1
PUNCTUATION MARKS IN ORIGINAL ARABIC TEXTS[1]
1. Introduction
A linguistic description of the use of different punctuation marks in the texts of Modern Standard Arabic (MSA) is hard to come by. This seems partly due to the relatively late introduction of these marks as we know them today into written Arabic texts which may explain the absence of any comprehensive treatment of punctuation marks in books on Arabic grammar.[2] Even in the few modern Arabic references which do touch on the rules for the use of punctuation marks, the treatment is sketchy and the rules seem to be mostly prescriptive, based on a number of contrived sentences rather than on a linguistic description of the actual usage of punctuation marks in authentic written samples representative of MSA texts. I therefore intend to study how punctuation marks are actually used in the written texts of MSA.
The objectives of the present study are, therefore, as follows:
(a) To describe the punctuation norms of Original Arabic (OA) texts.[3]
(b) To observe and record newly-emerging trends in the use of these punctuation marks.
(c) To assess the validity and comprehensiveness of the rules listed in Arabic references on the use of punctuation marks.
The research corpus is a randomly selected Arabic text which belongs to the literary genre of the short story. This OA text is a short story entitled al-Hafila Tasir 'Bus Walk,' by MAHMOUD AL-RIMAWI from Jordan. The corpus also contains a single passage taken from the OA text which covers most of the short story. This passage was depunctuated and presented to ten Arab teachers of Arabic at the university level (henceforth, the ATAs), who were asked to re- punctuate it. I closely examined the use of punctuation marks in the OA text, as well as in the ten different versions of the passage which was re-punctuated by the ATAs.
The research procedure used in this study can be summed up in the following steps:
(a) Different punctuation marks used by the writer of the OA text were sorted out first. For each punctuation mark, it was noted whether each of the ATAs used the same mark, a different mark, or no mark at all. The aim was to look for instances of correspondence and divergence in the use of punctuation marks by educated Arabs.
(b) The agreement of at least half of the ten ATAs with the OA writer, or among themselves, on the use of a particular punctuation mark in a given position was considered to represent a case of correspondence in punctuation usage. Conversely, the disagreement of at least half of the ATAs with the use of a particular mark by the OA writer, and their agreement to use an alternative punctuation mark, was treated as an instance of divergence.
(c) Instances of correspondence between the OA writer and the ATAs, or among the ATAs themselves, were taken to represent indicators of the norms of usage agreed upon by educated Arab writers, whereas cases of divergence were interpreted as signs of punctuation variation, instability, and/or newly-emerging trends in Arabic punctuation usage.
(d) High correspondance of usage with a given punctuation rule listed in Arabic references was understood to constitute evidence supporting the validity of that rule. On the other hand, a high ratio of divergence was taken to indicate that a given rule is not a true representative of the linguistic facts of punctuation usage, and that it therefore needs either to be rejected or modified.
The text analysis which I conducted revealed that the comma and the period constitute about 85% of the total number of punctuation marks found in the corpus. Other punctuation marks were scarcely used. Only four significant instances of correspondance were found in the whole set of data to represent all the uses of the colon, the question mark and the dash together. Likewise, only one instance of significant divergence, involving brackets (parentheses), was found. Such a small number of instances of these punctuation marks would be insufficient to warrant arriving at any reliable conclusions and/or generalizations in a study whose main objective is to find punctuation 'norms,' which can be safely deduced only by detecting as many cases of usage correspondence as possible. It has, therefore, been decided to deal exclusively with the comma and the period as representative of other punctuation marks, and to leave the examination of the uses of the other marks to researchers with larger corpora at their disposal.
2. results of the contrastive text analysis
The contrastive text analysis between the OA text and its ten duplicate versions revealed that the ATAs retained only 38% of the total number of the punctuation marks as originally used by the OA writer. This shows the relatively low level of correspondence in the use of punctuation marks between the OA writer and the ten ATAs. The above percentage also raises many relevant questions. Do these abundant cases of disagreement fall within the range of the permissible linguistic variation which is characteristic of normal language usage? If so, what are the rules which govern this wide variation in the use of punctuation marks in Arabic texts? In this variation catered for by the rules on the use of punctuation marks listed in Arabic references? These questions, and many others, are crucial for linguistic research on this yet unexplored territory of punctuation marks and their use conventions in Arabic texts. The present study is but an altempt to tackle some of these questions. Hence, it is important to carefully examine the following findings drawn from the contrastive text analysis of punctuation marks in original Arabic texts.
Results of the contrastive text analysis are reported separately below for the comma and the period. This section of the study is therefore divided into two major subsections, each reporting the results of the constrastive text analysis for one of the two punctuation marks as used in the OA text as well as by the ATAs in the ten versions of the OA duplicate passage. This means that the unit of analysis and discussion is a given punctuation mark. The rules of usage listed in Arabic references for each of the two major punctuation marks are listed first in every subsection. These rules are used as a background against which the actual usage of that punctuation mark is checked. The different uses of each mark are then examined as found in all the research data. Exemplary sentences are quoted from the corpus to illustrate specific uses of the given punctuation mark. It is perhaps necessary to point out that for the purposes of the OA text analysis undertaken in the present study, a sentence is defined as the minimum stretch of text which is structurally independent or self-sufficient in a way that would, at the end of it, allow a long pause in speech or a period, a question mark, or an exclamation mark in writing.[4]
2.1 the comma
Arabic references list many different rules for the use of the comma. It soon becomes clear from these rules that the comma is used in Arabic texts both in-tersententially, viz., to join one sentence to another, as well as intrasententially to set off different components within the same sentence. More specifically, the comma is used, according to the rules in Arabic references, between:[5]
a) short sentences which are semantically related and which together make up a complete and useful utterance;
(b) subordinate adverbial, adjectival or conditional phrases/clauses and their main clauses;
(c) items in a series, like days of the week, seasons of the year, etc.;
(d) the addressee in a vocative structure and the rest of the sentences; and
(e) the two constituents in a structure of apposition.
It was found from the text analysis that the comma was the punctuation mark most frequently used both by the OA writer and the ATAs. The comma represented 65% of the total number of all punctuation marks in the OA text. This was followed by the use of the period, which accounted for 20%. The remaining 15% of the punctuation marks was found to be shared by all the other marks used in the data. This high frequency of occurrence of the comma in the OA text should attract the attention of scholars, writers, teachers and learners of Arabic to the importance of the correct use and proper teaching/learning of this punctuation mark. The present paper, accordingly, seeks to give attention to the description of the different uses of the comma found in the corpus, as well as to arrive at an interpretation of the frequent occurrence of the comma in Arabic texts.
Out of the 28 commas originally used by the writer of the OA text passage which was presented, in a depunctuated form, to the ten ATAs, the ATAs used 11, 16, 17, 7, 16, 15, 14, 13 and 5 commas, respectively. Thus, the average number of commas used by the ATAs was 13. This represents a little less than 50% of comma agreement between the writer's and the ATA's uses of the comma in the original Arabic corpus. The above figures also show a wide range of disagreement among the ATAs themselves, with ATA «10 showing only 5 instances of agreement with the OA writer, and ATA «3 having 17 matches. It can be concluded from this that the range of discrepancy in the use of the comma by educated Arabs is wide indeed. But, as the principal objective of the paper is to discover the norms of use, instances of correspondence are to be given priority over those of divergence.[6]
2.1.1 instances of correspondence
Table 1 below shows that there are 17 instances in which the OA writer and at least half of the ATAs agreed to use a comma in the same position (positive correspondence). On the other hand, there are 8 instances in which half or more of the ATAs used a comma in the same position where the OA writer used a different punctuation mark (negative correspondence).
Table 1: Type, and frequency of comma correspondence[7]
Frequency / Positive Correspondence / Negative Correspondence9 / S1, S21, S26 / -
7 / S12, S18 / Sl2, S23, S28
6 / S11, S26 / S14
5 / S2, S7, S8, Sl0, S14, Sl6, Sl7, S20, S27, S28 / S6, S9, S15, S24
The above table shows instances of comma correspondence, both positive and negative. The fact that the same sentence, e.g. Sl2, appears under both positive and negative correspondence means, firstly, that the OA writer has used a comma in this sentence for which at least half of the ATAs also used a comma in the same position. This is a case of positive correspondence. Secondly, it means that the OA writer has used a different punctuation mark elsewhere in the same sentence where half or more of the ATAs used a comma instead (hence, negative correspondence). The figures under 'frequency' show, for example, that nine ATAs used the same comma as the OA writer in S1, compared with five in S2. It also shows that seven ATAs have agreed to use a comma in S28, where the OA writer used a different punctuation mark. All twenty-five instances of comma correspondence given in the above table are considered to be representative of agreed-upon usage of the comma in the Arabic punctuation system and are, therefore, examined closely below. Sentences with higher frequencies of agreement are tackled first.
A quick overview at the above twenty-five sentences in the OA text shows that the commas in these sentences can be subdivided, according to their position, into (a) sentence-final commas, and (b) sentence-medial commas. Commas in group (b) usually appear within sentences after dependent clauses, whereas those in (a) occur at the end of sentences. Table 2 below shows this positional distribution of the commas: ____
Table 2: Position of comma correspondence
Sentence-final / Sentence-medialS1, S6, S7, S8, S9, Sl0, S11, Sl2, S14, SI5, Sl6, Sl7, S18, S20, S21, S23, S24, S27 / S2, Sl2, Sl4, S26,
S26, S28, S28
It thus becomes evident that the single major use of the comma in Arabic is to mark the end of a sentence, a function typically reserved for the period in the English punctuation system. It is to be recalled that the first rule cited from Arabic references in Section 2.1 above states that the comma is used to set off, or join, 'short' sentences which together make up a 'long' utterance. The following are sample sentences from the corpus which demonstrate the above use of the comma in sentence-final positions:
(1 ) a – في المرة الأولى كنت أقود سيارتي، (S1)
b – أما هؤلاء الركاب، فقد سلموا أمرهم لب كما يبدو، (S20)
c – وانصرفوا يتبادلون فيما بينهم أحاديث صاخبة مفعمة بالرضا والفضول والمرح، (S21)
Nine ATAs agreed with the OA writer on using a sentence-final comma in each of the above sentences. Similarly, seven ATAs, together with the OA writer, concluded the following sentence with a comma:
( 2 ) إلا أن المفاجأة تمثلت في احتجاب الرؤية كلياَ، (S18)
Seven ATAs, however, used a sentence-final comma to end the following two sentences whereas the OA writer has used a period to terminate both sentences:
( 3 ) a – ولم يكن هناك سوى مقاعد قليلة متناثرة ومساحات فسيحة خالية، كحال الحافلات التي تستخدمها المطارات لنقل الركاب بين صالة المطار والطائرة، (S12)
b – [8][…] ربما لأعتقادهم أن الحافلات عندنا هي على هذه الشاكلة أو لثقتهم في براعة قيادتي، (S23)
It is to be noted that the six sentences in (1), (2) and (3) above vary in their length. The number of words in them ranges between six words in (l.a) to twenty-one in (3.a). Therefore, the reference to the use of the comma in Rule (a) in Section 2.1 after 'short' sentences is either unfounded or, at least, fuzzy. Yet, the fact that the OA writer, together with three ATAs, chose to conclude the 21-word sentence in (3.a) above by a period, rather than a comma, seems to suggest that there is a growing tendency in Arabic to terminate single, but long, sentences, with a period.
The above conclusion gains more credence when we look at the following sentence, for example, with respect to which the subjects were divided whether to use a sentence-terminal comma or a period:
( 4 ) ومع ذلك فما ان اعتليت كرسي القيادة حتى تمكنت على الفور من تشغيل الموتور والأنطلاق بالحافلة على الطريق المستقيم الذي أعرفه من قبل (S16)
The above sentence is fairly long. In (4), the OA writer, as well as five ATAs, used a sentence-final comma whereas the other five ATAs used a terminal comma.
We now move to examine sentences in the data where a comma is used medially to set off either (a) a main clause from a subordinate clause, or (b) a main clause from a fragment, viz., a phrase. A quick look at Table 2 above shows that there were seven sentences showing correspondence with instances of such commas in the corpus. Let us examine these medial commas in more detail, starting with instances of correspondence and moving on to those of divergence.
All four sentences below consist of main clauses preceded or followed by subordinate clauses:
( 5 ) a – ولما لم يكن أحد سواي مدركاَ لجسامة الخطر، ولما لم أكن قادراَ على متابعة أداء هذه المهمة المستحيلة وهي القيادة دون رؤية، فقد قررت أن أترجل (S26)
b – ولم يكن هناك سوى مقاعد قليلة متناثرة ومساحات فسيحة خالية، كحال الحافلات التي تستخدمها المطارات لنقل الركاب بين صالة المطار والطائرة (S12)
c – لكن الهرج والمرج كان يسودهم، فيما حقائبهم ملقاة كيفما اتفق هنا وهناك (S14)
d – […] كنت أقود حافلة فارهة عريضة، وكأطول ما تكون عليه هذه الحافلات الحديثة (S2)
In (5.a) above, nine ATAs agreed with the OA text writer on using the first medial comma and six ATAs on the second. Seven, six, and five ATAs agreed with the writer on the use of the medial comma in (5.b), (5.c), and (5.d), respectively. It thus seems that one of the standard uses of the comma in Arabic is to set off a subordinate clause from its main clause. This use is indeed listed in all Arabic references on punctuation marks cited in the bibliography (see AL-RASAN 1989: 23, MAHAIDAT and borini 1989: 68, ABU ajmiya 1998: 66; among others). There are many other instances in the corpus, however, which seem to run counter to the above rule and conclusion, as will be argued in section 2.1.2 below.
Finally, it remains to say a word on medial commas which are not used to separate two clauses as in the above sentences. In the following sentence:
( 6 ) […] فقد قفزت بلمح البصر الى الخارج، الى شارع ترابي […] (S28)
the OA writer used a medial comma to separate the noun phrase الى الخارج and الى شارع ترابي where the two NPs stand in apposition to each other. Half of the ATAs also used a comma, like the OA writer in the above sentences but the other half did not. The use of a comma to mark the boundary of the appositive relation thus seems only optional. Many Arabic references list this function of the comma among its standard uses in Arabic texts (see AL-RAWSAN 1989: 23, MUHAIDAT and AL-BORINI: 1989: 68; among others). However, the rules do not seem to account for this 'optionality' in the use of the comma. Such rules are, therefore, in need of modification.
2.1.2 instances of divergence
Before concluding this discussion on the different uses of the comma in the sample data, it may be worthwhile to look quickly at instances of divergence where the OA writer has used a comma, but at least half of the ATAs used a different punctuation mark instead. The following table shows these instances of divergence:
Table 3: Frequency of comma divergence[9]
Frequency / Sentences with Instances / of Divergence9 / S20
7 / SS, Sl3
6 / S28
5 / S2, Sl0, Sll, Sl6
Examining the above instances of comma divergence may, as was said earlier, shed light on areas of high punctuation variation and/or new punctuation tendencies in Arabic.
A quick look at the actual sentences listed in Table 3 above will reveal that in four out of the eight sentences, the OA writer's commas were replaced with zero punctuation marks by the ATAs, including those which appear with the highest frequencies in Table 3. The highest frequency of divergence is found in (S20) below:
( 7 ) أما هؤلاء الركاب، فقد سلموا أمرهم لي كما يبدو (S20)
Nine out of the ATAs used a zero punctuation mark instead of the comma used by the OA writer. This is a strong indication that a comma is not commonly used in Arabic to set off subordinate phrasal elements in sentences despite what the 'prescriptive1 rules on Arabic punctuation claim. (See AL-RAWSAN 1989: 23, MUHAIDAT et al. 1989: 68, and ABU AJMIYA 1998: 66, among others)
Likewise, five ATAs used no punctuation mark in the place of the two commas in the following sentence:
( 8 ) في المرة الثانية، بعد أسبوعين على الأولى، كنت أقود حافلة عريضة […] (S2)
Only one ATA used two commas as the OA writer had, while four used two dashes instead of the commas. Rules on punctuation usage in Arabic references do not, in fact, list this use of the comma. Rather, they state that double dashes are used to set off parenthetical sentence elements in Arabic texts (see AL-sheikh 1993: 156, al-rawsan 1989: 25, and abu-haltam 1988: 67, among others). The fact that five ATAs agreed not to use any punctuation marks around the parenthetical phrase بعد اسبوعين على الأقل shows that the above rule concerning the punctuation of parenthetical elements needs to be modified so as to accommodate such punctuation facts as the above. Moreover, the use of double commas, by the OA writer and one ATA, to mark off the parenthetical phrase in (82) seems to reflect a growing tendency in Arabic texts. This is probably due to the influence of punctuation norms in European languages, especially English, where setting off parenthetical sentence elements is one of the standard uses of double commas.[10] The above conclusion concerning parenthetical sentence elements gains more credence from the following two sentences: