Grammaticalization Cases Through the Four Linguistic shifts Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) Online version available in : http://arbitrer.fib.unand.ac.id | 2339-1162 (Print) | 2550-1011 (Online) | JURNAL ARBITRER SubmiSSion Track A B S T R A C T Recieved: October 22, 2020 Final Revision: December 31, 2020 Available Online: April 25, 2021 The present study aims at describing the grammaticalization phenomenon of some Hijazi Arabic (HA) elements that, to the best of the author’s knowledge, have not been dealt with. It argues that the elements xən, ʕsəs and lissa have gradually developed from the verb xalla, the preposition phrases (PP) ʕla asa:s and ila al-sa:ʕah respectively. To achieve this, the paper speculates that these developed forms have followed the prevalently agreed upon grammaticalization chains where they have undergone desemanticization (semantic shift), decategorialization (morphosyntactic shift), cliticization (morphophonological shift) and erosion (phonetic shift). These chains exhibit the grammaticalization stages of the elements under study. keyword Grammaticalization, linguistic-shifts, desemanticization, decategorialization, cliticization, erosion correSpondence E-mail: m.zah@tu.edu.sa Articles Grammaticalization Cases Through the Four Linguistic shifts Mohammad Ali Al Zahrani1 1 Associate Professor, Foreign Languages Department, Faculty of Arts, Taif University, Taif, Saudi Arabia Under Liscense of Creative Commons Attributioni-NonCommercial 4.0 International.DOI: https://doi.org/10.25077/ar.8.1.1-13.2021 I. INTRODUCTION Grammaticalization, as a linguistic term, was coined by Meillet (1912) and is used in the literature to convey two main meanings; one refers to the research framework accounting for the change, the other is the phenomenon itself (Hopper & Traugott, 2003: 2). The current paper uses the latter meaning where grammaticalization is a linguistic phenomenon referring specifically to some steps by which some elements have become more grammatical. More specifically, the paper describes some Hijazi Arabic (HA) lexical forms that have undergone some linguistics shifts (Section 4) to serve grammatical functions. The paper should contribute to the HA studies on grammaticalization. Several linguists have been exploring different linguistic aspects of HA (see, for example, Al Zahrani, 2020 for more studies on HA). However, with respect to grammaticalization, it has not received much attention. Eifan (2017) has explored some grammaticalized elements including ga:m ‘stand’, which marks the inceptive aspect, gaʕad ‘sit’, žālas ‘sit’, and fiḍil ‘remained’, which mark the progressive aspect. The syntactic distribution, the selectional properties and the hierarchical placement of these aspectual elements have been explored in Al Zahrani (2016). Also, some studies on HA have shown that the prefixes b-, derived from abγa ‘I want’, and ħ-, derived from ra:ħ ‘went/has gone’ have been grammaticalized to either mark future time references (Al Zahrani, 2013, 2020; Eifan, 2017) or hypothetical and counterfactual interpretations (Al Zahrani, 2020). Furthermore, Eifan (2017) has concluded that the prepositions ʕind ‘at’, maʕ ‘with’, and fi: ‘in, at’ show some grammaticalization features where they behave as markers for the possessive function: a function that can also be coded by the grammaticalized nouns ħagg and tabaʕ ‘belong to’. Another grammaticalized form that Eifan (2017) has explored is the subordinate conjunction ʕašān ‘because’. These previous HA studies have shown, to some extent, that the elements have undergone some grammatical changes, i.e., they have developed, on the one hand, from being purely lexical to being purely grammatical or moving 1 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) towards the grammatical pole; or from being less grammatical to more grammatical, on the other hand, as is the case with the afore-mentioned prepositions. The significance of the present study is that it particularly contributes to the HA literature by exploring some more forms/phrases that the grammaticalization studies to date, to the best of my knowledge, have not dealt with. The present study includes the verb xalla ‘allow’ and the prepositional phrases ʕala asa:s ‘on the basis of’ and lisa:ʔah ‘to the moment/time of’. The study also contributes to the field of grammaticalization, in general, as it uses the mechanisms of the grammaticalization phenomenon to describe the dialect of question. In essence, the paper shows that the forms under study have followed a grammaticalization path that results in some linguistic shifts by which they serve grammatical functions. For the descriptive sake, following Heine (1993), these linguistic shifts are crosslinguistically seen as four chains including desemanticization, decategorialization, cliticization and erosion (See section 3). The findings show that the verb xalla has developed to cover a variety of modal meanings ranging between permission, obligation, suggestion, adhortative, threatening, and imperative. The imperative form of xalla has developed a form functioning as an introductory particle. The prepositional phrases ʕala asa:s ‘on the basis of’ and lisa:ʔah ‘to the moment/time of’ have developed constituents functioning as cause and purpose subordinators, and as negative adverbial markers. However, due to the absence of diachronic evidence for the grammaticality developments of the elements under study, the paper presents the shifts and changes of the elements according to what appears to me from their morphophonological features. That is to say, all the developments of each form show constituents existing in HA and thus the order of these constituents is presented according to their morphophonological features at best. This method is not so unusual; as a matter of fact, investigating the variations and changes of the language through time without its diachronic data or any written documents is impossible (see, for example, Jarad’s (2013, 2014, 2017) and Fleischman’s (1982) studeis, to mention a few). This being said, unless otherwise stated, the examples come from the author’s intuition and/or introspection. The organization of the paper is as follows. Section 2 presents grammaticalization as a linguistic phenomenon. Section 3 introduces the theoretical assumptions needed in the current study. Section 4 and its subsequent Sections (4.1 & 4.2) explore the HA elements under study, and Section 5 concludes the paper. II. THE GRAMMATICALIZATION PHENOMENON Hopper and Traugott (2003: 5) defines the phenomenon of grammaticalization as “the process whereby lexical items and constructions come in certain linguistics contexts to serve grammatical functions, and, once grammaticalized, continue to develop new grammatical functions”. Thus, as a phenomenon, grammaticalization is concerned with “how grammatical items develop grammatical functions” (Hopper & Traugott, 2003: 7). Being developed to serve grammatical functions does not always imply a full change of the form from being lexical into grammatical. Put differently, Haspelmath (1999: 1044) argues that the most general definition of grammaticalization would not restrict the phenomenon to a full change of a lexical form to a grammatical form, but would admit that grammaticalization “shifts a linguistic expression further toward the functional pole of the lexical- functional continuum”. Against this background, grammaticalization shows some linguistic shifts of contentive elements via a grammaticalization path by which they serve grammatical functions. Notably, the grammaticalization path, as pointed out by Heine (2003: 589), is referred to in the literature as ‘pathway’ (Bisang, 1986), ‘path’ (Bybee et al., 1994: 14), ‘grammaticalization chain’ (Heine, 1993: 53), and ‘cline of grammaticality’ (Hopper & Traugott, 2003: 7). The paper uses these terms interchangeably. The grammaticalization path shows a gradual language change that Bybee (2015: 10) claims to be “an inevitable outcome of language use”. This gradual change, more or less, results in the loss of the semantic and phonological properties of the lexical items, as well as their syntactic freedom and morphological features (Norde, to appear in ORE : 1). Hence, the change that grammaticalization shows 2 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) is actually a composite change that encompasses simultaneous or successive “microchanges” at the different linguistic levels: syntax, morphology, semantics, and phonology (Norde, 2012; Norde & Beijering, 2014). These linguistic changes, as suggested by Hopper and Traugott (2003:7), as well as by many others (see references above), follow a “cline of grammaticality” that presents continuous series of changes that the lexical elements may undergo. For instance, Hopper and Traugott’s (2003:7) cline in (1) suggests that the items occurring to the very right edge of the cline are more grammatical than those in a preceding level of the cline. Hopper and Traugott (2003: 99) admit that grammaticalization is assumingly unidirectional. Put differently, the left end of the cline is less grammatical than the right one. 1. content item > grammatical word > clitic > Inflectional affix The literature shows dozens of examples that follow the cline in (1) (see, for example, Al Zahrani, 2020; Eifan, 2017; Hopper & Traugott, 1993, 2003; Jarad, 2014, 2017; Norde, 2012; Norde & Beijering, 2014, to mention a few). Some typical examples include the change of some lexical motion verbs into future markers as is the case with the English to be going to, the French aller and the Dutch gaan (Norde, to appear in ORE ); this is the case in all Roman languages (see, Fleischman, S. (1982), for more examples). Simialr cases are present in other langugaes such as the HA raaħ and abγa (Al Zahrani, 2020; Eifan, 2017), the Levantine varieties: bi-widd (Mitchell & Al-Hassan, 1994: 19), the Yameni Arabic yaʃaʔ (Versteegh, 2014), the Moroccan Arabic bat (Stewart, 1998: 110). Other examples also include the development of the indefinite articles out of the numeral ‘one’ as is the case in English a(n), French un(e), and Dutch een (Norde, to appear in ORE). An important point to be considered here is that the concept of unidirectionality does not necessarily require all crosslinguistic elements undergoing grammaticalization to move through all the stages of the cline of grammaticality, and this is evidenced by many grammaticalization studies (I show this with the grammaticalized constituent lissa, Section 4.2). Once again, the aforementioned studies have shown that the cross-linguistic elements follow the same cline of grammaticality, which in turn suggests a cross-linguistic regularity, which is an important aspect of grammaticalization. Heine (1993, 1997, 2003) argues that the crosslinguistic regularity is resulted by the cognitive processes that, as the examples have shown, cannot be language-specific. This is also evidenced by the regularity across languages where the concepts of some body parts are extended to function as locative markers (Heine, 1997, 2003; Heine & Kuteva, 2002). Examples include the English back, the HA Dahr ‘back’, the Danish bag and the Old Danish baker (Norde, to appear in ORE ) that can express the special notion of ‘behind’. Also, the HA waʤh ‘face” and ra:s ‘head’ have been developed as locative markers expressing front and the higher spatial notions. This concludes that grammaticalization is a linguistic phenomenon exhibiting a fact where concrete meanings develop to more abstract concepts (Heine & Kuteva, 2002). Having presented this overview, the paper now shifts to briefly present some theoretical assumptions needed for the present discussion of the grammaticalized elements. III. THEORETICAL ASSUMPTIONS: LINGUISTIC SHIFTS Following Heine (1993: 54), when contentive elements start to grammaticalize, they exhibit a number of linguistic shifts, each of which constitutes a distinct chain/continuum. Each chain shows a specific linguistic behaviour and this signifies that the chains are related to the four aspects of linguistics, referred to in the literature as ‘stages’ or ‘mechanisms of change’. These chains are the semantic shift (desemanticization), the morphosyntactic shift (decategorialization), the morphophonological shift (cliticization) and the phonetic shift (erosion). This section provides a brief definition of each shift and presents some of its basic and salient features. Desemanticization, coined by Heine and Reh (1982), is the first shift in a grammaticalization process. It is known as the semantic shift, and semantic bleaching. It is believed that early stages of grammaticalization must exhibit a shift in meaning 3 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) where the form gradually and slowly strips out from its contentive meaning (Hopper & Traugott, 2003). Heine (1993: 54) defines the semantic shift as the “process whereby in specific contexts a lexical item is emptied of its lexical semantics and acquires a grammatical function”. However, according to Heine and Kuteva (2002), desemanticization must be triggered by extension: context generalization, i.e., the element is used in new contexts where its meaning is extended to cover other more meanings and consequently increases its frequency. Hence, they argue that extension is the basic principle in grammaticalization. This paper considers extension an indivisible part of the semantic shift so that it treats them together under desemanticization. The term decategorialization was introduced by Hopper and Thompson (1984). Heine (1993: 55) classifies this shift as the second stage of a grammaticalization process. Decategorialization is a morphosyntactic shift in which a form increasingly loses some morphological and syntactic properties as well as some privileges of its grammatical category. Cliticization is the third linguistic shift of a grammaticalization process whereby a form gradually develops “to an operator of another item” (Vangaever, 2019: 260) in that it forms a part of the morphophonological structure of its complement, indubitably, after having lost its lexical content (Heine 1993: 55). In this shift, the linguistic element forms an independent constituent that is separate from its complement and then develops into a clitic. Erosion exhibits the fourth linguistic shift of a grammaticalization process by which the linguistic element undergoing grammaticalization loses its phonological substance and does not carry distinctive tone or stress (Heine, 1993: 56) and this may lead to fusion. Erosion is the last to happen in the linguistic shifts. Following this brief discussion of the linguistic shifts, it is important to emphasize that not all these shifts are relevant in each grammaticalization process as is the case in some of the elements under study. However, the semantic shift, including extension, is, on the one hand, the primary stage that “precedes all other shifts” (Heine 93: 58). On the other hand, the phonetic shift ‘erosion’ is not a necessary process in grammaticalization, and if it happens it is “the last to apply” (Heine & Kuteva, 2002: 42). In view of this discussion, when an element undergoes grammaticalization, extension implies that it gains some more properties due to its use in other linguistic contexts though it loses (some of) its semantic and syntactic (and probably phonetic) properties that it has as a pure lexical element. Notwithstanding, grammaticalization does not insinuate the elimination of the lexical item. Hopper (1991) argues that when a lexical element is grammaticalized, it may remain “as an autonomous element and undergo the same changes as ordinary lexical items” (Hopper, 1991: 22). This is the characteristic known as “divergence” (Hopper & Traugott, 2003: 118) or “split” (Heine & Reh, 1984: 57). In light of these linguistic chains, the paper now shifts to the discussion of the HA elements under study. IV. HA GRAMMATICALIZED ELEMENTS AND THEIR LINGUISTIC SHIFTS Due to the different behaviour of the elements under study, the paper explores them in two subsequent sections. It aims to explore the linguistic behaviour of the elements under study: the verb xalla and the prepositional phrases ila al-sa:ʕah ‘to the moment/ time’ and ʕla asa:s ‘on the basis of’. It also shows that their behaviour has undergone some shifts that consequently have developed some grammatical elements. Subsection (4.1) starts with the morphological derivation of the various forms of xalla and its inflectional properties. Then, in subsection (4.2), the paper discusses the prepositional phrases ila al-sa:ʕah ‘to the moment/ time’ and ʕla asa:s ‘on the basis of’ and shows that they are the source from which the grammaticalized lissaʕ and ʕasa:s have been developed. 4.1 The Verb xalla Following the majority of linguists including Ryding (1994, 2014; 2005), Bahloul (2008) and Al Zahrani (2013), the Arabic perfective form is the basic verbal form since it shows the least amount of inflection when derived from a radical 4 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) verbal root. Accordingly, the verb xalla is the basic form from which other root-related derivatives are created. Like other HA lexical verb forms, the basic form xalla derives perfective, imperfective and imperative forms inflecting for the phi features as the following three paradigms show (see Al Zahrani 2013 for more on HA morphology). The examples in (2a-d) show the perfective and imperfective forms of the verb xalla along with their inflectional properties indicating the grammatical features of tense, aspect, voice, person, number and gender. The basic notion of the verb xalla in these examples is permission/allowing. Further to this basic notion, the verb can be used in a causative function as in (3). Table 1. The Perfective Paradigm of xalla Table 2. The Imperfective Paradigm of xalla Table 3. The Imperative Paradigm of xalla Notice that all the subject pronouns appearing under the three Basic notion columns in the paradigms are indicated by the agreement morphemes. The basic notion of the verb xalla in HA indicates the act of allowing something/someone to do something or act in a specific way as the examples in (2a-d) show. Notice that the colon diacritic /:/ following vowel sounds indicates the length of the vowels. 2 3 Both the perfective and imperfective forms in (3) express a causative meaning; this implies an extension of the verb’s basic notion of permission. The verb is used in other general contexts where its meaning is extended to cover other interpretations. Furthermore to this causative-meaning extension, context generalization cases are provided with the imperative form of xalla in (4). 4 The example in (4a) can be causative, but it can also be used in different scenarios to convey different meanings. It can give permission for Nada to write, an obligation to the addressee that Nada should write, or a suggestion to the addressee to invite Nada to write. What we may assume in these causative and modal senses is that the addressee may have authority upon Nada to cause her, allow her, oblige her or invite her to write. This, however, is not the case in (4b) where neither the speaker nor the addresses can stop the rain. The meaning here can be a suggestion or a request from the speaker 5 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) to the addressee to wait until the rain stops. In like fashion, in (4c) the speaker asks the addressee to ‘leave it, wait’ until tomorrow. The examples in (4b-c) present cases where the basic meaning of the verb has extended to cover suggestion or request interpretations. Example (4d) exhibits another meaning where the speaker expresses his acceptance/satisfaction. In this example one can consider the verb as a particle to convey the meaning of ‘allow it, leave it’. In (4e), we can suggest a scenario where some students are asking their teacher to make the exam tomorrow instead of today as they haven’t prepared well. While the examples in (4a-e) can employ the other forms of the verb xalla, the examples in (4f-g) are particular to the imperative form xall. Examples (4f-g) show that xall conveys the meanings of ‘let alone’ and ‘be/stay’ respectively. The salient point that one can draw from the examples in (4a-g) is that they provide clear evidence for further meaning extension cases. That is, they exhibit more semantic meanings to the verb in question beside its permission and causative meanings expressed in (2a-d) and (3a-b). In relation to the morphosyntactic properties of the verb, the above examples show that the verb inflects and that the subject of the verb is indicated by the inflections signifying the phi features. However, the verb can also show a different case where the plural inflectional suffix –na can refer to third person singular features as in (5). of xall-i-na, as in (5c), by a referring expression such as ana wa Ali (I and Ali). Thus, the referring expression reinforces the meaning that –na is a plural marker. This analysis is not unusual, and it is on a par with the use of the English lets (without an apostrophe) as shown in (6) and (7), adopted from Hopper and Traugott (2003: 10). 6. “Lets give you a hand. (i.e., let me give you a hand.)” 7. “Let’s you and I take ‘em on for a set.” Quirk et al. (1985: 830) argue that the lets here is “no more than an introductory particle”. The presence of the first person plural in (7) is reinforced by you and I, and in this case the verb appears with an apostrophe: let’s. With reference to the four chains, examples (4a-e) and (5a-c) demonstrate desemanticization, i.e., they clearly show the semantic shift that unequivocally features the extension of the basic notion to other notions in other general contexts. Desemanticization is the first shift in a grammaticalization process. It involves a shift in meaning, which occurs only in a highly specific context (Hopper & Traugott, 1993, 2003). Stated differently, the contextualized scenario due to the frequent use in other contexts gives the rise for the appropriate interpretation, be it permission, obligation, suggestion or request. In this connection, the element undergoes the conceptual shift that Hopper and Traugott (2003: 51) describe as the first obligatory step in grammaticalization. The conceptual shift that we have noticed is actually triggered by the semantic generality of the verb, which is an important semantic property for its grammaticalization (Hopper & Traugott, 2003: 101). What the semantic generality implies is that lexical meanings that are subject to undergo grammaticalization are “quite general” and “basic words”, but not very specific and specialized terms such as whisper, chortle or squirm (ibid). The semantic generality of the verb of question is evident and it has shown that it is a quite general and basic word whose basic notion has been extended to cover a range of other meanings. Further to this conceptual shift, other frequent uses of the verb provide ample evidence for extending its basic meaning to adhortative — suggestion/ encouraging interpretations, again, in a highly specific context. This latter meaning has more 5 Both (5a) and (5b) show the third person plural suffix –na and this is compatible with the first translation of each example. However, the second translation of each example shows that – na can be used to refer to first person singular subject. This reference implies some ambiguity. Notwithstanding, the ambiguity resulted by (5a- b) can be solved by emphasizing the plural object 6 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) grammaticalization features as shown next in examples (8) through (12). The decategorialization of the verb is witnessed in (5) where the use of xall-i:-na ‘you-let- us’ refers to first person singular in a context expressing adhortative meaning. Thus, in addition to the foregoing semantic shift, (5) shows a morphosyntactic shift where the verb has lost some morphological and syntactic properties of its grammatical category. That is, the verb xall selects for the plural suffix –na not to indicate a speaker with first plural agreement features, but with first person singular features. This exhibits a remarkable stage of the grammaticalization of the verb of question as shown in (8). lexical verb following it. Notice that (9a-c) have the same translation: “let me write”. The linguistic shifts that the verb has undergone are represented in (10). 10. xall-i:-ni > xall-i:-na AND xall-u:-na > xalləna > xənna > xən In addition to the cliticization shift, the linear representation in (10) also shows the fourth shift: erosion, in which the verb, as suggested by Heine (1993: 56), has lost its phonological substance and does not carry distinctive tone or stress. If this analysis is right, the paper argues that xən functions as an introductory particle basegenerated in the left periphery of the clause after it has been stripped out from its basic notion, morphosyntactic and morphophonological features. So far, the grammaticalization process of the imperative form of the verb xall and its agreement features xall-i:-na and xall-u:-na shows that its use to indicated permission and obligation has extended to other more frequent uses that result in both semantic and morphosyntactic shifts. Following these two shifts, there are two other successive shifts, namely– the cliticization and erosion shifts where the verb functions as an independent constituent separate from its following constituents and then develops into a first person singular proclitic that has lost its phonological properties as shown in (10). These shifts are on a par with Bybee and Pagliuca’s (1985: 76) claim that “as the meaning generalizes and the range of uses widens the frequency increases and this leads automatically to phonological reduction and perhaps fusion”. Further to this, the grammaticalization of the expressions xalli:na and xall-u:-na to xən can be traced via the cline of grammaticality, represented in (1), where xən has been grammaticalized from the basic-form content item xalla into the grammatical element xən. Evidence for the grammaticalization of xall-i:- ni ‘you-let-me’ into the introductory particle xən springs from the fact that the particle xən can also be used in examples where its following lexical verb carries first person singular features as is presented in (11a), which is the counterpart of (9c) above. 8 The paradigm in Table 3 shows that there are only two imperative forms, which are used in (8a-b). Using Quirk et al’s (1985: 830) expression, these two imperative forms can be used as “introductory particles” with first person singular reference although they exhibit the plural suffix –na. This being the case, the previously mentioned features of the plural suffix –na, as well as the person, number and gender agreement morphemes –i– and –u–, represented in Table 3, are lost as shown in the different verb forms in (9a-c). One can consider the first person plural features expressed by –na frozen. Therefore, from now on, I gloss the verb xall as PRT (particle) without any hyphens. 9 The example in (9a) shows a morphophonological shift where the agreement morphemes –i– and –u– have shifted to the schwa sound /ə/ that is followed by the plural suffix –na (xalləna). (9b) shows another morphophonological shift in which more sounds have been deleted so the resulting form is xənna. (9c) presents another stage where the form has been reduced to xən and encliticized to the 7 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) In HA, a speaker may convey the meaning of “let me write” by uttering the statement in (9c) above, which has the first person plural morpheme n- attached to the verb –uktub ‘write’ = (n-uktub) [literally ‘we- write’], or the statement in (11a), which has the first person singular morpheme a- attached to the verb –ktub ‘write’ = (a-ktub) [literally ‘I-write’]. It should be mentioned that while a-ktub [literally ‘I-write’] can follow the grammaticalized from xən, i.e., the introductory particle, it cannot follow the form xall-i:-na [literally ‘you-let-us’] since the agreement features of the two verbal forms are incompatible. This fact provides another piece of evidence for the grammaticalization of xən. Another piece of evidence supporting the argument of the grammaticalization of xən is the fact that it can be followed by any perfective/imperfective form derived from its basic-form content item xalla, introduced in Table 1 and Table 2, as shown in (11b). In this latter example, both the grammaticalized element xən functioning as an introductory particle and the imperfective form a-xalli:kum co-occur. In (11b) the introductory particle xən conveys the adhortative interpretation (suggestion), whereas a-xalli:kum, according to its context, may convey either the causative interpretation or permission. Further to this analysis, (11b) also shows that the grammaticalization of xən does not eliminate the lexical derivatives introduced in Table 1 and Table 2. This shows the “divergence” (Hopper & Traugott, 2003: 118) or “split” (Heine & Reh, 1984: 57) characteristic of the basic form xalla (and its derivatives) in that it still exists in the dialect and may undergo more grammaticalization cases. On the flip side, the particle xən has been grammaticalized as a first person singular proclitic. However, it is not used for other persons (second or third). This explains the ungrammaticality of (11c). At the present juncture, it is worth reiterating that grammaticalization is a continuous process and grammaticalized elements do not stop at a specific stage as they may “continue to develop new grammatical functions” (2003: 5). This claim is evidenced in HA by Al Zahrani’s (2020: 176) argument asserting that although the HA grammaticalized elements b- and raaħ developed as future markers, they continued to develop other hypothetical functions. This being said, there are signs that xən can be used for other persons if and only if an adverbial element or a pronominal expression intermediates between the introductory particle xən and the verb carrying the agreement features of the second/third person as the examples in (12) illustrate. 11 12 (12) clearly shows the function of the element xən as an introductory particle in the left edge of the clause introducing the proposition following it. In these examples xən appears with verbs carrying second and third person agreement features. So far we have seen the linguistic behaviour of the verb xall across the four chains (shifts), and how it has developed some grammatical constituents through its grammaticalization journey, one of which is xən that serves as an introductory particle. The next subsection explore the PPs ʕla asa:s and ila al-sa:ʕah. 4.2 The Prepositional Phrases ʕla asa:s and ila al-sa:ʕah I briefly show that these two PPs have gone through some linguistic shifts that resulted in the two grammaticalized elements ʕsəs and lissaʔ. The expression ʕla asa:s is composed of the preposition ʕla ‘on, upon’ and the noun asa:s ‘base, basis, core, fundamental’ that constitute the preposition phrase (PP) [PP ʕla asa:s] meaning ‘on the basis of, on the fundamental that’. In like fashion, [PP ila al-sa:ʕah] is built from the preposition ila ‘to’ and alsa:ʕah ‘moment, hour, time’ and conveys the meaning of ‘to the moment of’. The literal use of the lexical phrase ʕla asa:s ‘on 8 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) a base’ and its plural variant ʕla usus ‘on bases’ refers to concrete bases/objects as in (13). Example (15a) shows the cause for the forgiveness whereas (15b) shows the purpose behind the forgiveness. Both cause and purpose are interpreted by the same PP expression ʕla asa:s. In their product of ten years of research, Heine and Kuteva (2002) have provided hundreds of grammaticalized elements, some of which are elements functioning as subordinators of cause and purpose clauses (see Heine and Kuteva (2002: 210-213)) as is the case with the HA element in question. Following Heine and Kuteva’s (2002) argument, the cause and purpose interpretations must have evolved significantly from the quite frequent usage of this PP, which in turn results in the semantic shift (desemanticization) featuring the meaning extension and the metaphoric usage of the PP ʕla asa:s. With reference to the cline of grammaticality, represented in (1) above, the grammaticalization journey of the PP ʕla asa:s can be summarized as in the linear development in (16). 16. ʕla asa:s > ʕlasa:s > ʕasa:s > ʕəsəs > (ʕ)səs What this linear development shows along with the metaphorical interpretations in (15a-b) is that Stage 1 in (16) started with two separate elements forming the PP. That is, the content item asa:s is basically selected as a complement by the preposition ʕla (Stage 1) that has been agglutinated to its complement (Stage 2). This latter stage discloses that the two elements have been united to create one syntactic unit functioning as a subordinate conjunction. The agglutinating process has resulted in a new grammatical element: a subordinator of cause and purpose. This process shows a morphosyntactic shift (decategorialization) by which the PP category has vanished, and this accounts for the ungrammaticality of (17) where I gloss the subordinator of cause and purpose as SUBO. 13 This literal interpretation of the content phrase ʕla asa:s and its plural variant becomes more generalized when used metaphorically. Traugott and Dasher (2002: 27) argue that desemanticization happens when metaphorical uses are involved. The illustration of this claim is shown by the phrases in (14a-b) where the PPs do not refer to concrete objects. 14 Recalling that context generalization gives rise to expanding the frequency of the expression, examples (14a-b) present context generalization cases wherein the basic meaning of the PP ʕla asa:s is extended to cover other contexts. By way of illustration, the basic meaning of a physical base, as in (13), is extended to metaphorically cover more abstract bases: aluxuwah ‘brotherhood’ and alsˤuħbah ‘friendship’. The NPs that can replace the abstract bases (14a-b) to convey metaphoric interpretations are definitely infinite and this suggests a high frequent usage of the PP. The metaphorical use of the PP is in the vein of Heine and Kuteva’s (2002) argument asserting that in grammaticalization concrete meanings are developed to more abstract concepts, and this is due to the semantic generality of the expression. Both the extension of the basic notion of a concrete basis and the frequent metaphorical uses of the PP ʕla asa:s trigger the semantic shift. This conclusion is corroborated by the two interrelated interpretations of cause and purpose illustrated by (15a-b) respectively. 15 179 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) The decategorialization of [PP ʕla asa:s] is reflected in the fact that the element ʕlasa:s has undergone some changes by which it has lost some of its morphosyntactic properties and lost the privileges of the original PP category; hence, it cannot function as a PP as shown by the ill-formed clause in (17). By pointing out that the phonological reduction follows the conceptual shift (Heine, 2003: 52), the reduced form ʕlasa:s is precluded when ʕla asa:s is used in its lexical meaning; this is supported by the ungrammaticality of (17). The preclusion of reduced forms is further evidenced by the fact that while the grammaticalized elements in (16) can replace the grammaticalized expression ʕla asa:s in (15a-b), only the lexical meaning of the PP ʕla asa:s is allowed in (17), as is the case in (13a-b). This suggests that there are two expressions: the grammaticalized ʕla asa:s that serves a grammatical function (Cf. the cause and purpose function in (15a-b)), and the lexical ʕla asa:s that expresses the basic lexical notion of a concrete base (Cf. (13a-b)). This linguistic paradox of precluding the reduced forms from conveying lexical meanings can be further supported by examples from English where the ungrammaticality of (18d) is attributed to the fact that the reduced grammaticalized constituent is gonna in (18c) cannot express the lexical meaning of is going to in (18a), although is going to can also function as a grammatical expression as in (18b) (examples adopted from Heine (2003: 51)). ʕasa:s. By this stage, the resulting structure ʕasa:s shows an independent constituent that is discrete from the original PP. Further to these grammaticalization chains, the final two stages (Stage 4 and Stage 5) exhibit the pure cause and purpose subordinate forms ʕəsəs / (ʕ)səs that no longer convey the lexical meaning of ʕla asa:s. Beside the previous phonetic reductions that happened simultaneously with agglutination, these two stages clearly show the phonetic shift (erosion) that the constituent has undergone. Stage 4 has two instances of the schwa sound whereas Stage 5 shows only the final schwa with the possibility of the deletion of the voiced pharyngeal fricative sound: the clitic ʕ-.These two stages demonstrate how in the phonetic shift the lexical content loses its phonological substance and and does not carry distinctive tone or stress (Heine, 1993: 56). Having discussed the grammaticalization of the cause and purpose subordinate conjunction across the four chains and seen its linguistic behaviour as a grammaticalized element, the section now shifts to the PP lisa:ʕah ‘to the moment’. To avoid repeating the facts discussed so far and to be more succinct, I start by the linear development in (19) that I assume the PP lisa:ʕah has gone through to develop the adverbial negative element lissaʔ ‘not yet’. 19. laa lisa:ʔah > laa-lissaʔ > lisa:ʔ-(X) > lissa(ʔ) The grammaticality development of lissaʔ represented in (19) suggests that it started out as a negated PP constituent: laa lisa:ʔah ‘no, to this moment, to the moment of’. Underlyingly, this stage shows the negative laa ‘no’, the preposition li- ‘to’ and the noun phrase al-sa:ʔah ‘moment/time/ hour’ = [NEG laa [PP li [NP al-sa:ʔah]. Notice that the lexical meaning of the PP lisa:ʔah ‘to this moment, to the moment of’ can be used in affirmative and negative statements as shown in (20). 18 (18) shows that there are the lexical expression is going to indicating a physical movement/motion and the grammatical expression is going to marking grammatical features. Following the semantic and morphosyntactic shifts, the final product in (16) (Stage 5) provides ample evidence for the other grammaticalization chains. To unravel this claim, recall that the agglutination of ʕlasa:s in Stage 2 results in the deletion of the final vowel of the preposition ʕla. Following this process, Stage 3 of the linear development in (16) shows a morphophonological shift (cliticization) by which the remaining part of the preposition ʕla is further reduced into the proclitic ʕ- that is cliticized into the noun asa:s and thus resulting in 20 Semantically, the NP lisa:ʔah is frequently used to cover a variety of time slots including ‘to this 10 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) moment’, ‘to this hour’, ‘to this time’ and ‘to that time’. Apart from the lexical use, (19) shows that the constituent lisa:ʔah is in the domain of the negative particle and this entails that it cannot occur in affirmative clauses (see example (21) below). Stage 2 of the linear development in (19) exhibits the morphosyntactic shift in which the negative particle laa is agglutinated to its complement and thus resulting in laalisa:ʔ. This stage also features the deletion of the intrinsic gender marker –ah. It is intrinsic because the gender inflection in al-sa:ʔah is grammatically inherent in the derivation of the noun form. This is not the case when the gender is contextualized as in adjectives such as ħilw- ah ‘beautiful.F’ and ħilu ‘beautiful.M’. Against this background, unlike adjectives, the deletion of the intrinsic feminine marker does not derive a counterpart masculine noun. In spite of this morphological analysis, the grammaticalization of the constituent laalisa:ʔ featuring the deletion of the intrinsic gender marker –ah has allowed for suffixing all subject agreement markers. Put it another way, Stage 3 of the grammaticalization development of laa lisa:ʔah introduces the constituent lisa:ʔ-(X) where (X) represents the subject accusative agreement suffixes shown in Table 4. have just’, taww-na ‘we have just’… etc. Besides, other functional elements can take these agreement markers and can stand alone without them such as the prepositions l-, fi, and ʔən, to mention a few. Against such a backdrop, the element lisa:ʔ- is considered a functional element at this stage and it can behave in such a way. It is worth mentioning that Stage 4 exhibits the grammaticalized form lissa(ʔ) ‘not yet’ which underlyingly features the agglutination of the negative particle laa. Hence, negation constitutively forms an underlying principle ingredient of the grammaticalized element lissa(ʔ), as illustrated by the examples in (21). Table 4 The subject agreement markers in lisa:ʔ- The paradigm suggests that the subject agreement markers can be suffixed to lisa:ʔ, and by deduction the form lisa:ʔ can stand alone without any agreement markers when it is nonreferential, which is the resulting constituent in Stage 4 of the grammaticalization development of laa lisa:ʔah = lissa(ʔ). Notice that Stage 4 also features the possibility of the deletion of the voiced pharyngeal fricative sound ʕ-; hence, it is represented in brackets in (19). The deletion of the agreement markers is not infrequent in HA; for example the adverbial time expression taww ‘just’ can take the agreement markers represented in Table 4: taww-i ‘I 20 In (21b) the grammaticalized negative-adverbial element lissa ‘not yet’ provides the negative response to the question in (21a). This negative- adverbial element features an intrinsic negative interpretation and this explains why it is compatible with negative clauses as in (21c), but not with affirmative clauses as in (21d). With reference to the cline of grammaticality, represented in (1) above, our discussion has clearly shown that lissa(ʔ) ‘not yet’ has developed from the negated content NP laa lisa:ʔah meaning ‘no, to this moment, to the moment of, to the hour of, to this time, to this hour’. However, the other stages of the cline of grammaticality are not witnessed here, i.e., lissa does not serve as a clitic or an inflectional affix. The following section concludes the paper. V. CONCLUSIONS AND RECOMMENDATIONS The present study shows that the forms xən, ʕsəs and lissa have evolved from the verb xalla, the preposition phrases ʕla asa:s and ila al-sa:ʕah respectively. Presumably, the gradual development of these elements has presented changes related to the different linguistic chains. That is to say, the forms have been grammaticalized through the cline of grammaticality exhibiting desemanticization (semantic shift), decategorialization 11 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) (morphosyntactic shift), cliticization (morphophonological shift) and erosion (phonetic shift). The paper has shown that the grammatical element xən functioning as an introductory particle has evolved from the content verb xall ‘notion of allowing’. While the content verb form and its perfective, imperfective and imperative derivatives inflect for the grammatical features of tense, aspect, voice, person, number and gender, the evolved element xən does not. The grammatically functioning cause and purpose conjunction particle ʕəsa:s has evolved from the PP ʕla asa:s. In like fashion, the negative adverbial element lissəʕ ‘not yet’ has evolved through a grammaticalization journey from the negated PP constituent [NEG laa [PP lisa:ʔah]. The paper is a contribution, on the one hand, to the growing literature of HA in that it helps gain a better understanding of the dialect. On the other hand, it contributes generally to the linguistic phenomenon of grammaticalization in that it enriches the literature exploring the elements changing through the cline of grammaticality and undergoing the four linguistic shifts. However, because the paper has only looked at three elements, future research on HA may explore more grammaticalized elements by which other studies may explore the features of all the elements that have gone through (or towards) the grammaticality pole. Further studies may also investigate HA and English causative expressions. They may compare the ways of causative expressions in English with the facts of HA, briefly shown in this paper, to make such HA expressions more familiar to the expert interested in grammaticalization but ignorant of HA. REFERENCES Al Zahrani, M. (2013). Morphosyntactic and Semantic Properties of Hijazi Arabic Modals. (PhD Theoretical Linguistics: Morpho-Syntax and Semantics Interface), University of Queensland, Brisbane, Queensland. Al Zahrani, M. (2016). Aktionsarten Projection and Subcategorization. The International Journal of Arabic Linguistics (IJAL), 2(1), 46 - 69. Al Zahrani, M. (2020). Hypothetical and counterfactual mood markers: hierarchical placement. The International Journal of Arabic Linguistics (IJAL), 6(1 & 2), 167-191. Bahloul, M. (2008). Structure and Function of the Arabic Verb. London: Routledge. Bisang, W. (1986). Verb-serialisierung in Jabêm. lingua, 70(2-3), 131-162. Bybee, Perkins, R. D., & Pagliuca, W. (1994). The evolution of grammar : tense, aspect, and modality in the languages of the world. Chicago: University of Chicago Press. Bybee, J., & Pagliuca, W. (1985). Cross-linguistic comparison and the development of grammatical meaning. In J. Fisiak (Ed.), Historical semantics - historical word formation (pp. 59-83): De Gruyter Mouton. Bybee, J. L. (2015). Language change. Cambridge: Cambridge University Press. Eifan, E. (2017). Grammaticalization in Urban Hijazi Arabic. (Master’s thesis), Manchester. Fleischman, S. (1982). The future in thought and language. Diachronic evidence from Romance. Cambridge Studies in Linguistics,(36). London. Haspelmath, M. (1999). Why is grammaticalization irreversible? Linguistics, 37(6), 1043-1068. Heine, B. (1993). Auxiliaries : cognitive forces and grammaticalization. New York ; Oxford: Oxford University Press. Heine, B. (1997). Possession: Cognitive sources, forces, and grammaticalization. Cambridge studies in linguistics(83). Heine, B. (2003). Grammaticalization. In B. D. Joseph & R. D. Janda (Eds.), The handbook of historical linguistics: Blackwell Publishing. Heine, B., & Kuteva, T. (2002). World lexicon of grammaticalization: Cambridge University Press.12 Mohammad Ali Al Zahrani/ Jurnal Arbitrer - Vol. 8 No. 1 (2021) Heine, B., & Reh, M. (1982). Patterns of grammaticalization in African languages. AKUP: Arbeiten des Kolner Universalien-Projekts Koln(47), 1-156. Heine, B., & Reh, M. (1984). Grammaticalization and reanalysis in African languages. Hamburg: Helmut Buske. Hopper. (1991). On some principles of grammaticization. In Traugott & Heine (Eds.), Approaches to Grammaticalization (pp. 17-35). Amsterdam/Philadelphia: John Benjamins. Hopper, & Thompson, S. A. (1984). The discourse basis for lexical categories in universal grammar. Language, 60(4), 703-752. Hopper, & Traugott. (1993). Grammaticalization. Cambridge: Cambridge University Press. Hopper, & Traugott. (2003). Grammaticalization (2nd ed.). Cambridge, UK ; New York, NY: Cambridge University Press. Jarad, N. I. (2013). The evolution of the b-future marker in Syrian Arabic. Lingua Posnaniensis, 55(1), 69-85. Jarad, N. I. (2014). The Grammaticalization of the Motion Verb “Raħ” as a Prospective Aspect Marker in Syrian Arabic. Al-ʿ Arabiyya, 47, 101–118. Jarad, N. I. (2017). Grammaticalization in Emirati Arabic. Arabica, 64, 742-760. Meillet, A. (1912). L’évolution des formes grammaticales: linguistique historique et linguistique générale. [reproduced in 1958] Paris: Champion, 230-280. Mitchell, T., & Al-Hassan, S. (1994). Modality, Mood and Aspect in Spoken Arabic: with special reference to Egypt and the Levant. New York: Columbia University Press. Norde, M. (2012). Lehmann’s parameters revisited. Grammaticalization and language change: New reflections, 73-109. Norde, M. (to appear in ORE ). Grammaticalization in morphology. Oxford Research Encyclopedia of Linguistics: Morphology,, to appear on: (https://oxfordre.com/linguistics). Norde, M., & Beijering, K. (2014). Facing interfaces A clustering approach to grammaticalization and related changes. Folia Linguistica, 48(2), 385-424. Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A Concise Grammar of Contemporary English. New York: Harcourt Brace Jovanovich. Ryding, K. (1994). Case/Mood Syncretism in Arabic Grammatical Theory”. In R. M. Rammuny & D. B. Parkinson (Eds.), Investigating Arabic (pp. 173-179). Ohio: Greyden Press. Ryding, K. (2014). Arabic: A Linguistic Introduction: Cambridge University Press. Ryding, K. C. (2005). A Reference Grammar of Modern Standard Arabic. Cambridge Cambridge University Press. Stewart, D. J. (1998). Clitic Reduction in the Formation of Modal Prefixes in the Post-Classical Arabic Dialects and Classical Arabic Sa-/Sawfa. Arabica, 45 104-128. Traugott, E. C., & Dasher, R. B. (2002). Regularity in semantic change. Cambridge: Cambridge University Press. Vangaever, J. (2019). Constructionalization at Work: The Emergence of the Latin Progressive Present Participle/Gerund Construction. Philologia Classica, 14(2). Versteegh, C. H. M. (2014). The Arabic language (Second edition. ed.). Edinburgh, : Edinburgh University Press. 13