Subordination and coordination are two syntactic features that are employed in writing in both English and Arabic. It is often argued, however, that the two languages differ in their preference for either syntactic relation (Cf. Introduction and Note 3). English, it is said, makes use of more subordination than coordination, while Arabic favors the use of coordination rather than subordination. The present paper aims not to prove or disprove this claim. Its main focus is finding out how these syntactic relations are dealt with in English-Arabic translation in general. To achieve this aim, the research makes use of three types of corpus: Texts originally written in English, texts originally written in Arabic, and texts translated from English into Arabic. The texts surveyed are taken from different disciplines. The reason behind this is that the researcher wants to get to a generalization about the hypothesis above, whether it is proven valid or not. In other words, by varying the texts, the researcher wants to get to general remarks from which several researches could be triggered. In each of the texts at hand, the instances of subordination and coordination  were counted. The number of occurrences were then turned into ratios and the results of the three parts of the corpus were compared so as to know whether the translated texts are more like those originally written in English or those originally written in Arabic. Recommendations conclude the paper.

1. Introduction

1.1 Coordination and Subordination in English

In English, compound sentences (coordination) are used to express related thoughts which are more or less equal and carry approximately the same weight; that is, when both clauses of the sentence are offered as new information; they are usually equal both syntactically and semantically. In other words, it is a paratactic relationship that holds between the clauses (Quirk et al., 1985: 918).

With subordination, unequal ideas are expressed. One clause carries more weight than the other. The subordinate clause is presented as given or known information rather than new (ibid: 919). The relationship at work here is hypotaxis; the superordinate clause and the subordinate clause(s) are in hierarchical hypotactic relationship.

Somewhere else in Quirk's (ibid: 920) it is stated that the second unit in a sequence of coordinated units "gains focal prominence from its position", and that such prominence applies to the final element in a complex sentence. This prominence means higher "communicative dynamism" than there is in the initial part of the information unit (ibid: 1356-7).

This argument does not, however, conflict with the discussion in the previous paragraphs regarding subordination. It is said above that a subordinate clause is semantically subordinate to the main clause, meaning that the information therein is often presupposed as given or known. Consequently, the normal position of a subordinate clause is initial rather than final, since the new information is often presented as we linearly progress along the information unit (i.e. sentence). When the subordinate clause is shifted to a final position, it gains more weight, semantically, than it is assigned by means of its syntactic level.

Quirk et. al. maintain that coordination is used when ease of comprehension is sought, but also hold that a compound sentence, "especially with and, is vague in that it leaves the specific logical relationship to the interference of the speaker" (ibid.:1040-1). In a complex sentence, on the other hand, the sentence may be difficult to understand since the "content of the sentence may presuppose knowledge that is not generally available" (ibid.: 987).

In general, it is argued that the use of subordination, rather than coordination, will immensely help in making one's  writing more mature, sophisticated, interesting and effective. This tendency is very clearly noticed in English writing textbooks. Oshima (1991: 165) gives an example paragraph with too many compound sentences. She says that "the overuse of coordination makes it both boring to read and difficult to focus on the ideas expressed". Oshima then gives another version of the same paragraph, which she considers "more effective because the sentences have been combined through he use of subordination as well as coordination". The two cited paragraphs are given below:

 In another writing textbook (O'donnell, 1986: 48-49), the same tendency (i.e. favoring subordination over coordination as a method of writing) is obvious even in an exercise where the students are supposed to fill in the blanks with coordinators. To give yet another example, consider the following sentence, which is one of so many instances of subordination taken from a textbook meant for students preparing for the Cambridge First Certificate in English Examination:

 To sum up, subordination as well as coordination are both used in English writing. Subordination, however, is preferred to coordination, and it is considered a sign of mature writing. A piece of writing with too many coordinators is seen as immature and requires rewriting.

1.2 Coordination and Subordination in Arabic

A compound sentence in Arabic consists of "more than one simple sentence conjoined by one of a closed set of conjunctive particles (Holes, 1995:215). These conjunctive particles are wa , fa, θumma, ?aw, ?am, bal and  la:kin.⁽¹⁾ The wa is the most common connective device in Arabic. It is used to connect words, phrases, clauses, and sentences. Unlike the English and, the Arabic wa is repeated before every item coordinated with the one before, no matter how many items are listed. But the fact of the matter is that the Arabic wa is not one but several particles, each having a different function. It may be used as a coordinator or as an adverbial. It could be used to express oaths or accompaniment and many other usages.⁽²⁾

A complex sentence, on the other hand, is one in which one or more of the non-verbal slots (subject, object, adverb, adjective) in a simple sentence is filled by a clause which, if taken out, makes a free-standing simple sentence (ibid: 215). There are four types of dependent clauses in Arabic. These are listed in the following table together with the particles used in each type:

Table no. (1): Dependent Clauses in Arabic and the particles used in each ⁽³⁾

Arabic, as already mentioned, and as it has frequently been claimed by researchers, both Arabs and non-Arabs, is a language that favors coordination at the expense of subordination. In a study by Joy Reid (Reid, 1992), a computer text analysis was applied to essays by both native English speakers and non-native speakers from Arabic, Spanish, and Chinese language backgrounds. The results found that Arabic writers use more coordinate conjunctions of any other language background. The tendency to do so was ascribed to first language transfer. Along similar lines, Ostler (1987) found that long sentences conjoined with coordinating conjunctions are typical of Arabic Writing. Much earlier, in 1967, Kaplan compared rhetorical and syntactic styles of English and Arabic and found that in English subordination is considered more elegant than, and hence preferable to, parallelism, while the opposite holds for Arabic. ⁽⁴⁾

Some Arab scholars adopted this point and defended it. Aziz ( 1989: 214 -5) states that Arabic opts for coordination when English would use other devices, especially subordination.

Whether in coordination or in subordination, it is just the normal way of thinking in Arabic to start the sentence with the known entity "followed by the clause which describes what is informationally new" (Holes, 1995: 215). Holes also noted that "placing the lengthy dependent clause first offends against the rhythmic principle which requires that the main break in the sentence comes no later than about halfway through, and hence, that the heavier' elements come later in the sentence" (ibid: 215-6). This does not mean that a subordinate clause never starts a sentence. When it does, it is presupposed as old information or shared knowledge (between the speaker and the recipient(s). When it is shifted forward in the sentence, it is thus assigned more importance.

2. Description of Corpus

The research corpus consists of three parts: Texts originally written in English, texts originally written in Arabic and texts translated from English into Arabic. Due to the limited scope of the paper, only short parts of the texts were studied (one page long in most cases). The researcher does not wish to focus on a certain text type. This is why the texts were randomly taken from different text types. Again, each type could make the corpus of a separate study. In this paper, the researcher is only after a generalization.

The method of analysis followed in studying the texts at hand was simple, but, the researcher believes, very direct and straightforward: The researcher counted every instance of subordination and coordination in the texts. Then a ratio of coordination to subordination was figured out.

2.1 The English Corpus ⁽⁵⁾

These are five English texts taken from different fields of study. The following table shows the number of subordination and coordination instances in each text. Only compound, complex and compound-complex sentences are considered. Simple sentences are not included in the count. The bottom row of the table shows the ratio of coordination to subordination in each of the texts. For this purpose, compound-complex sentences are regarded as instances of subordination rather than coordination:

Table no. 2 number of subordination & coordination instances in the English Corpus

In all of the texts of this part of the corpus (except for text no. 1, which is an abridged and simplified reader meant for students of English as a foreign language, and for reluctant native readers) the number of subordination instances exceeds those of coordination. In text no. 2 ( a paper on literary criticism), only one compound sentence is found. The rest are either complex or compound-complex. The ratio, as the table shows, is 0.03 : 1, which is surely for the favor of subordination.

 Text no. 3 ( a study conducted by the Center for Contemporary Arab Studies at Georgetown University, USA) was prepared and written by native speakers of English and meant not for the layman but for the educated elite. In the part of the study that was analyzed, complex and compound-complex sentences account for 25 instances altogether, while compound sentences are only two, a ratio of 0.08 : 1.

 Text no. 4 is an anthology with which students of English literature are familiar: The Norton Anthology of English Literature. Not only the audience, but also the genre of this text make it imperative that the language used be "perfect". Subordination here, again, is much more prevalent than coordination. There are six instances of coordination against eighteen of subordination. This makes the ratio 0.33:1.

 The last English text is a part of a file on the Gulf States, published by the Economist. As it is the case with most newspapers and magazines, this one is targeted at audiences of different walks of life and levels of education. In the part under survey, there are more simple sentences than any other type of sentence. Still, complex and compound-complex sentences amount to 22, compound sentences 15.

 It is obvious then that an English text written by native speakers of the language will exhibit more subordination than coordination. This is, as hypothesized at the outset, typical of most English texts, except for those intended for certain types of audiences, especially learners of English at beginning and pre-intermediate levels.

2.2 The Arabic Corpus ⁽⁶⁾

 This part of the corpus is taken from works by several Arab writers. Five texts were studied  and instances of coordination and subordination were counted. But counting sentences and classifying them into compound or complex was not an easy a job as it was when working with the English part of the corpus.

For one thing, the punctuation marks in Arabic are not strictly rule-governed as they are in English. The full stop, in particular, is not the only mark that marks the end of a sentence; a comma is more frequently opted for. Some Arab writers are beginning to use a full stop at the end of every sentence, though. Another reason that adds to the difficulty of counting subordination and coordination instances in Arabic is the fact that some Arab writers tend to write very long sentences, some of which could be a paragraph long with one full stop at the end and so many commas in between. In some cases a sentence' would have two or more subordinators, and even more coordinators than that. This abundant use of conjunctive devices adds to the difficulty of categorizing Arabic sentences into compound or complex. The following Arabic text (text no. 1) is an example on such a text:

وعندما يعود في المساء إلى البيت، فانه يستحق الراحة والطعام الهنيء. وفي نهاية الأسبوع، تكون فترة استراحة وأعمال أخرى. ففي بلده، أعلنت البلدية أنها ستبني جسرا جديدا على النهر الذي يخترق البلد. وهذا يعني تمويلا من المواطنين، وجزء منهم لا يريد الجسر لأسباب بيئية، وغيرهم لأسباب أخرى، ولذاك فللمواطنين تجمعاتهم التي يتبادلون فيها الآراء. وتستضيفهم محطات التلفاز والإذاعة. وتفتح الصحف صفحاتها لهذه الآراء المختلفة. وجون سميث، هو أيضا، عضو في عدة جمعيات علمية ومهنية ونقابية، يمارس من خلالها حقوقه العامة، لا يخشى أجهزة الأمن، ولا اتهام أحد له بأنه يعمل ضد النظام، أو ضد الرأس الأكبر، وهو يعيش في دولة القانون والمؤسسات. ويوجد توازن بين السلطات التي تتابع وتراقب وتحاسب، وعنده الانتخابات كل أربع سنوات لتكون كشف حساب لصاحب المنصب، وفرصة للطعن في أعماله وانتخاب غيره، والجميع أمام القوانين سواء. والمال العام محفوظ محترم، وجون يعيش في مدينة نظيفة، في بيت نظيف، له كرامته واحترامه.

 (مجلة أفكار: العدد ،162 ، آذار 2002)

In the paragraph above, twenty coordinating conjunctions (conjoining clauses) are easily spotted. There are also seven subordinators. There are nine full stops, but this is not indicative of the number of sentences.

Given this difficulty in counting coordination and subordinations instances, the researcher took another route to do the job: He counted the number of coordinators between clauses rather than the number of compound sentences. To count subordination instances, subordinate clauses rather than complex sentences were counted.

In the text above, (i.e., text no. 1) the ratio of coordination to subordination would then be 3:1, which is obviously in favor of coordination.

The researcher would mention here that even those subordinate clauses in the text above are conjoined by means of coordinators. Complex sentences, in turn, are conjoined with whatever follows them by means of a coordinator, mostly wa. But the important thing to notice and to highlight in this context is that there is an obvious presence of subordination, but not as obvious as that of coordination.

Text no. 2 is a paragraph taken from a book on socio-politics by Naser el-Deen al-Asad, in which he tries to make clear the meanings of some Islamic concepts. The book, the writer says in the introduction, is meant for the youth, who are not necessarily specialized in Fiqh or Shari'a sciences. The part of the text that the paper studied is a paragraph (page 61) of 120 words. Added to these, there are 14 coordinators (11 instances of wa, and 3 instances of fa ). In contrast, only four subordinate clauses are present. These are underlined in the text below.

وقد شاركت المرأة المسلمة مشاركة واسعة في الحياة السياسية والاجتماعية والعلمية. وحقها في ذلك مثل حق الرجل. فالبيعة والشورى والحرب ورواية الحديث النبوي وتفسير القرآن والتدريس، قام بها الرجال والنساء، وشاركت فيها المرأة جنبا إلى جنب مع الرجل، في مجتمع متكامل متكافل. وبيعة النساء للنبي ورد فيها قرآن، ولها أخبار في كتب السيرة والتفسير. وعدد منهن بايعن مع الرجال في بيعة العقبة، وفي مناسبات أخرى مثل: بيعة الرضوان، ويوم الفتح، وحجة الوداع. وجادلت امرأة النبي في أمر صدر من زوجها فنزل بمجادلتها قرآن وغلب الله رأيها وموقفها. وحين أراد عمر بن الخطاب  وهو أمير المؤمنين أن يقلل من المهور تصدت له امرأة وبينت له أنه ليس له الحق في ذلك واستشهدت بآية من القرآن، فنزل عند رأيها ورجع عن رأيه.

 (ناصر الدين الأسد: "نحن والعصر: مفاهيم ومصطلحات إسلامية")

The orderly use of punctuation utilized in the sample text above is not typical of most Arab writers, though. In text no. 3, for instance, (taken from a novel by Jabra Ibrahim Jabra) the use of the full-stop is not as orderly as it is in text no.2. Therefore, the number of full-stops does not equal the number of sentences. To get to a correct ratio of subordination and coordination, the researcher counted the subordinate clauses first. There are four altogether: one purpose clause, one nominal, and two adverbial clauses. Coordinators, on the other hand, were as follows: 8 instances of wa,  one bal, and one ?aw. This makes the total number 10 coordinators, more than double the number of subordinators.

Text no. 4 (The first page of an article by Dr. Abdulrahman Yagi) has 16 coordinating devices, most of which are instances of wa. Subordinate clauses account for 6 instances, a ratio of 3:1

The following table shows the number of subordination and coordination instances, as well as the ratio of coordination to subordination, in the Arabic texts:

Table no. 3 number of subordination & coordination instances in the Arabic Corpus

2.3 The translated Corpus⁽⁷⁾

This part of the corpus comprises parts of texts originally written in English and translated into Arabic. It is the Arabic version of the texts that the researcher studied. With this part of the corpus, the same method of analysis used for part two of the corpus was applied; i.e., counting coordinating devices and subordinate clauses.

Text no. (1) is a translation of text no. (2) of the first part of the corpus (i.e. the English corpus), in which the ratio of coordination to subordination was the least: 1:33 (see table no. 2). In the translated version of the same text, 24 subordinate clauses were cited, almost the same number as those in the original text. Coordinators, on the other hand, accounted for 14 instances, most of which are wa, the rest are fa. In fact, these coordinators are not employed to produce compound sentences. They are simply resumptive, not additive. They do not join dependent clauses but preface sentences and paragraphs. If we are to consider them as coordinators, we will have a ratio of 2:1, in favor of subordination. Obviously, this is not a normal feature of Arabic, which favors coordination over subordination.

The following table shows the results of the analysis of the five texts in this part of the corpus.

Table no. 4 number of subordination & coordination instances in the translated Corpus

As table no. (4) above clearly shows, the translated corpus are more coordinated than subordinated. This, of course, is not typical of Arabic discourse as proven by the analysis of part (2) of the corpus. This being the case, translators need to take a different approach to dealing with subordination and coordination when translating from English into Arabic. An Arabic text that abounds with complex sentences is by no means Arabic-like, just like an English text with more coordination than subordination. To suggest solutions, we now turn to the following section.

3. From Subordination to Coordination

Translation-wise, the analysis above is very worthy of study. In fact, it is the base of this research's main contribution. As has already been shown and proved in this paper, Arabic and English vary as to the frequency of use of subordination and coordination, with English making more use of subordination, Arabic of coordination. Therefore, getting to a translation that would sound natural in Arabic would entail a higher frequency of coordination than of subordination. But, as also noticed in the analysis of part three of the corpus, this is not a common feature of translated language. Arabic texts translated from English often tend to exhibit an unusually considerable presence of subordination, which is not in line with Arabic fondness for coordination. The result is an unavoidable unnaturalness⁽⁸⁾ ascribed to the translation output.

To naturalize Arabic translations, a translator needs to employ coordination more than he does subordination. But, is this possible in light of the issues of information load and focal prominence discussed above. In Quirk et. al. (1985: 1041) ten complex sentences are listed together with their correspondent compound sentences with and, but, or so. The following lists show the subordinators and their correspondent coordinators in Quirk's examples: