All instances of Tayyib and Tab that function as discourse-pragmatic features were extracted from the corpus. Some tokens of Tayyib and Tab were excluded from the sample. All occurrences of Tayyib that function as an adjective (lit. ‘good, kind-hearted’) were excluded from the dataset (n = 34). Moreover, adopting the convention of the variationist sociolinguistics (Tagliamonte 2006, pp. 88–97), tokens that occur in particular contexts were removed from the sample, including tokens found in direct quotations (n = 21) and interrupted speech (n = 2). Likewise, two Tab tokens were eliminated due to their co-occurrence with the English okay, which results in functional ambiguity. The remaining dataset consists of a total of 491 tokens of the variable TAYYIB (Tayyib n = 354 and Tab n = 137). In addition to coding each token of the variable TAYYIB for its two variants (full Tayyib and reduced Tab), each token was coded for discourse-pragmatic function (i.e., for both macro and micro functions and for functional classifications, see

Table 2. The discourse-pragmatic functions performed by TAYYIB.

Communicative Domains (Macro Functions)	Functions (Micro Function)	Description
Interpersonal	Subjective as a listener response (agreement and acceptance token)	Agreement to requested action; acceptance of what has been said
	Subjective as a listener response (acknowledgement token)	Acknowledgement and continuer
	Intersubjective as a tag-positioned	Check discourse progression; eliciting agreement
Textual	Transitional and organisational device	Mark the transition to a new topic/action
		Mark the transition into an ongoing topic/action
	Turn-exchange device	Turn-taking device to launch contribution
		Turn/topic-closing device to close the turn and terminate the topic
Interpersonal–textual	Mitigator and transitional device	Soften the directiveness of command/request act while marking transitions
		Reduce the force of disagreement while marking transitions

below) as well as the social factors: age and gender. Age was coded as young (aged between 18 and 20 years), adult (between 30 and 40 years) and older adult (55+ years), while gender was coded as male or female.

2.4. Data Analysis Methods

The current study employs both qualitative and quantitative methodologies to investigate the form, function and social variation in TAYYIB. In order to establish the functional repertoire and to facilitate the quantitative analysis of whether the variant choice is conditioned by its function, each token was coded for its function. To categorise discourse-pragmatic functions, Brinton’s (1996) approach was applied, which identifies two functional domains for the pragmatic markers: interpersonal and textual. The qualitative analysis of the discourse-pragmatic functions draws on the conversation analysis approach within the variationist sociolinguistics framework (e.g., Eiswirth 2020; Pichler 2013, 2016; Pichler and Hesson 2016). When assigning and coding discourse-pragmatic functions, a careful evaluation of the interactional environment and discourse organisation that triggered the use of such discourse-pragmatic TAYYIB is required. The contribution of linguistic and non-linguistic features, such as turn-placement, interaction structure and sequence organisation, prosodic and intonational features, and the co-occurrence of other linguistic and non-linguistic features were considered for the purpose of interpreting the intended function. It has been widely reported that these linguistic and non-linguistic features play essential roles in understanding and determining the intended meanings of the discourse-pragmatic features (e.g., Aijmer 2013; Aljutaily 2021; Alkhalil 2005; Couper-Kuhlen 2021a, 2021b; Du Bois et al. 2013; Maschler 2009).

To determine the contribution that function, age and gender make to the use of TAYYIB’s variants in NA, quantitative variationist sociolinguistics (Labov 1972) methods were applied. This includes distributional analysis of the raw and normalised frequency of all variants across the examined factors and the chi-squared test to determine the statistical significance. Regression analysis was conducted to evaluate the significance of several factors together and to identify interactions between independent variables (see Section 3.3).

3. Results

3.1. The Functions of TAYYIB

The current analysis demonstrates that TAYYIB is used in NA to perform various discourse-pragmatic functions, which represent three functional modes: interpersonal, textual and interpersonal–textual. As for discourse-pragmatic usage, TAYYIB is used for (i) interpersonal functions, either for subjective meanings that express the speaker’s beliefs and stance towards what is said or for intersubjective meanings that convey the speaker’s attention to the cognitive stances and social identities of the addressee (Traugott 2003, p. 124); (ii) textual functions when TAYYIB’s variants are used to mark the transition in the discourse to a new action/topic or as a turn-exchange device; and (iii) interpersonal–textual functions where TAYYIB is used simultaneously to mark the transition and perform intersubjective meaning as a mitigator, either to reduce the force of disagreement or to soften directive acts.3 The following sections discuss each function in detail with examples. Table 2 summarises the discourse-pragmatic functions performed by TAYYIB.

3.1.1. Interpersonal Functions

Extract (3): A book discussion
1	S042	yishraH lik mathalan [iTTayyia:r]↓
		It [the book] explains for you with examples, let’s say [a pilot]↓
2	S041	[uhm]
3	S042	shlo:n Tayyia:r yitHawwal 2ila: ta:jir
		How the pilot can be a businessman?
4		(…)
5	S041>	Tayyib
		okay
6		(1.05)
7	S042	2int mihnitik Tayyia:r .hh mafro:DH 2ink lamma: tiko:n tishtighil
8		Tayyia:r ti- tit3allam 2asa:siya:t ishighil lamma [tro:H lidd]owal
		As a pilot hh. you should le- learn the basics of business when you [travel
		to other cou]ntries
9	S041	[2i:h]
		[okay]

The speakers (adult females) in Extract (3) were discussing a book about entrepreneurship. Speaker S042 had completed all the chapters and summarised a chapter for S041 that she had missed. S041, as a recipient, produced multiple minimum responses to reflect her acknowledgement with the preceding turns and request continuation, and these included uhm (line 2), Tayyib (line 5) and 2i:h (line 9). According to Schegloff (1982), acknowledgement treats the previous turn as unproblematic because they do not give an opportunity to initiate a repair. Eiswirth (2020, p. 120) finds that acknowledgements are ‘both backward- and forward-looking’ because they accept what has been said in the prior turn and request continuity from the current speaker without taking any particular position in the current turn. This use of Tayyib is common in other Arabic varieties (Abdeljawad and Radwan 2016; Al-Harahsheh and Kanakri 2013; Aljutaily 2021; Ghobrial 1993; Ismail 2015), but it has not been found in Syrian Arabic (Alkhalil 2005). When using Tayyib as a backchannel and continuer, it is often produced with a relatively flat intonation contour, as in Extract (3), and this has been found by Aljutaily (2021) in Saudi Arabic.

Speakers use Tayyib turn-finally as an invariant lexical tag to check the recipient’s understanding and the progression of the discourse. Extract (4) illustrates this.

Extract (4): Eid
1	S036	bass tigo:l 2axh:f 2alga: 2aHad mithli: (..) £taxayiali: bil3i:d£
		But she said I am afraid to find somebody wearing like me (..) £in Eid party£
2	S035	[laughing]
3		(…)
4	S036>	gabil madri: kam sanah 2aw [laughing] £kan 3alina il3i:d£ Tayyib
		Before I don’t know how many years or [laughing] £we had the Eid party£ Tayyib/okay
5		(..)
6	S035	Tayyib
		okay
7		(…)
8	S036	ya3ni 3i- can 3i:d kidha: l3omman 2aboy w kidha:
		I mean Ei- it was Eid for my fathers’ relatives and that

In this extract, S036 (a young female) talked about one of her relatives wanting to buy a dress, but she was worried about the possibility of finding someone wearing exactly the same dress. In line 4, she started telling a story about when the same situation had occurred at an Eid party. At the end of her turn, S036 produced Tayyib with high intonation to check on the progression of her discourse and to ensure her addressee’s involvement in the story. In line 6, S035 replied with a minimum response Tayyib pronounced with relatively level intonation to indicate that she was following the speaker’s utterance and seeking continuity. Turn-finally, Tayyib is employed as an appeal to the recipients to request (non)verbal acknowledgement of their continued attention where it is received by Tayyib in the next turn. This use of Tayyib has not been found in other Arabic varieties but is equivalent to the English okay and right. Beach (1995, p. 272) and Beach and Lindstrom (1992, p. 330) state that the tag-positioned okay in English is employed to seek and ensure agreement and/or alignment with the following speaker. In the same line, okay is used to ensure that the listener is following the speaker’s discourse (Schleef 2005, p. 56) or to track the progression of the speaker’s speech (Heisler 1996, p. 297).

3.1.2. Textual Functions

Tayyib and Tab are frequently used by speakers to mark the transition to a new topic or action in the ongoing interaction. Extract (5) illustrates the transitional function of Tayyib where the speaker employed it to change the topic of the discussion.

Extract (5): Travel suspension
1	S024	2ana: 2ana: law 2ini: 2aHi- ma: 2aHiT sakso:kah .hh tlga:ni: 2aHallig
2		fi: 2ilbait
	If I would not shave my beard in a square style .hh I would shave it
		at home
3		(0.56)
4	S023	uhm
5		(..)
6	S024	bass 2isakso:kah £tbahdhlni: £ (..) 2aw 2digin law HaTi:t digin
		but, it is difficult for me to do the square style or if I want to keep the chin
7		(1.70)
8	S023	Tayyib .hh titwaga3 yiftHo:n isafar fi: um- shawwal↑
		okay .hh do you think they will resume the flight on um- shawwal↑
		(a month in the Hijri calendar)
9		(…)
10	S024	ma: 2a3tigid wallah
		I do not think so, wallah

In this sequence, the speakers (older-adult males) were discussing their habits regarding beard shaving. In line 1, S024 talked about his beard-shaving style and continued in line 5. Speaker S023 only produced the minimal response uhm in the first turn, but in the third turn, no response was produced. After the silence, S024 initiated a new topic and prefaced his turn with Tayyib to change the topic, which was followed by a question asking S023 about his expectation of cancelling the travel suspension in the next month. Introducing a completely new topic that is unrelated to its previous discussion and indicating a transition away from the previous topic into a new one is only found with Tayyib (not with Tab) and is typically followed by an interrogative sentence.

Marking the transition to a related topic is performed by Tayyib and Tab, followed by an interrogative sentence. This is exemplified in Extract (6) where two male friends were discussing generosity as an important characteristic of a good relationship. S013 initiated his turn with Tab, which prefaces a follow-up question asking about gender differences regarding generosity.

Extract (6): Generosity
1	S014	2ana: ya2sorni: dayman 2arddedah fi kil majlis ya2sorni: 2ilkari:m↓ (…)
2		wallah .hh 2nah ya3ni: fi3lan hi: 2iSSifah 2illi: tghaTi: 3iyo:b irrija:l
	I really appreciate generous people; I always say that I deeply appreciate
		them
3		(…)
4	S013>	Tab 2int tsho:f 2imar2ah 2ar- 2akram min 2irajol hh willa: 2irajol 2akram↑
		Tab do you think women are more generous than men hh or men are
		more generous↑
5		(1.02)
6	S014	um- ma: lah 3alagah
	um- gender is not related

Tayyib is found in the corpus to function as an organisational device in turn-taking sequences because it can be deployed to take a turn while planning or to close the topic/turn. The next extract illustrates the use of Tayyib as a turn-taking device.

Extract (7): Sleeping habits
1	S045	2at3aTTil lama: tighiT w hi: fi: [sri:rha:] (…) mori:H mori:H
2		bass 2inah qa(bi:)H
		I cannot do anything until she falls asleep in the bed (…), it is (the new
		bed) comfy but u(g)ly
3	S046	[oh]
4		(…)
5	S046>	Tayyib um uha hh wish 2ni: bago:l bism 2illah (…) 2i:h ya3ni: ma
6		giTa3ti:ha: min 2iqailolah
		Tayyib um- er- hh what do I want to say in the name of Allah (…)
		yes, I mean you did not stop her nap
7	S045	la:↓ la: la: hh 2in hadhi: hi: ma: 2abi: 2artaH
		No↓, no, no, hh that is I want to relax

Two new mothers were talking about their strategies to help their toddlers to sleep. S045, in line 1, said that she had bought a new bed that helped her. S046 took the turn while she was still planning her utterances and prefaced her turn with Tayyib to take control of the conversation in order to develop her upcoming contribution. In this usage, Tayyib often co-occurs with (un)filled pauses and other hesitation signals, as in the given example (line 5), wish 2ni: bago:l bism 2illah ‘what do I want to say, in the name of Allah’, and it introduces a new topic, detailed description or elaboration. This usage is equivalent to well as a turn-taking device (Aijmer 2013, pp. 34–35).

3.1.3. Interpersonal and Textual Functions

In addition to the interpersonal and textual functions performed by Tayyib and Tab, speakers frequently use TAYYIB to convey their attention to the addressee while shifting the topic/action. In this usage, Tayyib and its variants are more than transitional markers; they show the speaker’s involvement with the addressees and their interest in the ongoing interaction. Transitional TAYYIB is commonly used to soften a potentially face-threatening act (e.g., the effect of disagreement and misalignment, potential conflict, unwelcome assessment and criticism) while marking the transition to a new topic or action. TAYYIB is also frequently used with declarative or interrogative sentences to introduce suggestions and with imperative sentences to mitigate the effect of the directive act, as in the following extract.

Extract (8): Shopping plans
1	S035	Tab fi:h bilmawaqi3 ma: lgi:ti:↑ 2aw hh bissho:g mathalan
		Tab there are lots of choices on the fashion websites, you did not find↑
		or hh for example, in the malls
2		(..)
3	S036	ma- madri: ma 3ajabni: min zam:n ma riHt isso:g
		no- I dunno, I did not like them, I have not visited the malls for a while
4		(2.85)
5	S035	Tab ro:Hi: bokrah (..) ya3ni: 2idha: tigdri:n (…) hh ya3ni: minha:
6		um- tidawri:n fista:n lzawa:j um- mi:n↑
		Tab go tomorrow (…) I mean if you can (..) hh I mean to um-find a dress
		for the wedding, whose wedding↑

This extract is taken from a long sequence in which two young female speakers were discussing fashion shopping options. S036 was telling her friend that she needs to buy a dress for a wedding party the following week and S035 tried repeatedly to help her by giving several suggestions. In line 1, S035 prefaced her turn with Tab, followed by a declarative statement of potential places where S036 can find good options to purchase a dress. S035 employed Tab to soften her suggestion, and in this situation, Tab can be translated to ‘what about.’ Furthermore, in line 5, S035 provided another suggestion ro:Hi: bokrah ‘go tomorrow’ and again she initiated her turn with Tab followed by the imperative sentence to minimise the effect of the directive act. Thus, the effect of the request was lessened by Tab, which was used to distinguish it from a command and modify it to a suggestion or recommendation. The impact of the potential face-threatening act was also further mitigated in her multi-unit turn with the conditional clause 2idha: tigdri:n ‘if you can’ preceded by ya3ni: ‘I mean’ as a hedging and politeness device. This function of Tayyib and Tab as a softener or mitigation marker is found in other Arabic varieties (e.g., Al-Harahsheh and Kanakri 2013; Aljutaily 2021; Alkhalil 2005; Ghobrial 1993; Ismail 2015).

This section has provided a detailed review of the functional versatility of Tayyib and Tab in an NA interaction. The next section examines the distribution of TAYYIB variants across functional and social factors.

3.2. Distributional Analysis

The number of tokens of TAYYIB extracted from the recordings was 491.

Table 3. Overall distribution of the variants of TAYYIB.

	Tayyib	Tab	Total
N	354	137	491
%	72.1	27.9	100
Normalised frequency per 10,000 words	24.7	9.6	34.3

demonstrates the overall distribution of the two variants: Tayyib (full variant) and Tab (reduced variant). Tayyib is used substantially more often than the Tab form, accounting for approximately three-quarters of all tokens and occurring 24.7 times per 10,000 words. The reduced form is far less common, accounting for approximately 28% of the total and occurring only 9.6 times per 10,000 words. The following sections examine the effects of linguistic constraints on the form variation in TAYYIB.

As found in the qualitative analysis, the discourse-pragmatic feature TAYYIB is used in a wide variety of functions. To establish whether the patterns of formal and functional variation suggest grammaticalisation, it is necessary to investigate whether the use of specific variants (full or reduced) is associated with its discourse-pragmatic functions. The chi-square test reveals that function is an extremely significant factor in the choice of TAYYIB variants (χ² = 41.597, df = 2 and p < 0.00001). Figure 1 shows that the full and reduced forms have different functional trends. The distribution emphasises the full variant’s dominance in all the discourse-pragmatic functions found in the corpus. For the interpersonal domain, performing the meaning of the agreement and acceptance of some states of things, acknowledging and asking the speaker for continuation and checking discourse progression with the addressee are all expressed with the full form. On the other hand, the textual domain reflects a greater level of form variation. When using TAYYIB as a transition marker, speakers use Tayyib more frequently than Tab. A similar distributional pattern is found when performing mitigating functions while marking the transition to a new action/topic. However, the other textual functions (i.e., using the variable as a turn-taking or topic closure device) exhibit invariability because they are only produced in the full form.

The age and gender of the speakers were examined based on the tokens’ raw token number, relative frequency and normalised frequency to determine whether the distribution of TAYYIB indicates an ongoing linguistic change in NA. The chi-squared test indicates a highly significant relationship between the variant form and the social predictors: gender (χ² = 8.5086, df = 1 and p = 0.003535) and age (χ² = 57.086, df = 2 and p < 0.00001). Figure 2 presents the variation in the production of the variable TAYYIB by both gender groups and in the two younger generations. Older-adult speakers, on the other hand, show absolute invariability because they only used Tayyib. As for the differences between the gender groups, the reduced form is less frequently used in both groups, but it occurs at a higher rate in the speech of females relative to that of males (32% vs. 18.3%). The normalised frequency, as provided in

Table 4. Distribution of TAYYIB variants across speaker gender and age.

			Tayyib	Tab	Total	%
GENDER	Female	N	232	109	341	69.5
		Normalised frequency *	30.7	14.1	44.8
	Male	N	122	28	150	30.5
		Normalised frequency	18	4.1	22.2
AGE	Older Adult	N	108	0	108	22
		Normalised frequency	24.2	0	42.2
	Adult	N	118	53	171	34.8
		Normalised frequency	25.2	11.5	36.6
	Young	N	128	84	212	43.2
		Normalised frequency	24.6	16.2	40.8

* Normalised frequency per 10,000 words.

, suggests a strong association between women and the production of both variants because it occurs 44.8 times per 10,000 words compared to only 22.2 per 10,000 words among men. Looking at the form variation in the normalised frequency in the female speakers’ data, the full form is produced more than twice as frequently as the reduced form (30.7 vs. 14.1). However, the differences between the frequencies of the variants in male speakers’ production are significantly more distinct, with the full form occurring 18 times per 10,000 words compared to the reduced form, which is used only 4.1 times per 10,000 words. These results suggest that females are at the vanguard of a linguistic shift with the highest proportion of Tab, which comprises 79.6% (N = 109) of the total tokens of the reduced form, compared to just 20.4% (N = 29) among male speakers.

With regard to age differentiation patterns, both young and adult speakers produced the full form more frequently than the reduced variant, but the percentage of the reduced form is greater among the youngest group (39.6%, N = 84) than among the adults (31%, N = 53). While adults have a slightly higher normalised frequency of Tayyib (25.2) compared to the younger group (24.6), the latter recorded a higher rate of Tab of 16.2 times per 10,000 words, compared to just 11.5 times per 10,000 words for the former. The oldest group was found to be very conservative with their production, which was limited to the standard full form with a relatively comparable rate of the normalised frequency with the two younger generations. These results propose two opposite trends in the use of TAYYIB. While the production of the Tayyib variant is gradually decreasing over time, the use of the innovative-form Tab is increasing, and this is being led by younger speakers and females. These assumptions are examined in greater depth in the next section.

Considering the distribution of TAYYIB across social groups, it is vital to examine the normalised frequency for individual TAYYIB usage in order to detect the degree of inter-speaker variation. Figure 3 displays fluctuation patterns and indicates the frequency of total TAYYIB usage for all three generations. Each single dot represents an individual female speaker, whereas the triangles indicate individual male speakers in the dataset, and the x-axis indicates their ages. Analysing the frequencies of TAYYIB confirms the variation in its production across inter- and intra-age and gender groups. The normalised frequencies vary significantly from 0 to 100 occurrences per 10,000 words, with a few notable exceptions of four female outlier speakers whose totals increased to 138 and 187 in the younger group, 168 in the adult group and 239 in the older-adult group. This extreme variance between speakers within the groups is typical of women’s usage. This finding emphasises the need to examine individual speakers as a random effect, as is dealt with in the regression analysis in the following section.

3.3. Regression Analysis

Various factors influence the choice of Tayyib versus Tab, as the descriptive results have indicated. In order to provide a more in-depth analysis to evaluate the comparative impact of the various factors and thereby answer the research questions, two statistical tests were conducted: (i) a conditional inference tree test and (ii) mixed-effects logistic regression. The advantage of using the former is that it represents the subtle interactions between the factors in the data and enables the outcomes to be visualised (Tagliamonte 2011, pp. 153–55). Mixed-effects logistic regression was applied to determine the effects and significance of the independent variables with regard to variation. It combines the fixed-effect factors with the random effect (e.g., speaker), in which the random effect is considered to be a source of variation (Baayen 2014, pp. 350–53). Thus, it provides a better understanding of the importance of each factor because it recognises the potential effects of intra- and inter-speaker variation.

R Studio was used to conduct both tests (R Development Core Team 2018). The conditional inference tree test was conducted using the party package with the help of the function ctree (Hothorn et al. 2023). The package applied for the mixed-effects logistic regression model was lme4, where the function glmer was used (Bates et al. 2015).

3.3.1. Conditional Inference Trees

Figure 4 captures the interactions between the function and social factors, revealing that age is the most significant factor influencing the choice of TAYYIB variants. With regard to age, older speakers are very conservative, using only the full form with no marked pragmatic variation. In contrast, the production of young and adult speakers differs markedly depending on the function of the form. The interpersonal–textual and textual functions are grouped into one node because they behave similarly, exhibiting high form variation as opposed to the interpersonal function on the other node, which is exclusively performed by the full form. The production of TAYYIB for interpersonal–textual and textual functions is significantly affected by the speaker’s gender. As found in the distributional results, gender plays a significant role in conditioning the variation, whereby females have significantly higher rates of TAYYIB compared to males. Both adult and younger males are more likely to produce the full than the reduced form, whereas the production of female speakers is significantly influenced by their age. That is, younger females use the reduced variant to perform interpersonal–textual and textual meanings significantly more often than their counterparts, thereby suggesting that young women in particular are leading the change towards greater use of the reduced form.

3.3.2. Mixed-Effects Logistic Regression

Some categories of independent variables have little or no variability, as demonstrated by the distributional analysis. As a result, the data were re-classified as follows in order to perform the regression analysis. Function was not included in the model according to the micro functions to avoid empty cells because some of the categories have a low frequency (i.e., topic-closure, N = 4) and/or no variation (turn-taking, was only used with the full form). Thus, the regression includes functions based on the broad functional classifications, which constitute the following: interpersonal, textual and interpersonal–textual. However, the interpersonal tokens (N = 81) were excluded from the regression because they are only performed with the full form. As a result, the impact of function was tested by comparing the textual versus interpersonal–textual functions. The effect of age was examined by including young and adult data and, thus, generational differences were tested between adults and the young. The data for older adults (N = 108) were excluded from the regression because of the invariability in their choice of variant. After excluding the interpersonal tokens (N = 81) and older-adult data (N = 108), the final dataset comprised 302 tokens (full N = 142 and reduced N = 137).

The results of the mixed-effects model, which includes the reduced form as the application value and the full form as the reference level, are shown in

Table 5. Mixed-effects logistic regression of the combined effects of factors in Tab.

AIC	BIC	logLik	deviance	df.resid
392	410.7	−191.1	382.2	297
Scaled residuals:
Min	1Q	Median	3Q	Max
−1.3970	−0.7583	−0.3735	0.7791	1.9868
Random effects:
(Intercept)	Variance	Std.Dev.	No. of obs	No. of Speakers
Speaker.ID	0.6036	0.7769	302	40
Fixed effects:
Estimate	Std.Error	z value	Pr (>|z|)	Sig.
(Intercept)	0.50526	0.38265	−1.320	0.1867
Age Reference level: Adult Young	0.86117	0.39020	2.207	0.0273 *
Gender Reference level: Female Male	−0.93081	0.39918	−2.332	0.0197 *
Function Reference level: Inter-textual Textual	0.06948	0.28573	0.243	0.8079

Sig. refers to statistical significance, where * = p < 0.05.

. These results indicate that the social factors examined are the most important factors contributing to the variation in comparison to the linguistic factors. Speakers’ gender has a substantial influence on the choice of variant. Males were given a negative estimated value (−0.93081) and are significantly less likely to use Tab compared to females who are assigned as the reference level. Similarly, age appears to have an important impact on the variation. In comparison to the base-level adults, younger speakers have a positive estimate value (0.86117) with a significant p-value, thereby indicating that the younger generation is significantly more likely to use the reduced form than adults. These results support the findings of the distributional analysis for both social factors, which show that the groups of female and younger speakers are more likely to use the phonetically reduced variant. This can be interpreted as an indication of ongoing change led by younger speakers (females in particular) who are at the vanguard of this innovation. These predictions imply inverse tendencies, that is, men and adult speakers are more likely to use the full form than those who are young and women. Unlike the social effects, none of the linguistic predictors examined were statistically significant. The difference between the textual and inter-textual functions was insignificant, noting that the tokens of the interpersonal uses were not included in the model due to their formal invariability.

4. Discussion

Regarding the overall frequency of TAYYIB, full-form Tayyib is significantly more common than the phonetically reduced-form Tab, accounting for approximately 72% of the total tokens in Najdi Arabic. However, in Jordanian Arabic (Al-Harahsheh and Kanakri 2013) and Egyptian Arabic (Ismail 2015), a reversed pattern was reported where Tab accounts for approximately 65% and 70% of all tokens, respectively. These differences in linguistic change between NA and other Arabic varieties in the use of TAYYIB may be attributed to NA’s distinctive nature of being the more conservative dialect in maintaining numerous characteristics of Standard Arabic in contrast to the dialects spoken elsewhere in the Arabian Peninsula, as proposed by Holes (2015), Kusters (2003) and Versteegh (1997) based on their assessment of phonological and morphosyntactic qualities. Holes (2006, p. 27) characterises NA conservatism by stating that ‘in contrast to the peripheral regions, it [NA] has for much of its history been relatively impervious to outside influences.’ As a result, the switch from Tayyib to Tab in NA indicates that the change is still in its earliest phases, but it is more advanced in other Arabic varieties, such as Jordanian (Al-Harahsheh and Kanakri 2013), Egyptian (Ismail 2015) and Lebanese (Ayash 2016).

In terms of social factors, the choice of variants appears to be substantially affected by the speaker’s age, and a path of change for progress could be proposed. According to the regression and distributional analysis, speakers in the two younger generations (especially those between the ages of 18 and 20 years) had a significantly higher likelihood of generating the reduced variant Tab. Older speakers were more conservative than the two younger groups and they used the standard full-form Tayyib more often. These tendencies are consistent with the variation and change found in the English ok (Heisler 1996; Schleef 2008) and well (Beeching 2016) by which the use of these discourse-pragmatic features was prompted by younger speakers. With regard to varieties of spoken Arabic, TAYYIB trends in NA are in accordance with the well-established propensity for phonological and morphological variation, with the change typically being led by younger speakers, for instance, in Ha’il (Alammar 2017), in the Eastern region (Alaodini 2019), in Najdi (Al-Essa 2009; Al-Rojaie 2013; Alajmi 2019), in Medina (Hussain 2017) and in other Arabic varieties: Algerian Arabic (Dendane 2007), Bahraini Arabic (Al-Qouz 2009), Iraqi Arabic (Ahmed 2018), Kuwaiti Arabic (Taqi 2010), Jordanian Arabic (Al-Hawamdeh 2016; Al-Shawashreh 2016; Al-Wer 1991) and Syrian Arabic (Ismail 2007). The key findings of the aforementioned research are that younger Arabic speakers are the driving force behind language change due to their frequent use of innovative forms that differ from the standard practises of their speech communities, whereas older speakers are more likely to maintain their traditional forms of speech. Similar tendencies were found by Habib (2021, 2023) who investigated the discourse-pragmatic variation in Syrian Arabic. Habib (2021) reported that the younger generation (children aged between 6 and 18 years) led the change towards greater use of the innovative form ʔinnu: (‘that’) compared to the standard form yaʕni (‘I mean’). Additionally, the findings of the current study are in accordance with the general tendency of discourse-pragmatic variation and change found in English in which younger speakers tend to use the innovative variant more than older speakers (see Childs 2021; Denis and Tagliamonte 2016; Pichler 2013, 2016, 2021; Tagliamonte and D’Arcy 2004).

Speakers’ gender has a significant influence in conditioning the choice of TAYYIB variants, where the reduced-form Tab is more associated with females’ speech across the three generations. The findings of the distributional analysis and the conditional inference tree imply considerable interaction between gender and age, whereby younger females are leading the language shift towards greater use of the innovative Tab. This finding was anticipated and it is in accordance with the traditional perspective of linguistic variation, with Labov (1990, p. 215) theorising that women are often the language change leaders and they are often associated with the linguistic change that takes place from below the level of social awareness. With regard to variation and change in Arabic, the current results are also in accordance with those of Alammar (2017), Alqahtani (2015) and Hussain (2017) concerning phonological and morphosyntactic variation in Saudi Arabia and Habib (2021) regarding structural variation in Syrian. All these studies conclude that younger females drive the shift towards the new variants.

Regarding functional variation, according to the qualitative analysis, Tayyib and Tab perform a variety of discourse-pragmatic functions, which are in accordance with the findings of Al-Harahsheh and Kanakri (2013), Aljutaily (2021), Ayash (2016), Ghobrial (1993) and Ismail (2015). The results that emerged from the conditional inference tree and Figure 1 demonstrated that the variation in TAYYIB is highly conditioned by the pragmatic function. The full form was employed for all functional classifications and strongly favoured for interpersonal meanings, whereas the reduced variant was only employed for textual and inter-textual functions with no significant connection with either. This might be taken as evidence of the pragmatic development of both forms. As for the full-form Tayyib, it has experienced gradual processes of pragmatic expansion. It has developed from its propositional meaning (i.e., as an adjective ‘good’) to a textual meaning (i.e., primarily to signal the transition in the discourse) and subsequently has expanded to an interpersonal–textual meaning (i.e., to minimise the impact of less preferred responses while indicating transition). Following this, it evolved into the interpersonal function which maintains the meaning of acceptance and agreement and originated from the underlying propositional meaning of ‘good’ and ‘kind-hearted’ of the adjective Tayyib. This later stage of pragmatic expansion allows the use of Tayyib as a free-standing token to constitute a complete turn used by listeners as a response token to express their agreement with the preceding turn/accepting state of things, which carries the metalinguistic meaning of ‘I agree with/accept what has been said.’ In a comparable manner, speakers employ Tayyib as invariant final-tags to seek their addressees’ involvement. It appears that Tayyib has gradually expanded intersubjective interpretations because it is concerned ‘with the “self” of the addressee [as] paying attention to their presumed attitudes to the content of what is said’ (Traugott 2003, p. 128). The suggested expansion path for Tayyib follows the idea of unidirectionality hypothesised by Traugott (1982, p. 256) as ‘propositional > (textual) > expressive.’4 Additionally, the proposed path for Tayyib appears to agree with the path of the Chinese hao suggested by Wang and Tsai (2005). They assert that hao appears to follow Traugott’s (1982) historical steps, whereby hao evolved from the adjective to textual meanings as a closure signal, which subsequently developed into interactional meanings as the agreement/concession marker. The anticipated trajectory of Tayyib could be further proven by age association patterns with identifiable functional tendencies, as shown in the distributional data. Using Tayyib for interpersonal functions was entirely developed by the two younger generations, particularly adolescents under 21 years of age, who accounted for approximately 71% of the total interpersonal tokens. Conversely, the usage of the older-adult speakers (55+ years of age) was restricted to textual and inter-textual functions. This suggests that interpersonal meaning is the most recently acquired meaning for Tayyib and it could increasingly proceed to acquire new meaning in the near future.

The development of TAYYIB from the adjective (‘good’) to a discourse-pragmatic feature (‘okay, well, right’) appears to coincide with the cline of decategorialisation described by Hopper and Traugott (2003, p. 107) as the ‘major category (>intermediate category) > minor category,’ whereby the intermediate category consists of adjectives and adverbs, whilst the minor category of ‘closed-class items’ includes prepositions, conjunctions, auxiliary verbs, pronouns and demonstratives. According to Ferrara (1997, p. 371), the category of discourse marker is regarded as a minor category. Furthermore, Tayyib as a discourse-pragmatic feature has lost its ability to be inflected for number or gender. The loss of ‘morphological and syntactic properties [includes losing] the ability to be inflected or take on derivational morphology’ is a notable characteristic of decategorialisation according to Heine and Kuteva (2007, p. 40).5

The Tab variant exhibits multiple indicators of grammaticalisation given its semantic bleaching, pragmatic expansion and phonetic reduction. Tab experienced desemanticisation, which is defined by Heine (2003, p. 579) as ‘the loss in meaning, loss of lexical content or semantic reduction’ and as ‘weakening of meaning’ by Hopper and Traugott (2003, p. 94), whereas Lehmann (2015, p. 134) describes it as the ‘decrease in the semantic integrity of a sign.’ The qualitative analysis showed that the adjectival meaning in Tab is bleached out; thus, Tab is not used as a single response, compared to the full form, to signal acceptance. Heine and Kuteva (2007, pp. 39–40) suggest that desemanticisation is an immediate consequence of pragmatic extension, which entails employing the linguistic item in a new context. Furthermore, Tab has undergone phonetic reduction, whereby it becomes monosyllabic [tˁab] from the original disyllabic adjective [tˁajjib]. The steady loss of phonological substance, which is referred to as phonological attrition (Hopper and Traugott 2003; Lehmann 2015, p. 134), is one mechanism of grammaticalisation (Heine 2003). It is also known as phonetic erosion, which is defined as the loss of phonetic segments (Heine 2003, p. 585), including the loss of full syllables (Heine and Kuteva 2007, p. 43).

The supporting evidence regarding the variation from the apparent-time data reveals that the linguistic changes in semantic, pragmatic and phonological features in TAYYIB have taken place over a relatively short period of time and have been prompted by younger speakers, particularly females. The variation and linguistic change in Tayyib demonstrate many characteristics of grammaticalisation, namely, decategorialisation, pragmatic expansion and semantic weakening. Nevertheless, the inherent significance of Tayyib may still be recognised in some instances. This reflects the concepts of persistence described by Hopper (1991, p. 22) as ‘some of the traces of earlier meanings of an item undergoing grammaticalization are likely to survive in the form of the grammatical distribution of the item concerned.’ In comparison, the reduced Tab demonstrates the most advanced stages along the grammaticalisation cline, mainly the semantic bleaching, pragmatic expansion and phonetic reduction. Nonetheless, the data indicate that Tab has a higher frequency and wider interpersonal usage in Jordanian and Egyptian Arabic, suggesting that grammaticalisation in NA is likely to still be underway but in a considerably slower and more moderate pattern. The co-existence of both Tayyib and Tab reflects the phenomenon of ‘layering’, whereby ‘the older layers are not necessarily discarded when new layers emerge but may remain to coexist with and interact with the newer layers and interact’ (Hopper 1991, p. 22).6

5. Conclusions

The aim of this sociolinguistic study was to determine how the phonetic realisation of the TAYYIB differed between full and reduced forms in NA. The linguistic and social impacts on the variation in these forms have been investigated in this paper. It was important to establish whether social patterns of variation represent ongoing change, as well as if phonetic and pragmatic changes in discourse-pragmatic aspects suggest grammaticalisation. According to the qualitative examination, TAYYIB fulfils a wide variety of discourse-pragmatic functions. TAYYIB variation is consistent and is influenced by functional and social factors according to the quantitative analysis.

The current investigation was unable to review the development path of TAYYIB because it was based on apparent-time data. Therefore, diachronic data will be required to create a comprehensive view of the grammaticalization cline of TAYYIB in Arabic. Investigating the change in the use of TAYYIB in other varieties of Arabic would improve comparability between related dialects and provide insight into the mechanism of discourse-pragmatic variation in Arabic varieties. Because the current study and the previous research have examined TAYYIB in conversational data, there is a need to study the variable in other interactional settings and text genres, considering the intrinsic multifunctionality and context dependency of the discourse-pragmatic features.