Contact-Induced Change in an Endangered Language: The Case of Cypriot Arabic

Abstract

Cypriot Arabic (CyAr) is a severely endangered Semitic variety spoken by Cypriot Maronites. It belongs to the group of “peripheral varieties” of Arabic that were separated from the core Arabic-speaking area and came into contact with non-Semitic languages. Although there has been a renewed interest since the turn of the century for the study of CyAr, some aspects of its structure are still not well known. In this paper, we present and analyze a number of developments in CyAr induced by contact with Cypriot Greek. Our methodology for investigating such phenomena makes a novel contribution to the description of this underrepresented variety, as it was based not only on existing linguistic descriptions and text corpora in the literature, but mainly on a vast corpus of naturalistic oral speech data from the Archive of Oral Tradition of CyAr. Our analysis revealed the complexity of investigated contact phenomena and the differing degrees of integration of borrowings into the lexico-grammatical system of CyAr.

1. Introduction

Cypriot Arabic (CyAr), also known as Sanna or Cypriot Maronite Arabic (CMA), is a severely endangered Semitic variety spoken by Cypriot Maronites of the Kormakitis village. It belongs to the group of “peripheral varieties” of Arabic (Borg 1985, 2004; Roth 1973–1975; Grigore 2019), a group of Arabic varieties that separated from the core Arabic-speaking area and came into contact with non-Semitic languages (Borg 1994; Roth 2006–2007). The language was presented to the research community for the first time by Boustany (1957) and was briefly described more than a decade later by Newton (1964). However, the first extensive descriptions on the basis of fieldwork were the ones by Roth and Borg in 1973 and 1984, respectively. Since then, most studies of the language (e.g., Procházka 2006–2007; Kossmann 2008, 2010; Panayidou 2012, 2013; Grigore 2019; Mion 2017, 2018; Walter 2020) have been based on the data published by these scholars. New data were collected in the early 2010s by Gülle (2014, 2016), and more recently by the research team of the Cyprus Ministry of Education, Culture, Sports and Youth in an endeavor of creating an Archive of Oral Tradition of CyAr (see Section 2).

A question that was extensively discussed in the above-mentioned publications concerned the structural impact of Cypriot Greek (CyGr) on CyAr and whether the latter could still be considered a Semitic variety or, on the contrary, a mixed language. Even though answering this question is not within the goals of the present study, in this section we will present an overview of scholars’ position on the subject in order to highlight the importance of new descriptions of conduct-induced change in CyAr that can form the basis for further discussion. The position about the mixed character of CyAr was first formulated by Brian Newton, who, in his early account of CyAr in 1964, considered the Greek and Arabic elements present in the language of equal structural weight with no one prevailing over the other (1964, p. 51), and his position was also echoed in the title of his paper “An Arabic-Greek dialect”. Newton’s opinion was based mostly on the massive borrowing of loanwords from (Cypriot) Greek that entered CyAr with their morphological apparatus. However, we also need to remember that Newton’s paper appeared long before any systematic study of the phenomenon of code switching and/or spontaneous borrowing, both discursive phenomena related to the bilingual ability of speakers.

Thomason and Kaufman (1988, pp. 106–7) and Thomason (2001) endorsed the mixed language scenario on the basis of the evidence presented in Newton (1964) and concluded that CyAr is a mixed language that exhibits an extreme pattern of “borrowing of morphology (and phonology) along with lexicon”, whereby Greek items enter CyAr with their morphology, a phenomenon that affects, according to them, all grammatical subsystems. They even predict that the Greek will eventually supersede Arabic morphology and will end up being used with Arabic elements as well.

On the opposite side, Roth (1973–1975, 1981, 1986, 2000, 2004) and Borg (1985, 1994, 2004, 2011) argued that, despite extensive influence from Greek in all areas of grammar, CyAr is undeniably a Semitic variety that retains most of its structural characteristics. This viewpoint was also reflected in their choice to follow the descriptive framework traditionally used for the description of Arabic varieties to analyze CyAr. Notwithstanding this fact, Roth, in her 1973–1975 (pp. 22–23) description of the CyAr verbal system, admitted that she chose to do so “par parti pris pédagogique, et par commodité en vue de comparaisons ulterieures” and that this approach made her more aware of the difficulties of using well-established notions in Arabic dialectology to describe a rapidly changing language such as CyAr. Borg was also highly aware of the strong influence exerted by CyGr on CyAr and commented on this influence on several occasions. He did not, however, admit that any of the changes that can be ascribed to CyGr influence modify the essence of CyAr as an Arabic variety and refuted the idea that it could be treated as a mixed, pidginized or creolized variety of Arabic (Borg 1994).

Gülle (2014) argued that the massive borrowing of words from CyGr and the high degree of code switching are due to the high competence of CyAr speakers in CyGr. He claimed that, contrary to other cases of language shift whereby decay precedes death, CyAr dies “with its boots on”, to paraphrase Hamp (1989), because of the lack of “proper semi-speakers”—in other words, because of the lack of a generation of speakers who would have acquired and used the language imperfectly. In his comprehensive study of CyGr, Cypriot Turkish (CyTr) and CyAr, Gülle underlined the parallel development of similar structures in the three varieties, ascribing them to intense contact between their speakers in a predivided Cyprus (Gülle 2014). According to Gülle (2016), CyAr bares none of the expected signs of a moribund variety, such as breakdown of grammatical categories, agrammatism, loss of subordinated mechanisms or extreme phonological variation, distortion and hypercorrection, but does exhibit massive lexical borrowing—including adverbials in adverbial clauses—with the loss of systematic integration (Gülle 2016, p. 45). Nevertheless, he considered the last two trends as possible instances of code switching due to the bilingual competence of CyAr speakers. He also ascribed the lack of any sustained signs of structural decay in the CyAr variety he observed to a rapid loss of speakers caused by the interruption in intergenerational transmission, which was the indirect effect of the massive displacement of the Kormakiote population away from their village after 1974. The same position regarding the lexical impact of CyGr on CyAr was taken up by Vassili in his MA dissertation (2010) on CyAr. Vassili considered most CyGr segments used by CyAr speakers as instances of code switching and single words as cases of spontaneous borrowing. Thomas (2000), in his turn, also treated CyGr segments as cases of code switching and commented that the massive resort to code switching makes it “difficult to determine if a word is a genuine loan” (and thus, integral part of the lexicon of Sanna), “or simply a response to a speaker’s environment” (that is spontaneous borrowing or an instance of code switching) and therefore “… inhibits the study of the limits of CMA’s lexicon”. Massive borrowing of single words or expressions from CyGr could also be seen as part of a process of relexification (see below) whereby CyAr loses specific areas of its native lexicon and non-native words from CyGr flow into supplement (Karyolemou 2021).

Lüdtke (1976–1977) considered Maltese and CyAr as similar cases of varieties with mixed morphology. He also underlined (1976–1977, p. 213) that one of the main factors (although not the only one) that condition the emergence of such mixed systems is of sociolinguistic nature, namely the existence of ‘communal bilingualism’, that is, the situation whereby a majority of speakers are in a position to master both the native and the dominant system. Comparing a number of peripheral Arabic varieties according to a set of features taken from all linguistic levels and the lexicon, Procházka (2006–2007) argued that CyAr occurs today as the most estranged of all. However, he also warned researchers against the risk of hastily pronouncing the divergent nature of a variety on the basis only of lexical borrowings: “the high number of non-Semitic loanwords in these dialects often leads one to jump to the conclusion that they have developed significantly away from their Arabic sisters. But such an assertion should be handled with caution” (2006–2007, p. 117). He remains, however, silent on the issue of the mixed character of CyAr.

Kossmann (2008) supported the view that, in order to decide about the mixed character of CyAr, we also need to consider the nature of the Greek component used by CyAr speakers, something that had rarely been done. Using data from Borg (1984, 1985, 2004), Newton (1964) and Roth (1973–1975), Kossmann analyzed the CyGr elements that occur in CyAr more thoroughly, concluding that they are quite different from what one would expect had they been used by a monolingual CyGr speaker. In light of these observations, Kossmann concluded that we cannot consider all CyGr elements as cases of code switching, whereby, by definition, the segments keep their source identity. However, he also rejected the mixed language hypothesis and claimed that, despite the strong influence of CyGr on CyAr, it would be exaggerated to consider the Greek and Arabic component as of equal par: the Arabic component is clearly preponderant. In light of his observation on CyAr and other cases of centuries-long contact, Kossmann (2010) formulated the parallel morphology hypothesis, whereby “two morphological systems, differentiated according to etymological origin, function side by side in the same language” (Kossmann 2008, p. 20, see also Kossmann 2010) without mixing; the CyGr elements keep their morphological identity, whereas the CyAr elements keep theirs. Kossmann (2010, p. 483) argued that the existence of parallel morphological systems is a well-attested phenomenon in many languages worldwide and does not necessarily qualify only mixed languages.

The idea of the existence of two different paradigms that are valid for CyAr and CyGr elements is also advocated in Panayidou (2012, 2013). Panayidou’s work on the adjective–noun order led her to hypothesize the existence of two functional sequences, a Greek and an Arabic one, that differ both morphologically and syntactically and are valid for different semantic adjectival classes: the Greek structure accommodates the adjective classes of shape, nationality and borrowed Greek color, whereas the Arabic structure accommodates the adjective classes of quality, size and native Arabic color, the former being embedded into the latter. However, Panayidou did not draw any conclusions regarding the mixed character of CyAr.

It is obvious from the overview presented above that scholars disagree on the mixed or predominantly Semitic character of CyAr. Note however that no study argues that Greek has taken over structurally. Even the idea that the Greek element is at least equally important as the Semitic one in CyAr is clearly refuted by Kossmann (2008, p. 19): “The main language is clearly a form of Arabic, while the Greek elements are statistically relatively marginal. One can write a grammar of the Kormakiti language using solely the Arabic part of the lexicon, as Borg did, but it would be preposterous to do the same using only the Greek part”.

2. The Scope of the Paper

Contact-induced change in CyAr is a dynamic and still on-going process. The changes that resulted after the forced displacement of the Maronite population out of Kormakitis village in 1974 (Gülle 2014, 2016; Karyolemou 2019) and their relocation in the south of the island intensified contacts with the Greek-speaking population, increased the number of mixed marriages and confirmed the trend towards language shift: today, no member of the Kormakitis Maronite community is to be considered monolingual in CyAr. According to estimations based on field work with and within the community (Karyolemou 2010), a language shift in favor of GyGr has been completed for ages 40 and below. The process of language shift was already noted by Newton (1964) and confirmed by Roth (1973) prior to the massive displacement out of Kormakitis, but it certainly intensified after the events of 1974 with new areas of contact-induced change being added to the old ones.

However, as up until very recently there had been no written CyAr literature, or any kind of written texts in CyAr for that matter, we are in the dark regarding CyAr in its diachrony. Our earliest source of information about the structure of the language, which can be used for comparison with contemporary data from the Archive of Oral Tradition, is Boustany (1957), who does not offer a systematic description; Newton (1964), by contrast, overviews several phenomena, but with a small number of examples elicited from only one speaker, whom he met during his short visits to Kormakitis in May and June 1963. More extensive data in the form of a corpus of 14 small stories were recorded and transcribed by Borg (1985), and a certain number of isolated examples are also found in Roth (1973–1975, 1986, 2000, 2004) from her own fieldwork in 1972, 1973 and 1974; as mentioned above, several examples are also included in Gülle (2014, pp. 225–30), whose fieldwork was conducted in December 2010 and August 2011 and included four speakers, and finally, the most extensive corpus of oral data is the aforementioned Archive of Oral Tradition. The Archive of Oral Tradition of Cypriot Arabic is a corpus of natural speech data collected since 2013 within the framework of the Research Programme for Documenting and Revitalizing Cypriot Arabic funded by the Cyprus Ministry of Education, Culture, Sports and Youth. It comprises 81 interviews with a total duration of 235h37m with 40 female and 41 male CyAr speakers, of 61 to 100 years of age. All speakers originated from Kormakitis village and are considered native or good speakers of CyAr. The interviews were conducted by members of the community who had been trained in the principles of ethnographic research. The collection of data is still in progress and aims to include all active speakers of the language.

The lack of extensive historical data for the language invites us to be extremely vigilant with hypotheses about the influences exerted on CyAr by other varieties spoken in the Cypriot context or by varieties spoken by communities with which the Maronite community of Cyprus has diachronically had relations, such as Lebanese or Standard Arabic.

Using data from the Archive of Oral Tradition for CyAr, in this paper, we present and analyze several under-described phenomena in CyAr that could be attributed to the influence of CyGr. More specifically, in the area of phonetics and phonology, we discuss the phonetic inventory of CyAr vowels and consonants, the phonetic realization of geminate occlusives, yod occlusivization and manner dissimilation in clusters of plosive consonants; in the area of the lexicon and morphosyntax, we discuss extensive borrowing, partial relexification and its consequences for the syntactic properties of NPs; finally, we discuss the case of the fi structure to express the directional and locative functions (Karyolemou 2019). Our goal in this paper is to offer a wide range of naturalistic data from the largest corpus of recorded oral speech of CyAr and their analysis in order to enrich our knowledge of the structures of the language and of possible contact-induced change. Such a descriptive work based on the analysis of a large corpus of naturalistic oral data is important, as it can form the basis for future studies to engage in a more informed discussion regarding whether CyAr is a mixed language or not.

3. Contact-Induced Change in Phonetics and Phonology

Contrary to the case of Old Arabic, the vowel system of CyAr does not distinguish between short and long vowels (Borg 1985, p. 12). Apart from vowel quantity, CyAr differs from Old Arabic in exhibiting a third vowel height beyond the high vs low contrast, namely the mid height; specifically, CyAr features two mid vowel phonemes, namely /e/ and /o/, and thus, regarding its monophthongal vowels, it has a five-vowel system: /a e i o u/. Even though this vowel system is identical to the one of CyGr, it is not the case that the mid vowels /e/ and /o/ are due to influence from CyGr, for they have developed internally following predictable patterns of diachronic change from Old Arabic (Borg 1985, pp. 69–71) and are also found in various other spoken Arabic varieties. Regarding the phonetic realization of CyAr vowels, the low vowel is impressionistically realized as a central [ɐ], while the /e/ and /o/ vowels are realized as mid vowels [e̞] and [o̞], respectively, and these realizations are perceptually of the same quality as those of the CyGr respective vowels, something though that awaits empirical confirmation. Furthermore, the allophonic realizations observed in other Arabic varieties which are foreign to the CyGr vowel system, such as [ɑ] and [æ] for /a/ (e.g., Thelwall and Sa’Adeddin 1990), are not found in CyAr. Thus, the vowel system of CyAr appears to phonetically coincide with that of CyGr.

Regarding diphthongal vowels, even though /aj/ and /aw/ (and marginally /ej/ and /uj/) are reported to exist in the vowel inventory of CyAr (Borg 1985, pp. 11, 44–45), native speakers appear to have difficulty recognizing them as single vowels: specifically, through the process of the documentation of CyAr, the creation of teaching materials and especially its teaching, the authors of the current study have observed that CyAr speakers intuitively syllabify words such as /ʕajn/ “eye” and /nawm/ “sleep (noun)” as [ˈʕa.in] and [ˈna.um], respectively. We propose that this de-diphthongization into a biphonemic vowel sequence is due to contact with CyGr, which does not tolerate diphthongs; e.g., in CyGr [xa.i.ˈðe.fko], “I caress” has four syllables, while Standard Modern Greek (SMG) [xai̯.ˈðe.vo] has three. This is also evident in the form of Standard Greek as used in Cyprus, namely Cypriot Standard Greek (Arvaniti 2010), in which SMG diphthongs are broken up into two different syllables; e.g., the SMG word [ˈɣai̯.ða.ɾos] “donkey” is rendered as [ɣa.ˈi.ða.ɾos] in Cypriot Standard Greek. In CyAr, a counterargument would be that there are a couple of CyAr words stressed on the antepenultimate syllable which contain a diphthong in the penultimate or ultimate syllable; e.g., the words /ˈku.lːaj.le/ “every night” and /ˈnu.sːu.lːajl/ “midnight” should contain the diphthong /aj/ or else they would be stressed on the fourth syllable from the end, which is generally not allowed in Arabic varieties. However, we propose that CyAr phonology has been influenced by CyGr in allowing stress to fall beyond the penultimate, as this is observed to a certain degree in CyGr: e.g., [ˈɣɾa.fu.me.nde] “we write”, [po.ˈt͡ʃi.tʰːe.me.ɾu] “over there”, [ˈpɾo.te.le.fte.os] “penultimate”, [ˈpo.ksi.lːi.cʰːin] “scolding”, etc. Thus, if we assume that CyAr phonology has been influenced by CyGr phonology in this respect as well, then we can accept the breaking up of diphthongs even in the words [ˈku.lːa.i.le] and [ˈnu.sːu.lːa.il].

The prohibition against diphthongs in CyGr is part of a wider constraint of CyGr phonology regarding semivowels in general: no semivowels emerge at the surface level, be they in the nucleus (i.e., as part of falling diphthongs) or in onset or coda position (i.e., as approximant consonants). We argue that this general prohibition against surface semivowels in CyGr also holds in CyAr, in which the semivowels /j/ and /w/ of Old Arabic have evolved into the fricatives [ʝ] and [v] when in onset position and to full syllabic vowels [i] and [u] when in the nucleus—either on their own (e.g., /ʕjt/ → [ˈʕit] ‘feast’; cf. plural /ʕajat/ → [ʕa.ˈʝat]) or even after another vowel (with which we claim do not form diphthongs). It can be argued that the change from semivowels to fricatives or vowels is not only a diachronic change, but also constitutes a synchronic alternation. At a deeper morphophonemic level, the existence of /j/ and /w/ may be postulated in the root of words, and depending on their position in the syllable, they may surface as fricatives or vowels. For instance, the root for “faucet” is |ʕ-j-n|, which becomes /ʕajn/ → [ˈʕa.in] in the singular and /ʕajun/ → [ʕa.ˈʝun] in the plural; similarly, the root for “blind” is |ʕ-w-ɾ|, which becomes /aʕ(a)waɾ/ → [ˈa.ʕa.vaɾ] in the singular masculine, /ʕawɾa/ → [ˈʕa.u.ɾa] in the singular feminine and /ʕ(u)wɾ/ → [ˈʕuɾ] in the plural.1

CyAr differs from most Arabic varieties in lacking the uvular consonant [q] and almost any consonants of the guttural place of articulation, namely the glottal consonants [h] and [ʔ], the voiceless pharyngeal fricative [ħ] and the pharyngealized (‘emphatic’) consonants such as [tˤ], [dˤ], [ðˤ] and [sˤ]. The only pharyngeal consonant in the CyAr phonetic inventory is the voiced fricative [ʕ] with no voiceless counterpart. Other than this consonant, CyAr coincides with CyGr in not utilizing places of articulation behind the velum. From a historical perspective, this reduction in the consonant inventory from Old Arabic to CyAr may be due to influence from CyGr (Borg 1997, p. 539).

One other aspect of contact-induced change is the phonetic realization of geminate voiceless occlusives in CyAr, which, contrary to other Arabic varieties, differ from their singleton counterparts not only in terms of the increased duration of the occlusion, but also in being aspirated. Aspiration as a phonetic correlate of geminate voiceless occlusives is a characteristic of CyGr (Armosti 2011a, 2011b), which is carried over to CyAr, resulting in an identical set between the two varieties of aspirated geminates contrasting with unaspirated singletons: [p] vs. [pʰː], [t] vs. [tʰː], [k] vs. [kʰː], [c] vs. [cʰː], [t͡ʃ] vs. [t͡ʃʰː]. Moreover, similarly to CyGr (Armosti 2011a), in CyAr there is no singleton–geminate contrast in the case of the alveolar affricate, for it is always realized as a geminate [t͡sʰː], as revealed by our phonetic analysis of instances of this sound in the Archive of Oral Tradition. Regardless of the CyGr influence exerted on the CyAr system of geminate consonants, CyAr still differs from CyGr in exhibiting the geminate voiced plosives /bb/ and /dd/, which do not exist in CyGr: e.g., [tuˈbːane] ‘fly (the insect)’, [miˈdːate] ‘rolling pin’. In the analysis of the oral corpus, the phonetic realization of these geminates ranged from fully voiced [bː] and [dː] to fully devoiced [pː] and [tː], with the latter exhibiting no aspiration. Thus, CyAr has a tripartite contrast among [t], [dː] and [tʰː] in the case of the coronal plosives and [p], [bː] and [pʰː] in the case of labial plosives, which is foreign to CyGr: e.g., [ˈʃata] ‘hers (plural possession)’ vs. [ˈʃadːa] ‘(he) tightened it (female object)’ vs [ˈʃatʰːa] ‘it rained’. Interestingly, the result of the process of post-lexical full assimilation of the definite article /l/ to a word initial /p/ is [bː] and not [pʰː] (see Christodoulou et al. 2021); e.g., /l#ˈpape/ → [ˈbːape] ‘the door’ (while ‘door’ is [ˈpape]). This is a reflex of the Old Arabic /b/, which, in CyAr (arguably under the influence of CyGr), has changed into a voiceless [p] when singleton, but retains its voicing when geminated. The same phenomenon is observed in a limited set of lexical items beginning with a /t/ that historically originated from a /d/: e.g., /tist/ → [tist] ‘dish’ vs /l#tist/ → [dːist] ‘the dish’. However, in most cases, the aspirated version of this geminate results: e.g., /tik/ → [tik] ‘rooster’ vs /l#tik/ → [tʰːik] ‘the rooster’ (Borg 1997, p. 234).

The phonology of CyAr has also been influenced by having borrowed certain phonological processes from CyGr phonology. One of them is “yod occlusivization”, i.e., the fortition of an underlying glide /j/ into a dorsal plosive after obstruent consonants and the rhotic /ɾ/: e.g., /pjut/ → [pcut] ‘houses’ < */bjuːt/; /θjep/ → [θcep] ‘clothes’ < */θjeːp/ (Borg 1997, p. 224). This phenomenon is not encountered in other Arabic varieties, but is characteristic of CyGr: e.g., /antɾoˈpi̯azzo/ → [andɾoˈpcazːo] ‘I shame’, /paɾaˈmiθi̯a/ → [paɾaˈmiθca] ‘fairytales’. Interestingly, this process in CyAr extended to environments in which it does not occur in CyGr, such as after sibilants: e.g., /ˈmeʃje/ → [ˈmeʃce] ‘walker.fm’. One other process borrowed into CyAr from CyGr is manner dissimilation of a plosive into a fricative when followed by another plosive: e.g., similarly to the CyGr realization of /kt/ as [xt] in /pleˈktos/ → [pleˈxtos] ‘knitted.m’, in CyAr, /ktilt/ is realized as [xtilt] ‘you.s.m killed’ (Borg 1997, p. 224). Compared to CyGr, CyAr phonotactics allow more consonantal clusters in all word positions, something that facilitated the spreading of the dissimilation rule to clusters not found in CyGr, such as in /tk/, /pk/, etc.: e.g., /ˈtkuɾa/ → [ˈθkuɾa] ‘males’, /pkum/ → [fkum] ‘I get up’.

4. Contact-Induced Change in the Lexicon, Morphology and Syntax

In this part of the paper, we will discuss lexical borrowings and the concept of partial relexification, as well as the possible effect of massive borrowing on syntactic changes in NPs. We will then proceed with a more detailed discussion of the syntactic changes in the directive and locative functions by considering the multifunctional CyAr preposition fi ‘to’ and its current usage. The fi structure was specifically chosen because it has been signaled out as a probable influence of CyGr, at least in some of its uses, in previous works, e.g., Borg (1985, 2004) and Gülle (2014, 2016).

4.1. Lexical Borrowings and Partial Relexification

While documenting the language, we observed variable performances with utterances ranging from almost everything in Greek (ex. 1 and 2 below) to almost nothing in Greek (ex. 3 below), which attests to the varying degrees of influence from CyGr, even by speakers within the same age group.2

(1)	Yatí panta kanlíχon …, l-δeca kant ipoχreomeni na sintirí l-kalurkés.
	‘Because they had always … the village had the obligation to provide for the nuns’
	HΖ3_ΠΚf_19-09-2015_A_1_ΚΤ (87y)3

(2)	U ost alok, akke ilikia ta naχni, ana evδomintatessaron χronón, kápote áθela mas l-δi pkyaχutna aχχók allá entje- entje kamarónnumen l-korifí tel katástasi
	And until now, to the age we are, I am seventy-four years old, sometimes unwillingly, the eye turns us up there but it’s not- it’s not that we are proud about the top of the situation
	HΖ1_AΧm_02-08-2015_A_2_KT (74y)

(3)	… vurοs χarap, ute rijce tekuyllu, χarap, χarap, χarap. Raχ il-Lárnaka tel Lapeθ, mnawna l-Lárnaka tel Lapeθ. L-Lárnaka tel Lapeθ raχ kal, ekef kal, kanikapel eχte mara cijnet kal, kant aχχullát il-forn kanistanter attaχrop teimpsek şi caríf teakol. Má kantaχrop kal. Kantintor il-χops tela. Alok aş tasáy pikúl yapati.
	… he left running, he didn’t even turn back to talk to him, he left, left, left. He went to Larnaca of Lapithou, on this side to Larnaca of Lapithou. He went to Larnaca of Lapithou, he said, he stayed, he said, he saw a lady baking, he said, she was next to the oven, so he was waiting for her to leave in order to take a bread to eat. She was not leaving, he said. She was watching over her bread. Now, what to do, my father says.
	HZ6_ΦΤf_21.11.2015_A (71y)

Several researchers (Roth 1973–1975, 2000; Borg 1985, 2004) have already noted the loss of native lexicon. Lexical reduction is not surprising, if we consider the social and interactional limitations set on the language. Used exclusively for intracommunal and/or intrafamily communication, CyAr remained restricted to familiar and ordinary areas of interaction and never developed a vocabulary to refer to new concepts. Thus, for instance, with the exception of occupations relating to traditional activities—farmer, shepherd, priest, teacher—all other professions are not represented in the language and are borrowed from CyGr. It would not even be accurate to talk about lexical loss in this case as it is most likely that the language never developed the vocabulary corresponding to modern professions (Gülle 2014). The same applies in the case of ethnic names: with the exception of a generic term for non-Maronites, χawtík (litt. ‘ceux-là’ (Roth 2004), pl. χfetík), almost exclusively used to designate Greek Cypriot people, the remaining ethnic names are all borrowed from CyGr. More puzzling is the replacement of all numerals after ten by CyGr numbers, with the exception of tens, hundreds and thousands.

From the speakers’ point of view, some cases of borrowing seem to be well received and generate no negative reactions; speakers tend to consider them as part of CyAr vocabulary. Such is the case for ethnic names and, to a lesser degree, for professions. For numerals, on the other hand, the situation differs: speakers tend to consider the lack of numerals after 10 and their replacement by CyGr numbers as problematic. This became obvious to us from the negative reactions of CyAr speakers who were trained at the University of Cyprus as language educators for the Summer Linguistic Camp held in Kormakitis in 2014. Most of them were in their late 40s or 50s and were considered native or very good speakers of the language, which they continued to use among themselves even if not on a daily basis. They had no professional qualifications in language teaching but shared a strong commitment towards maintaining and reviving CyAr (Karyolemou 2021). While they generally acquiesce to the use of CyGr words for professional and ethnic names, they vividly reacted to using CyGr numerals to remedy the missing CyAr numbers, even though they were themselves already regularly using them orally. On the contrary, they suggested that new numbers be devised based on the counting system currently in use in other Arabic varieties. This solution was finally adopted, and new numbers were devised according to Standard Arabic (Karyolemou 2021).

Another area of systematic and extensive borrowing from CyGr concerns complementizers and conjunctions, since CyAr has not developed Semitic terms for these functions—probably because of its oral nature, which often dispenses with systematically marking syntactic relations (Borg 2004)—and, therefore, has taken over almost the entire system of complementizers and conjunctions from CyGr, as shown in examples 4–6 below.4

(4)	… entje kantácaref oti kantatrúχ arka tinye
	She didn’t even know that there would be so many people
	HZ6_ΦΤf_21.11.2015_A12 (71y)

(5)	… afú má pitrúχ il-Kormatjiti, ttarúχ oχre δeca
	… since you are not going to Kormakitis, I will go to another village
	HZ6_ΦΤf_21.11.2015_A13 (71y)

(6)	Allá raχu antáχt u tcalmu ll-arápika
	But they went over there and they learned Arabic
	HΖ1_AΧm_02-08-2015_A_4_KT (75y)

The extent of borrowing becomes more obvious in narrative passages, wherein the succession of events needs to be marked by means of functional words. Observe extract (7) below from a 99-year-old speaker of CyAr,5 who narrates a frightening experience from WWII,6 in which CyGr ammá ‘but’, afú ‘since, when|well’, amma ‘when’ and δjoti ‘because’ serve to link together the various events and obtain narrative flow.

(7)	ammá u kanyukcáw pompes calena, afú lli-proto l-layle ta ruχna, kanisa[v]unna mnaδa … amma kuanrúχ tannaχtca kuanχótt mnaδa fi l-maχurna δjoti milli ptómata má mannaχter tannaχtca ...
	But also, the bombs would fall on us, since the first night that we went, they did that thing for us [sign in front of the nose to indicate a mask] when we were trying to pass by we had to put that thing on our nose because we could not pass through all those dead bodies ...
	HZ4_AΧKm_19.11.2015_A (98y)

We tend to believe that the systematic and massive borrowing of CyGr terms for specific semantic fields or entire grammatical classes and the replacement of native lexicon, where it existed, could be seen as a case of partial or selective relexification (Muysken 1981; Lefebvre 2005) or relabeling (Lefebvre 2008) from the dominant language (CyGr). Because the terms relexification and relabeling have been almost exclusively used to explain how creole or mixed languages emerged out of substrate African languages with a lexicon borrowed from dominant European languages under specific social conditions, we need to clarify that we do not make any such claim with regard to CyAr, since an important part of its lexicon still remains of Semitic origin. If we follow Muysken (1981), however, according to whom relexification should not be treated as a unitary phenomenon but as a continuum of lexical transfers from a providing to a recipient language,7 then there is no reason not to apply the concept of partial or selective relexification/relabeling to the process of extensively borrowing in any situation of intense and prolonged contact that affects some parts of the lexicon and not others, as long as it concerns entire grammatical classes or lexical fields.

Some borrowings from CyGr are fully integrated in CyAr according to the morphological rules of Arabic despite differences in the morphological processes of derivation between Greek and Arabic. For instance, CyAr paδpún is formed on the CyGr word ποταμόν ‘river (acc.)’ by extracting the four consonants ptmn, then applying an unharmonized scheme with a low vowel in the initial syllable CaCCuC > patmún > paδpún (pl. ftamín) (Borg 1985, p. 118). The same process applies to the word χmin ‘short’ < CyGr χαμηλόν for which consonant extraction has resulted in the formation of a three-consonant root χm(l)n and, by application of the scheme CCiC, the word χmin (pl. χman).

Given the amount of borrowings from CyGr, it is quite remarkable that such examples of integrated borrowings are extremely rare and practically limited to a handful of cases, as already underlined in Roth (2004) and Borg (1985, pp. 108, 153): e.g., CyAr kammín (pl. kmenín) < CyGr καμίνιν ‘furnace’, kişnár (pl. kşenír) < CyGr ξινάριν ‘pickaxe’, and a few others. An explanation might be that they represent older borrowings from a time where CyAr was L1 for all/most speakers in the community, with CyGr being familiar to only a few, and the language was still dynamic enough to integrate borrowings. Non-adapted loanwords, on the other hand, were arguably the result of subsequent extended bilingualism; as certain items of the CyAr lexicon gradually receded, non-adapted loanwords were considered not to be a problem anymore; on the contrary, they proved useful, as they facilitated communication. According to Roth (2004), among the receding lexicon, there might have also been older integrated borrowings.

More often, CyGr borrowings are minimally integrated in CyAr, retaining most of the morphological, phonological, semantic and syntactic properties of the source language, thus resembling spontaneous borrowing or code switching (Gülle 2014, p. 102) more. In the following examples, the CyGr feminine noun istoria ‘story’ retains its semantic and morphological characteristics, as can be seen from the use of the feminine form of the third person singular of the CyAr copula e and the feminine demonstrative of proximity aδi (ex. (8)), whereas the CyGr words poδílato ‘bicycle’ and yofiri ‘bridge’ retain all their characteristics and are minimally integrated by taking the definiteness marker l- that assimilates to the initial consonant of the accompanying noun (ex. (9)). Finally, in example (10), the CyGr verb stínnun ‘set up’ occurs in its regular form, whereas the noun stíllo ‘pile’ retains its gender (mas.) and number (sing.), as shown by the use of the masculine singular indefinite article eχen ‘a’, case mark (acc.) and consequently its syntactic function (object in acc.) in relation to the verb ‘stínnun’.

(8)	Aşşik	kitlúχ	aδak	aş	istoria	e	aδi
	wh	kill.past1PL	dem.dist.MSG	wh	storyFSG	cop.3SG.F	dem.prox.FSG
	How did they kill that one, what (a) story is this?
	HZ4_AΧKm_20.11.2015_B3 (98y)

(9)	Misek	il-poδílato	raχ	aχχullá	l-yofiri …
	take.past.3SGM	art-bicycle.NSG	go.past.3SGM	dem.adv.	art.-bridge.NSG
	He took the bicycle, he went over there to the bridge…
	HΖ3_ΠΚf_24-10-2015_Β_1_ΚΤ (87y)

(10)	Pikacituon	anaχχullá	u	stínnun	eχen	stillo	χok
	prs.put. 3PL-obj.pro	adv.prox.	conj.	raise.prs3PL	num.	pile.MSG	adv.
	They put them over here and raise a pile on top
	HΖ2_ΚΠm_05-08-2015_A_1_KT (81y)

4.2. Massive Borrowing and NP Structure

The massive introduction of borrowings, here nouns, from CyGr has also other consequences with regard to the structure of the noun phrase (NP) in CyAr, in particular when the head noun (HN) of the NP is specified by an adjectival phrase (AP). In what follows we will focus on the syntactic properties of the NP, i.e., the order of appearance of the HN and AP, given that it differs in the two languages: in CyGr, unless there are stylistic reasons, the order of appearance is “modifying adjective + modified noun”, whereas in Arabic the order is, generally speaking, “modified noun + modifying adjective” (see, however, Fehri 1999; Al-Shurafa 2006 for prenominal adjectives). In her extensive discussion of adjective–noun order in CyAr, Panayidou (2012, 2013) argues that adjectives follow different morphological and syntactic rules depending on whether they are native or borrowed from CyGr. Arabic adjectives (mostly quality, size and native color adjectives) follow rules of non-concatenative morphology, and thus occur post-nominally even when the modified item is a borrowed item. This is illustrated in example (11) below where the native adjective of size γpar (pl.)8 ‘big’ appears after the borrowed HN kamminya ‘furnaces’ and in example (12), where the CyGr-borrowed term kámara is followed by the adjective γpire.

(11)	Ana	kuntsáy	kamminya9	γpar	χtir
	per.pron.1SG	imperf.make.1SG	furnace.NPL	big.PL	quant.
	As for me, I used to make very big furnaces
	HΖ2_ΚΠm_05-08-2015_A_1_KT (81y)

Note that both the indefinite article eχte and the adjective γpire agree in gender with the borrowed item, which retains its morphological and syntactic characteristics.

(12)	Kant	eχte	kámari	γpire
	be.imperf.3SG	art.-one.FSG	room.FSG	big.F
	It was a big room
	HZ1_AΧm_02-08-2015_A (75y)

In example (13) below, however, contrary to predictions, the quality adjective CyAr kayse ‘good’ (f.) appears in front of the CyGr HN it modifies, thus violating the rule observed in Arabic:

(13)	ómorfi	kaθaroprósopi	kayse	ráftena	kayse	traγuδistina	kayse	χoreftina
	beautiful.FSG	nice-faced.FSG	adj.FSG	tailor.FSG	good.FSG	singer.FSG	good.FSG	dancer.FSG
	Beautiful, nice-faced, (a) good tailor, (a) good singer, (a) good dancer
	IA3_ΓΣf_01.10.2015_A (~75y)

The position of the adjective in front of the noun does not seem to be random, given that it occurs in the same position three times in a row. The change in the place of the adjective might be triggered by the presence of the CyGr-borrowed nouns ráftena, traγuδistina and χoreftina: the adjective complies to the order observed in the language to which the HN belongs, as the HN has a preeminent position in the NP.

In the following examples ((14) and (15)), however, the native adjectives are preposed to native NHs (examples from Karyolemou and Kanikli Forthcoming):

(14)	γpire	cayara,	cayarua
	big.FSG	cloud.FSG	cloud.dimin.FSG
	Big cloud, small cloud
	IA2_ΣΚf_01.10.2015_B (~75)

(15)	kullon	ntviltu	mnawnke	allik	il-caták	illi-pkyut
	indef.prn.	pass.born.past.3PL	adv.dist.	dem.dist.PL	art.-old.MSG	art.-house.MPL
	Everybody was born over there in those old houses
	IA4_ΚΦf_07.11.2015_A (75)

In examples (14) and (15), we could consider that the transfer of the CyGr syntactic rule in NP+AP is completed since it applies now to entirely native NPs without being triggered by the presence of a borrowed CyGr item. The use of mixed NPs where a CyGr noun is combined with a CyAr adjective (as in (13)) might be the path that facilitates structural transfer because of the syntactic preeminence of the HN, which impacts the use of the CyGr order in NP. This is of course only a hypothesis and needs to be further examined.

The high number of CyGr occurrences in CyAr—that still need to be extensively studied, because they do not all resort to the same contact phenomenon—is attributed by many researchers to the high degree of bilingualism of Kormakitis’ speakers. From a minority language maintenance point of view, it might be that CyAr has been maintained for so long despite strong pressure from the dominant CyGr community towards acculturation and linguistic assimilation partly because of its ability to extensively borrow from and converge to the dominant language (see for instance Borg 1985 on the phonological convergence of CyAr to CyGr). As Hill and Hill (1977) already noted a long time ago, sometimes minority languages that have been in contact with a dominant language over a long period of time might open the door of their lexicon to dominant influences to preserve other, more important structural levels, and to be able to survive. Although this assertion gives language an active role in protecting itself from attrition and loss, the idea that structural flexibility or convergence in a contact situation might allow a small language or certain aspects of it to survive over a longer period of time cannot be dismissed as much as it cannot be proven. Roth (2000, p. 132), for instance, commenting on the maintenance and relatively high occurrence of the relative clauses in the speech even of less competent speakers, which she finds to be at odds with the attrition and loss suffered by other types of subordinate clauses, states that a possible explanation could be that “la structure de la relative dans les énoncés produits par des locuteurs de bonne et de mediocre compétence et la structure de la construction relative en grec chypriote présentent de grandes analogies” and offers several examples where the structure of relative clauses in the two languages is exactly the same.

4.3. The Directive and Locative Function

As in many other Arabic varieties, fi is a multifunctional connector in CyAr and appears in a variety of syntactic environments. In particular: (a) it introduces a nominal with a directive or locative function often (but not exclusively) after a transitive verb; (b) it heads prepositional phrases that bear the beneficiary thematic role of ditransitive verbs; (c) fi-phrases function as prepositional complements. Although there are also several other functions (Karyolemou and Kanikli Forthcoming), here we will concentrate on (a), whereby fi introduces a nominal to mark location and direction with or without a transitive verb. These uses of fi can be illustrated in the following examples (from Karyolemou and Kanikli, forthcoming):

(16)	Ju	fi	l-moδca	telna
	come.past-3PL	prep.	art.-place.MSG	poss.
	They came to our place

(17)	Kuntrúχ	as pumen	fi	ll-oχar	illi-θkyaca	l-maronítika?
	go.past.1SG	part.say.pr.1PL	prep.	art-otherPL	art-village.FPL	art.MaroniteNPL
	Did you go, let’s say, to the other Maronite villages?

(18)	Eh	l-δeca	şaitna	pittukaca	fi	l-δisi
	interj.	art.-village.FSG	poss.	prs.fall.3SG.F	prep.	art.west.FSG
	Our village falls to the West

(19)	Kecetín	fi	lli-snie	tannakol	u	takket	il-pape
	seat.ptcp.PL	prep.	art.-table.FSG	subj. eat. 1PL	conj.	knock.pst-3SG.F	art.-door.FSG
	As we were sitting at the table to eat, (and) somebody knocked at the door [the door knocked]

In examples (16)–(18), the introduction of fi marks the direction, whereas, in (19), fi marks the location after a verbal adjective. In other words, in these data, fi is used to introduce a nominal with a directive or locative function. An initial remark concerns the almost identical syntactical structure of the same sentences in CyGr, which leads to the hypothesis that their presence in CyAr could have been triggered by the influence of CyGr and that this is a rather recent development. This hypothesis is further supported by the fact that Borg’s examples collected mostly during his fieldwork in 1984 and subsequent visits in Kormakitis do not display such a usage. On the contrary, Borg (2004, p. 3) explicitly contrasts CyAr (together with Maltese) with other Levantine varieties, in which the use of fi in this syntactic environment is indeed observed. Especially in the case of the nominal after a movement verb, he underlines: “Whereas most Arabic colloquials insert a directional or locative preposition after the verb, here both CyAr and Maltese show less explicit grammaticalization of this formal relationship and ordinary dispense with a particle”. The following examples from Borg (2004, p. 3; 1994, p. 57) illustrate the non-explicitation of the movement towards a place after the directive verbs <raχ> ‘go’ (20-22) and <χitu> “take” (23), as well as the absence of a locative preposition in (24):

(20)	ana	reχ	illi-mpakkál	taştri	peδ
	pron.1SG	go.part.MSG	art.grocer.MS	subj.buy.1SG	egg.MP
	I am going to the grocer’s to buy eggs

(21)	l-nes	piruχu	kúllayle	sala
	art.-people.MPL	prs.go.3PL	every nightFSG	vespers.FSG
	The people go every night to vespers

(22)	ruχu	lli-χkali	lákuon
	go.imp.2SG	art.-field.MP	find.imp.2S-obj.pro.
	Go to the fields and find them!

(23)	pkyaχtúχ	il-payt	tel	carús
	prs.take.3PL.-obj.pro.	art.-house.MSG	poss.	bride.FSG
	They accompany/take him to the house of the bride

(24)	lakáyt	ill-ummak	illi-Mtine
	meet.past.1SG	obj.mar.-motherFSG.poss.	art.-town.FSG
	I met your mother in the town (Nicosia) (from Borg 1994)

With regard to the CyAr examples (20)–(24) that show no use of the directional or locative fi, we need to note the following two remarks. The first remark concerns specifically the CyGr verb πάω ‘to go’, which appears in some local varieties with a non-prepositional NP, e.g., πάμεντε την χώραν ‘we are going [to] the city’ (city = the capital city, Nicosia), πάμεντε το χωρκόν= ‘we are going [to] the village’. The possibility that the use of the verb raχ (‘to go’) without fi in CyAr would be influenced by this basilectal CyGr structure, which is very marginally used today, cannot be easily proven because of the lack of written documents in CyAr that would allow us to inform such an assumption. However, it cannot be dismissed either, since we do have evidence of the retention by CyAr of basilectal CyGr elements, which have been levelled out in contemporary CyGr because of the influence of SMG at least at the lexical level: for instance, CyAr still uses words such as protsa ‘fork’ and kandila ‘glass’, which are obsolete/obsolescent in CyGr.

A second remark concerns the fact that, even in CyGr, the verb πάω does not always require the use of the preposition σε ‘to’ (just as in SMG); in many cases, the place of the preposition σε remains vacant, e.g., πάω θάλασσα ‘I am going to the beach’, πάω σινεμά ‘I am going to the cinema’, πάω στρατό/στρατιώτης ‘I am going to the army’, πάω σούπερμαρκετ ‘I am going to the supermarket’, πάω ψώνια ‘I am going shopping’, πάω κομμωτήριο ‘I am going to the hairdresser’s’ (Valiouli and Psaltou-Joycey 1995; Gülle 2016). Gülle (2014, pp. 164–65) comments on this as follows: “In written Greek, the locative or directive is usually marked with an inflected form of stos [sic] (which is probably a grammaticalization of the preposition for ‘in’, σε, as a proclitic on the article) and noun in the accusative. In spoken Greek (in Cyprus and on the mainland), this preposition is usually missing, and the directive or locative is only marked with the accusative case. Since Kormakiti Arabic lacks any morphological case, these constructions are not marked in this language”.

As a matter of fact, at least one of the examples in Borg’s corpus (1985, p. 173 also quoted in Gülle (2014, p. 165)), namely the expression πάω στρατό/στρατιώτης ‘I am going to the army’: ana smacat oti eprepen tarúχstratyoti ‘I heard that I was supposed to go to the army [literally ‘to go soldier’]’, could be considered as a transfer from CyGr probably prompted by the surrounding context. This example, however, cannot be counted as a case of an unmarked directive (contrary to Gülle 2016, p. 165) even though it contains the verb tarúχ ‘to go’ for the following two reasons: (a) stratyoti ‘soldier’ does not indicate a location, and, therefore the expression cannot be interpreted as a movement towards a place (directive); (b) tarúχ stratyoti ‘to go soldier’ is a collocation that combines one element from CyAr and another one from CyGr and presupposes no syntactic relation between the two words; it is a part of lexicon, not of syntax and indicates a (change in) state, i.e., ‘to become/to be a soldier’. Similar cases are in CyGr πάω σχολείο (litt.) ‘I go [to] school’ = ‘I am a pupil’ and πάω πανεπιστήμιο (litt.) ‘I go [to] university’ = ‘I am a student’ and also πάω ράψιμο ‘I am studying to become a tailor, I am an apprentice tailor’, as compared to πάω σινεμά ‘I am going to the cinema’, πάω θάλασσα ‘I am going to the beach (=sea)’, where the verb πάω retains its original meaning. According to Valiouli and Psaltou-Joycey (1995, p. 301), who analyzed in detail such structures in SMG, “the verb and the noun form a group, a synthesis, whereby the noun can acquire dynamic elements that allow it to express an action and energy, whereas the accompanying verb (gο, be, remain, etc.) can act as an empty verb that lends its characteristics (tense, aspect, person and number) to the noun transforming it into a verb” (our translation). As a consequence, collocations such as tarúχ stratyoti should not be included in any investigation of the directive usage of fi in CyAr.

Gülle (2014, pp. 164–65) argues that the use of fi with a locative meaning (ex. 18–19 above) might be an influence from Lebanese Arabic, but the use of fi to mark the directive is clearly an influence from (Cy)Gr, because such a use does not occur in other Middle Eastern varieties. Although a parallel influence from two different sources for two different uses of the same structure is not impossible, it is very uncommon; hence, Gülle’s assumption should be more carefully examined. Two further statements by Gülle (2014, pp. 165–67), namely (a) that the fi directive might be a relatively new development because it occurs in the speech of a young semispeaker and (b) that the fi locative might be the outcome of contact with Lebanese Arabic as it is suggested by the speakers themselves, are not confirmed either by our data from the Archive of Oral Tradition of CyAr or by sociolinguistic evidence. Regarding the former, the sporadic use of the fi directive in our sample of CyAr speakers aged 55+ years, who have a very good knowledge of the language, suggests that this might not be a new development that marks the speech of semispeakers, but must have been around for quite a long time before spreading down to the younger generation of speakers in their 40s. As for the latter, the high frequency of contact with CyGr and low frequency of contact with Lebanese Arabic together with the high degree of bilingualism in CyGr does not support the view of extensive influence from Lebanese Arabic, but rather from CyGr. However, we need to acknowledge that the Maronites of Cyprus feel strongly affiliated to the Maronite community of Lebanon (Karyolemou 2018) and consider Lebanese Arabic as a reference variety. We believe that the new uses of fi to mark the locative and directive functions that have developed most probably due to CyGr influence might have been reinforced by speakers’ positive attitudes resulting from the fact that fi is used in the prestigious variety of Lebanese Arabic to mark at least one of the two new functions, the locative function. As for the other, namely the directive function, its emergent nature in some but not all of the older speakers invites a more thorough investigation before any generalizations are made.

5. Conclusions and Future Directions

The investigation of the selected contact phenomena revealed the complexity of their nature, as structures and lexical items borrowed from CyGr show different degrees of integration into the lexico-grammatical system of CyAr. It is an uncontroversial fact that many of the changes that CyAr has undergone—some of which have been discussed here—are innovations when compared to other Modern Arabic varieties and thus render it more similar to CyGr. However, this is only part of the story. At the level of phonetics and phonology, for instance, while in some respects CyAr is becoming similar to CyGr, at the same time it diverges from it by extending the borrowed rules, such as yod occlusivization and manner dissimilation, to environments not encountered in CyGr. At the lexical level, partial relexification and the minimal morphological integration of borrowings has further consequences for the CyAr NP structure, which converges to CyGr. However, the marking of previously unmarked functions (locative and directive functions) results in increased similarity not exclusively with CyGr, but also with other Modern Arabic varieties. Thus, CyAr emerges as an interesting case of a severely endangered language with variable degrees of mixture resulting from the contact of a Semitic variety with an Indo-European fusional one at different points in their history of coexistence. However, even though contact-induced change is observable at all levels of grammar and the lexicon, we need more descriptions of contact-induced change as the basis for any discussion about whether CyAr is a mixed language and of what kind (Matras and Bakker 2003; Bakker 2003). Crucially, future studies should be based on the collection and/or analysis of modern data in order to reach more informed conclusions regarding the status of the language; the current study, which provides descriptions of contact phenomena in CyAr that are informed by the analysis of a uniquely big and diverse corpus of naturalistic oral speech data, can serve as a new reliable reference point in this ongoing discussion.