Journal of Linguistics & Literature Vol 1, No 1, 2016 25

The Phonotactic Adaptation of English Loanwords in Arabic Anwar A. H. Al-Athwary* Abstract The phonological modifications made to English loanwords in Modern Standard Arabic (MSA) have come as a response to cope with the phonetic and phonological constraints in MSA sound system. These adaptations of loanword pronunciation clearly reflect the areas and effects of phonetic and phonological interference between the two languages in contact. For this purpose, more than 300 English words borrowed by MSA are analyzed. At the syllabic and prosodic level, mechanisms like cluster simplification, syllabic consonant conversion, gemination, etc. are found at work and by far systematic in MSA borrowings. Generally, it has been found that most of the regular adaptations at syllable level such as declusterization, syllabic consonant conversion, consonant lengthening and vocalic glide insertion, are motivated by linguistic constraints inherited in the phonological system of MSA rather than by extra-linguistic motivations. Keywords: phonotactic adaptation, Modern Standard Arabic, ______________________________________________________ *Najran University, KSA Introduction The phonological adaptation of loanwords is of two kinds: adaptation at the segmental (individual-sound) level and adaptation at the phonotactic (syllabic or prosodic) level. The present article focuses on the second type, i.e. the syllabic modification of English borrowings in Modern Standard Arabic (MSA). MSA follows its own rules and has its own characteristic types of syllable structure. The following facts on MSA syllable patterns are agreed upon among Arab and non-Arab linguists:

The Phonotactic Adaptation of English Loanwords in Arabic 26 Unlike English, vowel-initial syllables never occur in MSA; all syllables always begin with a single consonant. This statement is supported by the fact that “all vowels, when recorded (in isolation by spectrograph), are initiated with a glottal stop ? ” 1 (al-Ani, 1970, p. 22). Therefore, all native words and foreign words as well, which are supposed to start with a vowel are initiated with a glottal stop before the vowel. Unlike English, too, initial consonant clusters are not permissible, and the onset, which is a basic constituent of the syllable, is occupied by only one consonant. Final and medial two-consonant clusters are possible like in qaşr /CVCC/ „a palace‟ and ?aħmar /CVC-CVC/ „red‟. In the case of medial consonant sequence, the first member is the coda of the preceding syllable and the second one is the onset of the following syllable as in the word ?aħmar above. This implies that three-consonant clusters are not allowed. On the other hand, the sequence of two vowels is disallowed anywhere in MSA syllable structure (al-Matlabi, 1984, p. 235). There is a third specific type of sound sequence, which involves the sequence of two identical consonants and technically called “geminate”. By definition, gemination is a process by which a consonant is doubled 2, obtaining a long consonant as a result. Unlike English, gemination in MSA is abundant and serves a significant morphological and semantic function, and therefore contrasts with their corresponding simple consonants. Consider the following native minimal pairs in (2) in which l and k are geminated: (2) a.

b.

qalam

(n)

„a pen‟

qallam-

(v)

„to cut ones nails, flower beds, etc.‟

?akal-

(v)

„to eat‟

?akkal-

(v)

„to feed‟.

Distributionally, the geminates occur word-medially as in (2) and word finally as in ħaqq „a right‟, watadd „a peg, wedge‟, etc. The former occurrence is the most frequent and always comes in intervocalic position.

Journal of Linguistics & Literature Vol 1, No 1, 2016 27 It should also be mentioned that all consonants (including the semivowels) are potentially involved in the gemination process. The syllable nucleus [+ syllabic] should be either a short vowel or a long vowel. Thus, the number of syllables in an utterance will be identical to the number of vowels therein. This is in line with the universal principle of syllabification.3 Five syllable patterns are permitted in MSA (see al-Ani, 1970, p. 87, al-Matlabi, 1984, P. 238 and Hijazi, 1998, pp. 80-81) (C= consonants, V= a short vowel and V: = a long vowel): (3) (i)

/CV/

as in wa „and‟

(ii)

/CV:/

as in fi: „in(side)‟

(iii) /CVC/

as in qiţ „a cat‟

(iv) /CV:C/

as in ba:b „a door‟

(v)

as in milħ „salt‟.4

/CVCC/

The pattern ending with a vowel is an open syllable (CV and CV:) and that ending with a consonant is a closed syllable (others). Moreover, the first pattern (CV) is classified as a short syllable and all the rest as long syllables. The first three patterns have a higher frequency of occurrence. 2. Statement of the Research Problem With the modern technological and educational developments that have taken place in the world and their recent echoes in the Arab world, a need has arisen for transferring many technical concepts into Modern Standard Arabic (MSA) in many fields. As a result, Arabic finds itself face to face with an immense number of foreign terminologies, especially from English. The consonant cluster systems of English and MSA are totally different as mentioned in the introduction above. So, the adaptation of English consonant clusters into MSA involves several constraints. This paper, therefore, attempts to address the following question: how are MSA borrowings from English accommodated at the syllabic (phonotactic) level into the phonological structure of MSA? It aims at investigating the

The Phonotactic Adaptation of English Loanwords in Arabic 28 possible mechanisms and strategies used by MSA to incorporate English clusters into the phonological system of MSA. Research Methods Being a language of formal speech, formal education and mass media, MSA loanword materials meant for the analysis are compiled from different sources such as Arabic newspapers (as the main source), some Arabic magazines, textbooks, and dictionaries, and a number of Arabic websites. An additional source is the linguistic background and intuitions of the researcher himself as a native speaker of Arabic. The number of loanwords collected is more than three hundred items. The data collected are analyzed by using the descriptive and analytical methods. The Integrated Model Of Loanword Nativization Having a comprehensive view of various phonological models of different linguistic schools seems to be necessary so as to provide an adequate and rule-based explanation of all observations in the borrowing process. It also participates in overcoming the problems that may appear in analyzing loanword data. Borrowing, as far as methodology is concerned, serves as one of reliability and validity tests against various theories of phonology to make sure that whether a given theory will survive as a reliable Linguistic model or not. Danesi (1985c, p. 2) described the gathering and analysis of loanwords as “one of the most revealing empirical procedures for testing, developing or modifying theories of phonological structure”. Another significant role played by lexical borrowing is as “evidence for the psychological reality of linguistic processes” and this would mean that the process of Loanword adaptation “provides insights into the inner workings of a phonological system and offers evidence which can be used to evaluate opposing theories” (Thornburg, 1980, p. 524). It is generally accepted that there is no Linguistic theory or model that is ultimate or final. They are always subject to rejection, modification or revision, even those that have been proved to have a great degree of resistance against “test of time”.

Journal of Linguistics & Literature Vol 1, No 1, 2016 29

The phonetic, phonemic and generative views are the most common phonological theories developed by the end of the 19th century and throughout the 20th century. Each model has its own arguments and perspectives regarding borrowing and loanword adaptation. Though these models seem to have elements of truth in them, each one fails, on its own, to provide a full account and interpretation of various mechanisms of Loanword assimilation. Consequently, they have proved to be inadequate. Danesi (1985c) provided a convincingly critical account of different models of loanword assimilation stated above, showing on one hand, their contributions in the field of loanword nativization, and arguing their main shortcomings, on the other. He explains how such theories have been “trapped within the perimeter of their theoretical domains, pointing out that the main problem being that these models underline monolithic theories, that is, they are formulated in such a way so as to be accepted or rejected in toto (1985c, p. 12). These arguments are summarized in a form of a comparison between the three models as shown in table (1) below. The convenient solution for such methodological dilemma, as it seems, is to bring the divergent linguistic views together and follow a very selective approach to them while dealing with and analyzing loanword data. In this respect, Danesi notes that: The integration of the insights of different schools of phonology is the only really workable modus operandi in methodology… in using an “integrated model” of phonology, analysts will have various theoretical alternatives at their disposal to account for the data. They are not enclosed within the perimeter of one particular theoretical domain.

(Danesi 1985c, p. 41) What we need, then, as Danesi suggested, is an integrated or convergent model which makes use of implications and perspectives of various phonological theories as far as loanword assimilation is concerned. In other words, this model will include elements of any theory that seem to explain a certain phenomenon best. It seems, therefore, the most adequate one from an explanatory standpoint.

The Phonotactic Adaptation of English Loanwords in Arabic 30 Table (1): A comparison of the three models of loanword assimilation

SHORTCOMINGS

METHODLOGICA L PROCEDURES

LOANWORD TREATMENT

ADOPTORS

PHONETIC MODEL

PHONEMIC MODEL

GENERATIVE MODEL

Pre-structuralistis (Hermann Paul 1886) and many structuralists

Structuralisms (Pioneered by Weinreich 1953, 1957)

Generative philologist (initiated by Hyman 1970, 1973)

Loanwords are explained in terms of physical phonetics: a foreign sound is replaced by the most closely related (i.e. phonetically approximate) unit in the native phonetic inventory

Explained in terms of phonological system: a foreign sound is replaced by the closest phoneme that is „felt‟ to resemble it. Reference is made not to phonetic facts but to systemic ones and not to phonetic substitutes but to substitutes reflecting constraints of allophonic and distributional nature.

Can be viewed in terms of rules, which specify morphemic structure, and by rules which refer to surface phonetic structure. This means that the morphological categories, which were ignored by phonemic analysts, are involved in this model.

A contrastive analysis of the borrowing language and donor language in terms of phonetic approximation.

Structural phonemics and cognitive viewpoint of nativization.

Generative phonology in which foreign sounds are perceived in terms of underlying forms and binary distinctive features.

Has no explanation that can reconcile divergent patterns of phonetic approximation 2

Does not account for nonphonological (morphological) phenomena in loanword adaptation

Some loanword phenomena cannot be explained in terms of abstract underlying forms, but rather as examples of a strategy that maps the phonetic shape of the foreign word to its closest native phonetic sequence. Then we again have come full circle back to the theory of phonetic approximation

Journal of Linguistics & Literature Vol 1, No 1, 2016 31 The integrated model proposed by Danesi is based on Italo-Canadian loanword data from English, and consists of two main principles (1985c, pp. 21-39): i) The Paradigmatic Principle (PP) and ii) The Phonological Synchronization Principle (PSP). PP refers to the interpretation of foreign item morphologically. So, this principle will not be discussed here because is beyond the scope of the present paper. 4.1 The Phonological Synchronization Principle (PSP) The foreign sounds of the item in question are interpreted in terms of the syllabic, prosodic, phonemic and phonetic patterns (as the case may be) of the borrowing language. This principle implies two basic processes: a. Sounds and sound patterns that are identical in both the donor and borrowing languages will not be modified in any way. b. Differences in pronunciation will activate either repatterning processes which are tied to the phonological system, or simple phonetic substitution mechanisms. Thus, such model is eclectic in the sense that it has incorporated the three basic types of adaptation mechanisms of a morphological, phonological and phonetic nature that are attributed respectively to the generative, phonemic and phonetic models. The two principles stated earlier interact with each other to generate a nativized form which is, in most cases, indistinguishable in form-class shape and phonological configuration from native words. Such interaction, which occurs in determining the final shape of a loanword, is not a new concept. Danesi (1985), however, considered the morphological adaptation mechanisms as autonomous rather than extensions of phonological ones. The borrower, he noted, seems to make some adjustments purely on the basis of the internal paradigmatic requirement of the native language

The Phonotactic Adaptation of English Loanwords in Arabic 32 without any reference to the phonological configuration of the incoming item (p. 23). 5. ANALYSIS AND RESULTS It is obvious from the quick comparison between the two phonological systems of English (as the source language) and MSA (as the recipient language), given in the introduction, that the areas of interference do exist at both the phonetic level and the syllabic level. The involvement of the two levels in loanword adaptation is natural. Wells (2000) explained that in the borrowing process “the incorporating of a loanword from one language into another may involve not only the sounds (phonetic segments, phonemics), of which the word‟s pronunciation is compared, but also the positions in which those sounds are used (syllable structure, phonotactics), the phonetic processes they undergo (phonological rules) and their accompanying suprasegmental features (duration, stress/accent)” (p. 10). In this way, not only do the speakers of MSA replace the exotic English phonemes by familiar MSA ones, they will also reorganize the way the sounds are arranged to conform to MSA phonotactics. In the Integrated Model of Loanword Nativization, the adjustments to syllable structure of incoming words are an example of phonological repatterning (Danesi, 1985c, p. 37). This fact manifests itself clearly in a number of sub-mechanisms that operate within this framework and which include the following: declusterization, syllabic consonant conversion, consonant lengthening (gemination), vocalic glide insertion, syllable deletion, monophthongization, and change in vowel duration (i.e. quantity). As can be seen below, these sequential and prosodic modifications in syllabic structure of English loans show a high degree of regularity. In the following analysis, I will confine myself to the first four mechanisms. 5.1 DECLUSTERIZATION One of the clear-cut phonotactic constraints in MSA phonology is that consonant clusters (CCs) exist only word-medially and word-finally but never word-initially and they are always made up of two elements. Therefore, in the context of arabicized loans, the English initial consonant

Journal of Linguistics & Literature Vol 1, No 1, 2016 33 clusters (ICCs) will be subject to the process of cluster split or declusterization. It is made possible by means of two repatterning methods or mechanisms: i) the anaptyctic vowel insertion and ii) the prefixation of the prothetic syllable /?i-/. The anaptyctic vowel insertion is the most frequent mechanism whereby an anaptyctic vowel is placed after the first member of English ICCs and after the second member of English medial consonant clusters (MCCs) that consist of three elements. The examples in (4 a-b) will suffice: (4)

a) ICCs: kiri:m < cream kiristal < crystal firi:zar < freezer fulu:r < fluorine bulu:tu < Ploto burunz < bronze b) MCCs: kunturul < control 5 kumbiyu:tar < /k∂mpju:t∂/ „computer‟ ?ilikturu:niyy < electronic

In the case of the three-element sequences across word boundaries, the vowel is inserted after the first element of the sequence as in the loanword aysikrim from /aIskri:m/ „ice-cream‟. The consonant sequence may, however, remain as it is as in banknu:t6 from English „banknote‟ because the syllable structure of this compound is in line with that of MSA phonology. That is, the syllables /CVCC/ plus /CVC/ of this word are permissible in MSA. As can be observed in (4 a-b) above, the phenomenon of sound harmony (i.e. the spreading of the following vowel or the following glide) in this mechanism is clearly manifested. That is, the quality or category of the

The Phonotactic Adaptation of English Loanwords in Arabic 34 intrusive vowel is always determined by the quality of the vowel occurring in the subsequent syllable. At this point, it may be appropriate to note that such phenomenon seems to be general in languages which do not admit ICCs in their phonological structure (see for example Paradis and Lebel, 1994, p. 82 on Fula (a western African language), and Sharma, 1980, pp. 83-84 on Central Pahāri Language in India). It should be also noted that the intrusive vowel is usually short as i and u in (4 a-b). The short vowel a, however, might be used as an anaptyctic vowel, but it optionally alternates with i in loanwords like in (5) below: (5)

tara:nzi:t ~ tira:nzi:t < transit bala:zma ~ bila:zma < plasma fala:š

~ fila:š < flash

balasti:k ~ bila:sti:k < plastics bala:ti:n ~ bila:ti:n < platinum The vowel harmony is also maintained here because a and i still belong to the same category, i.e. both are front vowels. Now, we turn to the second mechanism of declusterization, i.e. the prefixation of the prothetic syllable ?i-. It is so striking but less frequent. The mechanism strictly applies to the English ICCs that start with the sibilant /s/, e.g. /str-/, /st-/, /sk-/, etc. To split such consonant sequences, the prothetic syllable ?i- (i.e. the glottal stop ? + the short front vowel i) is prefixed before the cluster constituting, with the first element of the cluster, a new syllable of the type /CVC/: (6)

?istira:ti:jiyyah < strategy ?istarli:niyy < sterling ?iskitš < sketch ?isti:riyu < stereo ?ista:ti:kiyyah < statics ?iskuwa:š < squash

Journal of Linguistics & Literature Vol 1, No 1, 2016 35 The break of the ICCs in this way makes the syllabic structure of English loanwords permissible to MSA phonology: in the new created syllable, the first /C/ (i.e. ?) performs as the onset of the syllable, the /V/ (i.e. i) as the nucleus and the first element of the consonant cluster as the coda. Needless to say that the use of the glottal stop ? is necessary here in order to conform with MSA phonological rule which states that syllables should always start with a consonant. In the case of the fist example in (6), i.e. ?istira:ti:jiyyah, the two mechanisms are involved, i.e. the prefixation of ?i- and the insertion of the anaptyctici after the second element of the three consonant cluster as the nucleus of the second created syllable (i.e. /-tr-/). The prothetic ?i- might be prefixed to some loanwords that do not have initial clusters like ?ismant from English „cement.‟ This case can be accounted for as a case of false analogy. There are some cases in the data where the two declusterization rules may alternatively apply as can be seen in (7): (7) a. Vowel Apantyxis

b. i-Prefixation

kili:ni:kiy

?ikli:ni:kiy

< clinical

bila:ti:n

?ibla:ti:n

< platine

bila:zma

?ibla:zma

< plasma 7

In other borrowings, the alternative application of declusterization mechanism in (7b), however, is infrequent and in many other cases seems to be unacceptable by MSA speakers.8 It would be very odd, for instance, to encounter loan forms like the ones in (8): (8) *?ikrista:l to refer to English „crystal‟ *?ibrunz to refer to English „bronze‟ *?ifla:š to refer to English „flash‟, etc. The discussion of the ICCs being split in English loanwords of MSA gives rise to a very important and controversial point which calls for some comment and clarification.

The Phonotactic Adaptation of English Loanwords in Arabic 36 Some Arab linguists (see al-Qinai 1998 & 2000) claimed that MSA terms in modern times (than ever before) tend to maintain the initial clusters of foreign vocabulary. al-Qinai (1998), for instance, supportively quoted alShihabi (1955) who in turn noted that “This rule (i.e. the rule of ICCs declusterization) has become somewhat out dated in Modern Standard Arabic wherein the flux of loanwords has made initial clusters of two consonants permissible without the need to insert any short or long vowels.” (p. 299). To provide an evidence to his view, al-Qinai argued that words like the French „stade‟ (i.e. stadium) can be either adapted as ?ista:d or sta:d in Arabic and the English item „tramway‟ as tra:m (ibid). Note that the statement concerning the English loan tra:m contradicts with al-Qinai‟s (2000, p. 21) notation where the MSA correlate of the English „tram‟ is transcribed as tira:m, i.e. with the break of English ICC. In fact we don‟t agree with such point of view for many reasons. First of all, it can be argued that al-Shihabi‟s earlier observation may hold true only with regard to many Arabic regional dialects, which permit initial clusters in their own syllable structures like dialects spoken in Al-Sham9 and AlMaghrib countries. This phenomenon can be accounted for by the fact that such dialects were largely influenced by foreign languages during the colonization era, especially by the French language and culture. For MSA, the case is totally different and as we discussed above that the rule is so strict that all foreign ICCs should be declusterized. In case of any anomalies occur (see al-Qinai, 2000, p. 15) they have to be considered as exceptional cases that are highly influenced by the indigenous dialects and by the level of education as well. The second argument is that there is a linguistic evidence against such claim. If we want to apply the MSA morphosyntatic rule of “?al-prefixation” (where ?al- is the definite article meaning „the‟) to the loanword sta:d stated by al-Qinai, we get the unacceptable word *?al-sta:d, and of course the correct alternative is ?al?ista:d, i.e. by prefixing the prothetic syllable ?i- to the word before adding ?al-. Thus, it can be inferred that accepting ICCs in MSA borrowings may cause many problems which, in turn, may lead to linguistic complexity. Another linguistic support drawn from my personal observations is related to the

Journal of Linguistics & Literature Vol 1, No 1, 2016 37 fact that English vocabulary with ICCs represents areas of interference to untrained Arabic learners of English in most Arab educational institutions. They usually and unconsciously insert a short vowel and break the ICCs of many English words. Finally, it can be concluded that the non-existence of ICCs in MSA native syllable patterns is marked as one of its languagespecific characteristics. Consequently, the repeatedly ICCs split of English loanwords in MSA is a matter that is related to the very phonological and morphological structure of Arabic (i.e. MSA), the change of which becomes impossible. Despite the fact that MSA has already incorporated a lot of foreign items into its lexicon as a response to the need-filling motive, to use Hockett‟s (1970) terms, this should not be utilized as a justification to modernize the language by wrong means, i.e. by altering or modifying its basic structure. The natural reaction towards the incoming items is rather to adapt them so as to conform to MSA morphophonemic structure. 5.2 SYLLABIC CONSONANT CONVERSION Vowels in English (both pure and diphthongal) typically behave as the nuclei (i.e. peaks) of the syllable and are assigned the feature [+syllabic]. Furthermore, the sonorants /ļ/, /m,/ and /ņ/ may also constitute the peaks of some types of final syllables as in English words like „little‟, „racism‟ and „mutton‟, respectively. These consonants become [+syllabic] and labeled as syllabic consonants. In the context of MSA syllables, however, [+syllabic] is only granted to the vowels and never to consonants which, as a result, never occur as syllabic consonants. The syllabic consonants found in some English loanwords are, therefore, converted into non-syllabic ones. In other words, each syllable whose peak is a syllabic consonant is changed to a syllable with a vowel as the syllable nucleus and the syllabic consonant as its coda margin, as in (9) below: (9)

a.

di:zil < /di:zļ/ „diesel‟ ni:kil < /ni:kļ/ „nickel‟ ka:bil < /keıbļ/ „cable‟ mu:di:l < /modļ/ „model‟

b.

dijital < /dıd3ıtļ/ „digital‟

The Phonotactic Adaptation of English Loanwords in Arabic 38 şandal < /sændļ/ „sandal‟ kirista:l < /krıstļ/ „crystal‟ c.

kartu:n < /kα:tņ/ „carton‟

The epenthetic vowel is either i, as in (9a), a as in (9b), or u as in (9c) whose quality is influenced by the spreading vowel in the preceding syllable, hence the vowel harmony occurs as in (9a-b) (both vowels in each case are front) except in mu:di:l (9a) and kartu:n (9c). The influence of English orthography (i.e. spelling) is evident, particularly in (9 b-c) wherein the quality of the intrusive vowel is affected by the vowel letters of English final syllables. The process of English syllabic consonant conversion can be couched in the form of a general rule in (10), (C, refers to the syllabic consonant): (10)

Eng. {C,}

MSA →

{VC}

Among the total number of the adapted syllabic consonants (8 cases as in (9)), the sonorant /ļ/ constitutes the sizeable number (7 cases (9a-b)) with only one case of the sonorant /ņ/ (9c). The syllabic sonorant /m ,/ is attested in one word, viz. ru:mati:zm from the English /ru:m∂tIzm,/ „rheumatism‟. It is, however, an exceptional case to the rule in (10) where the syllabic /m,/ is somewhat maintained and partially dealt with as a final consonant cluster, i.e.-sm. Alternatively, the English syllabic /m,/ is deleted so that we get the loan form ru:mati:z. 5.3 CONSONANT LENGTHENING The technical term for consonant lengthening or doubling is usually known as “gemination”. Blanc (1952) defined it as “the prolongation of the continuants and a longer closure of stops” (quoted by al-Ani, 1970, p. 77). As it has been discussed earlier that geminates are so abandunt in MSA and occur word-medially and word-finally with the former being the most frequent.

Journal of Linguistics & Literature Vol 1, No 1, 2016 39 In MSA loanwords, the germinates usually occur word-medially and, of course, intervocalically. The lateral l, and the semi-vowel y are the only geminates found in MSA loanword data as can be seen in (11): (11)

ġurilla < /g∂rIl∂/ „gorilla‟ fanilla < /flænl/ „flannel‟ millimitr < /mılmi:t∂/ „millimetre‟ millilitr < /mılıli:t∂/ „milliletre‟ filla/villa < /vıl∂/ „villa‟ ?ayyun < /aı∂n/ „ion‟

There is one case in which the geminate n occurs word finally. It is the loanword ţann from the English „ton‟. The gemination taking place here is in analogy with native words like ħadd „boundary, limit, penalty‟, ţall ‘drizzle, dew‟. It is convenient here to point out that consonant lengthening is treated as a syllabic phenomenon by many linguists. al-Ani (1970, p. 77) referred to it as “identical clusters”, and in Danesi‟s Model of Loanword Nativzation it is classified as one of the syllabic repatterning mechanisms. Danesi (1985c) provided phonological evidence to such classification. Referring to Ingria (1980), Lefen (1980), and Stemberger (1984), he argued that: In terms of syllable structure, the doubling process [i.e. gemination] can be explained by positing that length is a nonsegmental feature … It belongs to what Clements and Keyser (1984) call the CV-tier. The double consonants are, it would seem, ambisyllabic constituents filling two nonnuclear positions in a syllabic tree. (p. 37)

The Phonotactic Adaptation of English Loanwords in Arabic 40 The following metrical tree in (12) is given by Danesi for the English loanword „fatto‟ fact‟ in Italo-Canadian, an Italian variety spoken in Canada ($ = syllable boundary):

(12) C

V $

f

a

C

C

$

V

Segmental tier CV-tier

t:

o

In loan forms, and also in native words, a geminate occurs within a single morpheme, and is, therefore, called, in Hayes‟ terminology, “true geminate” as opposed to “fake geminate” which takes place across morpheme boundary. Moreover, true geminate cannot be broken up by epenthesis or metathesis rules (Hayes, 1986, p. 327). The split of the geminate -ll- in fanilla, for instance, by an epenthetic vowel results in the unacceptable *fanilila. The true geminate like -ll- in loanwords can be presented as in (13).

(13)

C

C

l: However, where the syllable boundary is concerned, the first member of the doubled consonant occurs as a coda of the preceding syllable, and the second always as an onset of the following syllable. The syllabic repatterning mechanism in English loanwords may be illustrated by analyzing two arabicized items, i.e. ġurilla and ayyu:n cited in (32) above in terms of metrical trees as can be seen in (35) below:

Journal of Linguistics & Literature Vol 1, No 1, 2016 41

(14)

a.

gurilla < g∂rıl∂

English Form $

$

$

C

V C

V

g

∂

ı l

r b.

V V

a ∂

$

C

V ∂

$

C

V C

V

ġ

u: r

i

$ C

C

l:

V

a

ayyu:n < a‟∂n

English Form $

Arabicized Form

Arabicized Form

$

$ C

n

C V



a

$ C C V C

y: u: n

The fact that geminates in Arabic performing a morphological and semantic function has nothing to do with geminates in loanwords. An explanation for the point could lie in the fact that these functions are related to gemination that is pertained to certain morphological patterns of native morphemes that do not follow in the available loanword data. As a sort of example at point is the doubling of the second consonant of MSA triconsonantal verbal root faεal- or/and the doubling of the second consonant plus the lengthening of the second vowel (vowel gemination) of the same pattern. This yields the morphological patterns faεεal-, and faεεa:l, respectively. The output is so productive as can be observed in (15) and which is not applicable to loan forms:

The Phonotactic Adaptation of English Loanwords in Arabic 42 (15)

faεal?akalto

faεεal ?akkal-

eat‟ „to feed‟

?akka:l „one who eats much‟

rasam- _____ „to paint‟ đaħakto laugh

faεεa:l

rassa:m „a painter‟

đaħħak„cause to laugh‟

đaħħa:k „one who laughs/ makes others laugh

much‟ jamad„to freeze‟

jammad-

_____

„cause to freeze‟

Thus, the presence of the geminates in loanwords of MSA is merely stimulated by the tendency on the part of Arab speakers to assign such characteristic to loans as a mechanism of nativization. 5.4 VOCALIC GLIDE INSERTION MSA disallows the combination of two (or more) different vowels in its phonotactics; this constraint suggests that a nucleus of MSA syllable is always composed of a single vowel and never be a sequence of different vowels. The expression “different vowels” in the previous paragraph should not be understood as to otherwise state that a sequence of two “similar vowels” might exist in MSA syllable structure. Rather, the only implication the expression may have is that “long vowels (in their underlying forms) can be considered as sequences of two identical vowels” (Mahadin, 1996, p. 44-5). To avoid the occurrence of two (or more) vowel sequences in MSA borrowings from English, MSA resorts to a phonological rule at work in order to alter such unacceptable sequences to acceptable ones; that is the rule of vocalic glide insertion.

Journal of Linguistics & Literature Vol 1, No 1, 2016 43 5.4.1VOCALIC GLIDE INSERTION RULE : The types of vowel sequences that found in English loanwords, and which are subject to the rule of vocalic glide insertion are mentioned in (16) below: (16)

a. two pure vowels such as /Io/, /υe/ /Iæ/, etc., b. a diphthong such as /I∂/, /oI/, etc and, c. a diphthong + a vowel such as /I∂υ/, /aIo/, etc.10

Now, consider the following examples: (17) a. ?ulu:mbiya:d < /∂Impıæd/ ‘Olympiad’ jiyu:lu:jiya < /dӠı’ɒl∂dӠı/ ‘geology’ ?infiluwanza < /ınflu’enz∂/ ‘influenza’ kalsiyu:m < /kæls ı∂m/ ‘calcium’ b. ra:diyu: < /reıdı∂υ/ ‘radio’ malu:riyu < /m∂le∂rı∂/ ‘malaria’ fi:diyu:< /vıdı∂υ/ ‘video’ kafiti:riya < /kæf∂tı∂rı∂/ ‘cafeteria’ The vocal glide insertion rule is applied in a total of 28 cases in the loanword data. The intrusive glide is either y (the most frequent= 26 cases) or w (very rare = 2 cases), and that is determined by the spreading of the preceding vowel, i.e. it is a back glide w after a back vowel as the last example in (17a) and a front glide y after a front vowel as the rest of examples in (17). The split of vowel sequences by means of a vocalic glide in English loanwords can be couched in a form of a rule in (18). (18)

English {VV}

MSA →

{VGV} G = the glide y or w

The Phonotactic Adaptation of English Loanwords in Arabic 44 The intrusive glide should not be treated as a member of a diphthong. If we analyzed it as such, we would have a sequence of two vowels (the created diphthong plus a vowel, preceding or following it). The alternative and adequate analysis is that the glide of the rule in (18) can safely be considered as a separate consonant (i.e. semi-vowel) whose function is to break the unpermitted vowel sequences in MSA loans. This observation also operates in some cases where replaces the first element of the English vowel ?i:dyulu:jiyyah from /aIdI‟ol∂d3I/ „ideology‟ and /fIzI‟ol∂d3I/ „physiology‟ or sometimes substitutes a hirwi:n from /her∂υIn/ „heroin‟.

the vocalic glide sequence as in fisyulujiya from diphthong as in

The glottal stop ? and the glide h may have the same function of the vocalic glide as stated by the rule in (18). The former may be inserted to break vowel sequences within words, while the latter is inserted when the MSA bound plural morpheme -a:t is suffixed to singular loanwords ending with the vowel u: as in si:na:riyu:-ha:t (pl.) „scenarios‟ and ša:mbu:-ha:t (pl.) „shampoos‟ . 3. CONCLUSION The analysis has revealed that the foreign items are adapted at the syllabic level. This would mean that not only do the speakers of MSA replace the exotic English phonemes by familiar MSA ones (the segmental adaptation), they also reorganize the way the sounds are arranged to conform to MSA phonotactics. The phonological analysis serves a very significant pedagogical purpose of drawing the attention of English teachers in the Arab homeland to most of the predictable linguistic interferences by bringing home to them the contrastive aspects of the phonology of Arabic and English. The syllabic analysis of English borrowings in MSA has revealed the following areas of interference which English teachers should take into account when introducing English to Arab learners of English: English ICCs (initial consonant clusters) may be repeatedly declusterized by the students because they are lacking in the MSA phonetic system as discussed above. The absence of gemination in English may represent

Journal of Linguistics & Literature Vol 1, No 1, 2016 45 another area of interference. For instance, The English sounds /m/, /d/, l/, /t/ in words like “common”, “sudden”, “pillar”, “butter”, etc. may be considered as geminates by Arab learners of English and mispronunciations as /komm∂n/, etc. may repeatedly occur. In addition to the influence of the Arabic gemination, this is partially affected by the orthographic form of English. The same constraints faced in adapting English clusters into the syllable structure of Arabic can be investigated again from the point of view of Optimality Theory and TCRS (Theory of Constraints and Repair Strategies) proposed by Carol Paradis and her colleagues NOTES 1. For reading convenience, the following phonetic symbols are used in the transliteration of Arabic words (both native and borrowed): ?

glottal stop (‫)ء‬

Ħ voiceless pharyngeal fricative (‫)ح‬

x

voiceless uvular fricative (‫(خ‬

Ţ emphatic voiceless plosive (‫)ط‬

ς

voiced fricative (‫)ع‬

pharyngeal

Ş

emphatic voiceless alveolar fricative (‫)ص‬

ġ

voiced uvular fricative(‫)غ‬

J

voiced lamino-palatal affricate (‫)ج‬

š

voiceless fricative(‫)ش‬

Đ emphatic voiced denti-alveolar plosive (‫)ض‬

q

voiceless uvular plosive (‫)ق‬

palatal

denti-alveolar

2. In MSA orthography, gemination is indicated by the optional diacritic symbol [w] (?at-tašdi:d) rather than by doubling the consonant. 3. One of the universal principles of syllabification is that every [+syllabic] segment is placed in nucleus of a syllable. Consequently, the number of syllables in any particular utterance is equal to the number of [+syllabic] segments in

The Phonotactic Adaptation of English Loanwords in Arabic 46 that utterance (Kahn 1976 – as quoted by Abu-Salim 1998: 296). 4. Mahadin (1996) claims that there are six syllable patterns in Arabic, rather than the five patterns cited in (3). He adds the long closed syllable /CV:CC/. This pattern, however, does not exists in MSA. 5. The MSA loanword ?al-kunturu:l from English „control‟ has come to mean „a section or department of examination/evaluation in school, university, or even in the Ministry of Education‟ in many Arab countries such as Egypt, Yemen, etc. This particular meaning, which does not exist in English, has most probably undergone a partial semantic shift, i.e. the word ?al-kunturu:l has taken on a partially new but related meaning. 6. In the Egyptian dialect, however, an anaptyctic/i/ is insertedafter /k/ in the loan / banknu:t/. 7. The last two examples in (7) are also reported in Shahin (1986: 322). 8. In older loanwords from Greek, Latin and other languages, the mechanism in (7b) is the most frequent and is followed in loanwords like (see Ali 1987: 108). i. ?afla:ţu:n < „plato‟ (Greek) ii. ?isţabl < „stable‟ (Latin) iii. ?uşţu:l < „fleet‟ (Greek), etc. 9. The term „Al-Sham‟ refers to the following Arab countries: Palestine, Syria, Jordan and Lebanon. 10. Sometimes one-vowel element is omitted to avoid the two-vowel sequences. The case at point is the omission of /∂/ in the English sequence „diphthong +∂‟ like /∂I∂/ and /a∂/ as in fayru:s from /vaI∂r∂s/ „virus‟.

Journal of Linguistics & Literature Vol 1, No 1, 2016 47 APPENDIX LIST OF ENGLISH LOANWORD DATA This is a list of the 300 loanwords, which are thought to be borrowed from English, and constitute the basic data of our analysis in this study. All items are nouns unless specified. SOURCE (ENGLISH)

ARABICIZED WORDS

Agenda

ajindah

Album

ulbu:m

Aluminum

alamunyu:m

Ampere

ambi:r

Anaemia

ani:miya

Asphalt

asfalt

Aspirin

asbiri:n

automatic (adj.)

u:tuma:ti:kiyy

Bacteria

bakti:riya

baking bowder

bikinġ bawdar

Balloon

ba:lu:n

Bank

Bank

Banknote

banknu:t

Bar

ba:r

Basalt

ba:zilt/ bazalt

Baseball

bi:sbu:l

Battery

baţţa:riyyah

Beer

bi:rah

Benzene

binzi:n

Bilharzia

bilha:risiya/ bilha:rsya

Billion

bilyu:n

The Phonotactic Adaptation of English Loanwords in Arabic 48 Biscuit

basku:t

Boot

bu:t

Bronze

Burunz

bulldozer

Bildawzar

Bus

ba:ş

Cabin

kabi:nah

Cable

ka:bil

cafeteria

kafiti:riya

Cake

Kayk

calcium

ka:lsiyu:m

Camera

ka:mira

capsule

kabsu:lah

Captain

ka:btin

caramel

Karamil

carbohydrates

karbuhi:dra:t

Carbon

karbu:n

carbonate

karbu:na:t

carbuhydarte

karbuhi:dra:t

card

Kart

caricature

karikati:r/ karikatu:r

carton

kartu:n

cartoons

(afla:m) kartu:n

casino

ka:zinu:

cassette

ka:sit

catalogue

katalu:j

cathedral

katidra:iyyah

CD (computer)

si: di:

cement

ismant/ asmant

Journal of Linguistics & Literature Vol 1, No 1, 2016 49 centimetre

Santimitr

central

sintira:l

ceramic

si:rami:k

cheque/ check

ši:k

chimpanzee

ša:mba:nzi

chips

Šibs

chloride

kulu:ri:d

chlorine

kulu:r

chocolate

šukula:tah

cholera

Kulira

cholestorole

kulusturu:l

chorus

ku:ras

christmas

Kirismis

cigar

si:ja:r

cigarette

sija:rah

cinema

si:nama

classicalism

kila:si:kiyyah

clinical (adj.)

iklini:kiyy

cobra

ku:bra

cocaine

ku:ka:i:n

cocktail

ku:kti:l

colonel

kuluni:l

commandos

ku:ma:ndu:z

Commonwealth

Kumunwilθ

computer

kumbiyu:tar

confederation

Kunfidraliyyah

congress

Kunġris

consul

Qunşul

The Phonotactic Adaptation of English Loanwords in Arabic 50 control

kunturu:l

coupon

kubu:n

cream

kiri:m

cricket

Kirikit

crystal

kirista:l

décor

di:ku:r

democracy

di:muqra:ţiyyah

dictator

dikta:tu:r

diesel

di:zil

digital (adj.)

Dijital

dinosaur

di:naşu:r

diploma

diblu:m

diplomacy

dibluma:siyyah

disc

Disk

disco

Disku

dish (sattelite)

diš/ duš

doctor

duktu:r

dollar

du:la:r

drama

dira:ma

dynamite

di:nami:t

dynamo

di:na:mu:

electron

ilikturu:n

enzyme

inzi:m

fahrenheit

Fahrinhayt

fantasy

fanta:ziya

fashism

fa:šiyyah

fax

fa:ks

federalism

fidra:liyyah

Journal of Linguistics & Literature Vol 1, No 1, 2016 51 FIFA

(al-) fi:fa

film

Film

filter

Filtar

flannel

Fanillah

flash

fala:š/ fila:š

fluorine

fulu:r

folklore

fulkulu:r

freezer

firi:zar

fuse

fiyu:z

gallon

ja:lu:n

garage

jara:š

gas

ġa:z

gene

ji:nah

general (a military officer)

jinara:l

geology

jiyu:lu:jiya:

glucose

juluku:z

golf

julf/ ġulf

gorilla

Gurilla

gram

jira:m

granite

jira:ni:t

graphite

jira:fi:t

greenwich

jirinitš/ ġirinitš

guitar

jita:r

hallo

alu:

hamburger

hambu:rjar

hard disk

ha:rdisk

hardware

ha:rdwi:r

helicopter

hi:lukubtar/ hulukubtar

The Phonotactic Adaptation of English Loanwords in Arabic 52 heroine

hirwi:n

hockey

hu:ki

holocaust

huluku:st

hormone

hurmu:n

hydrogen

hi:druji:n/ i:druji:n

ice-cream

aysikri:m

Icon

ayqu:nah

Ideology

i:dyulu:jiyyah

imperialism

imbirya:liyyah

inch

inš

influenza

infiluwanza

insulin

insu:li:n

intercom

intarkum

internet

intarnit

ion

ayyu:n

jacket

ja:kit

jazz

ja:z

jeans

ji:ns/ ji:nz

jeep (car)

Jayb

jelly

Jili

jocky

ju:ki

kangaroo

Kanġar

karate

ka:ritiyyah

kerosene

ki:rusi:n

kilobyte

ki:lubayt

kilogram

ki:lujira:m

kilometre

ki:lumitr

kilowatt

kiluwa:t

Journal of Linguistics & Literature Vol 1, No 1, 2016 53 lamp

Lambah

laptop

la:btub

laser

Layzar

liberalism

libra:liyyah

lobby

lu:bi

logistics

lu:jistiyyah

lord

lu:rd

malaria

mala:riya

marathon

maraθu:n

mark

Marikah

marines

ma:ri:nz

Marx

ma:riks

masonry

ma:su:niyyah

massage

masa:j

mechanic

mi:kani:kiyy

mechanisms

mi:kanisma:t

medal

mi:da:liyyah

megabyte

mi:jabayt

metre

Mitr

microbe

mi:kru:b

microfilm

Mikrufilm

microphone

mikrufu:n

microscope

mi:krusku:b

microsoft

mikrusu:ft

microwave

mikruwi:f

mile

mi:l

militias

mili:šiya

milliard

milya:r

The Phonotactic Adaptation of English Loanwords in Arabic 54 milliardaire

milya:rdi:r

milligram

millijira :m

milliletre

millilir

millimetre

Millimitr

million

milyu:n

millionaire

milyu:ni:r

mobile

mu:bayl

model

mu:di:l

modem (computer)

mu:di:m

monologue

munulu:j

motor

mutu:r

mouse

Maws

Nazi

na:ziyy

nickel

ni:kil

nion

niyu:n

nitrate

nitra:t

nitrogen

ni:truji:n

nylon

naylu:n

ohm

u:m

Olympiad

ulumbiya:d

OPEC (countries)

ubik

opera

ubira

orchestra

urkistra

oxide

uksi:d

oxygen

uksiji:n

ozone

uzu:n

panorama

banora:ma

parachute

barašu:t

Journal of Linguistics & Literature Vol 1, No 1, 2016 55 parliament

barlama:n

penalty (football)

Balanti

penicillin

binsili:n

pepper

bi:ba:r

petrol

bitru:l/ butru:l

phoneme

funi:m

phonograph

funuġra:f

phosphate

fusfa:t

piano

biya:nu

pizza

Bitza

plasma

bala:zma

plastics

balasti:k

platinium

balati:n

Ploto

bulu:tu

police

buli:s

pop (music)

Bub

poster

Bustar

powder

Bawdar

pragmatism

birjma:tiyyah

professor

burufisu:r

proof

buru:fah

prostate

burusta:tah

protein

buruti:n

protocol

burutuku:l

proton

burutu:n

pyjamas

bija:mah

radar

rada:r

radicalism

radika:liyyah

The Phonotactic Adaptation of English Loanwords in Arabic 56 radio

ra:diyu:

receiver (satellite)

risi:far

remote control

rimu:t kunturu:l

rheumatism

ru:mati:zm/ rumati:z

robot

Rubut

romanticism

ru:mansiyyah

routine

ru:ti:n

rugby

Rujbi

salad

Salaţah

saloon (car)

şa:lu:n

saloon/salon

şa:lu:n

sandal

Şandal

sandwich

sa:ndawitš

sardine

sardi:n

satellite

Satalayt

scanner

iska:nar

scenario

si:na:riyu:

secretary

sikirti:r

seminar

simina:r

senator

sinatu:r

shampoo

ša:mbu:

shorts

šu:rt

sketch

iskitš

software

su:ftwi:r

squash

iskwa:š

statics

istati:kiyyah

stereo

isti:riyu:

sterling (adj.)

istarli:niyy/ istirli:niyy

Journal of Linguistics & Literature Vol 1, No 1, 2016 57 strategy

istra:ti:jiyyah

studio

isti:diyu:

superman

su:barma:n

supermarket

su:barma:rkit

symphony

sinfu:niyyah

tactic

takti:k

tang

ta:nj

taxi

ta:ksi

technocratism

tiknuqra:ţiyyah

technology

tiknulu:jiya:

telefax

tilifa:ks

Telegraph

tiliġra:f

Telephone

tilifu:n

Telescope

tilisku:b

Television

tilifizyu:n/ tilfizyu:n

Telex

Tiliks

Tennis

Tinis

(The) Catholic

(al)-ka:θuli:k

thermometer

tirmu:mitr

Ton

Tann

Torpedo

ţurbi:d

Tragedy

tira:ji:diya

Tram

tira:m

Transistor

tira:nzistu:r

Transit

tira:nzi:t

Trillion

tirilyu:n

Tuna

tu:nah

Typhoid

tayfu:d

The Phonotactic Adaptation of English Loanwords in Arabic 58 Uranium

yu:ra:niyu:m

Vaigra

fayjra/ vayjra

Vanilla

fani:liya

Varnish

warni:š

Vaseline

fasili:n

Veto

fi:tu/ vi:tu

Video

fi:diyu:/ vi:diyu:

video clip

fi:diyu: kilib

Villa

fillah/ villah

virus

fayru:s

visa

fi:za

vitamin

fi:ta:mi:n/ vitami:n

volt

fu:lt

watt

wa:ţ

web

Wib

whisk(e)y

Wiski

winch

Winš

yard

ya:rdah

yoga

yu:ġa

zinc

Zink

Journal of Linguistics & Literature Vol 1, No 1, 2016 59 REFERENCES Abu-Salim, Isam. (1998. Stress and Syllabification in English. Arab Journal for the Humanities, 62: 295-313. Al-Ani. S. (1970). Arabic Phonology .The Hague. Mouton and Co. Ali, Abdul Sahib M. (1987). A Linguistic Study of the Development of Scientific Vocabulary in Standard Arabic. Kegan Paul: London. al-Matlabi, Ghalib F. (1984). fi l-?aşwa:ti l-luġwiyyah: dira:satun fi ?aşwa:ti l-madi l-εarabiyyah [On language sounds: a study of Arabic vowels]. Ministry of Culture: Baghdad. al-Qinai, Jamal S. (1998). Transliteration of Foreign Proper Nouns in Arabic: Graphological Proximity vs Transcription Equivalence. Arab Journal for the Humanities, 63, pp. 282-311. al-Qinai, Jamal S. (2000). Morphophonemics of loanwords in Arabic. Studies in the Linguistic Sciences, 30 (2), pp. 1-25. Danesi, Marcel. (1985c). Loanwords and Phonological Methodology. Marcel Didier Inc.: Québec. Hayes, Bruce. (1986). Inalterability in CV Phonology. Language 62 (2), pp. 321-351. Hijazi, Mahmoud F. (1998). Madxal ?ila εilmu l-luġah [An Introduction to Linguistics]. Dar Qibaa: Cairo. Hockett, Charles F. (1970). A course in Modern Linguistics. Oxford and IBH Publishing: New Delhi. Mahadin, Radwan S. (1996. Phonemic Representation of Long vowels in Arabic. Language Forum, 22 (1-2), pp. 41-62. Paradis, Carole & Lebel, Caroline. (1994). Contrasts from Segmental Parameter Settings in Loanwords: Core and Periphery in Quebec French. In C. Dyck (ed.), Toronto Working Papers in Linguistics 13

The Phonotactic Adaptation of English Loanwords in Arabic 60 (1). Proceedings of the MOT conference on contrast phonology, pp. 75-94. Shahin, Abdulsabur. (1986). ?al-εarabiyyah: luġatu l-εulu:m wa tiqniyyah [Arabic: the language of science and technology]. Dar al-I‟tisam: Cairo. Sharma, D.D. (1980). A Study of Loan Words in Central Pahāri. R.K. Malhotra, Punjab University: Chandigarh. Thornburg, Linda. (1980). Arabic Loan Phonology: The Assimilation of English Lexical Items. Linguistics, 18, pp. 523-542

Wells J.C. (2000). Overcoming Phonetic Interference. English Phonetics: Journal of the English Phonetics of Japan, 3, pp. 9-21.