Diversity in Language: Contrastive Studies in English and Arabic Theoretical Applied Linguistics

Diversity in Language

This page intentionally left blank

Diversity in Language Contrastive Studies in Arabic and English Theoretical and Applied Linguistics

Edited by Zeinab M. Ibrahim Sabiha T. Aydelott Nagwa Kassabgy

The American University in Cairo Press Cairo • New York

The American University in Cairo Press Cairo and New York Copyright © 2000 by The American University in Cairo Press 113 Sharia Kasr el Aim Cairo, Egypt All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the American University in Cairo Press.

Dar el Kutub No. 17881/99 ISBN 977 424 578 4 Printed in Egypt

Contents

Foreword Contributors A Note on Transliteration and Transcription of Arabic Words Abbreviations

v vii xi xiv

Arabic Language: Distinctive Features 1 El-Said Badawi An Opinion on the Meanings of i'rab in Classical Arabic: The State of the Nominal Sentence. Summary in English

1

2 Huda M. M. Ghali The Syntax of Colloquial Egyptian Proverbs

5

3 Devin Stewart Understanding the Quran in English: Notes on Translation, Form, and Prophetic Typology

31

Arabic and English: Comparative Studies 4 Nagwa Kassabgy and Mono Kamel Hassan Relativization in English and Arabic: A Bidirectional Study

49

5 Mohammad Al-Khawalda The Expression of Futurity in the Arabic and English Languages

70

6 Jehan Allam A Sociolinguistic Study on the Use of Color Terminology in Egyptian Colloquial and Classical Arabic

77

iv 7 Nancy G. Hottel-Burkhart The Canons of Aristotelian Rhetoric: Their Place in Contrastive Arabic-English Studies

Contents

93

Writing: Learning Style and Form 8 Maha El-Seidi Metadiscourse in English and Arabic Argumentative Writing: A Cross-Linguistic Study of Texts Written by American and Egyptian University Students

111

9 Cynthia May Sheikholeslami and Nabila el-Taher Makhlouf The Impact of Arabic on ESL Expository Writing

127

10 Loubna Abdel-Tawab Youssef Teaching "Form" in English Verse to Arabic Poetry Readers

147

Language Acquisition: Attitudes and Comprehension 11 Christopher W. Horger Dialectal Analysis of Freshman Writing Students' Attitudes toward American and British Dialects

162

12 Abdel-Hakeem Kasem The Acquisition of the English Copula by Native Speakers of Lebanese Arabic: A Developmental Perspective

179

13 Salwa A. Kamel Categories of Comprehension in Argumentative Discourse: A Cross-Linguistic Study

193

Follows English section

Foreword The need for creating a forum for an exchange of ideas and understanding in the fields of English and Arabic linguistics and teaching led to the First International Conference on Contrastive Rhetoric, held at the American University in Cairo (AUC), in February of 1999. This book includes manuscripts based on some of the presentations made at that conference as well as a number of papers by several other scholars. The opinions and ideas expressed in the manuscripts do not necessarily reflect our own, but we believe that they will contribute toward opening the field for further research. The focus of the book, which has been loosely organized into four sections, is on English and Arabic linguistics and teaching. The first section focuses on the Arabic language: its philosophy of tense, syntax, and the teaching of the Quran; the focus of the second section is on comparative studies; the third section looks at writing; and the focus of the fourth section is on language acquisition, especially in terms of learners' attitudes and comprehension. We would like to thank all the authors who submitted manuscripts, including those whose papers are not part of this volume. We are particularly honored to include the contribution of Dr. El-Said Badawi, a distinguished scholar in the field of Arabic linguistics and sociolinguistics. Our thanks also go to those who willingly gave of their time to review the various manuscripts and provide us with valuable insights. The editors also appreciate the support received from Mr. Mark Linz, Ms. Pauline Wickham, and Mr. Neil Hewison of the American University in Cairo Press. Finally, we dedicate this volume to our families, friends, and colleagues in the field. We are grateful to have had the opportunity to create a forum for an exchange of ideas, and we hope this volume will inspire future development in the fields of Arabic and English applied and theoretical linguistics as well as sociolinguistics.

Zeinab M. Ibrahim, Sabiha T. Aydelnott, Nagwa Kassabgy, editors

This page intentionally left blank

Contributors Jehan Allam ("A Sociolinguistic Study on the Use of Color Terminology in Egyptian Colloquial and Classical Arabic") is a senior Arabic language teacher in the American University in Cairo (AUC). Her research interests include issues related to teaching and sociolinguistics. She is currently involved in research on youngsters' effect on language.

Sabiha T. Aydelott (editor) teaches in the Freshmen Writing Program at AUC. She has a doctorate in education, with specialization in reading and writing, from the University of Tennessee, Knoxville. She has taught in Pakistan, Iran, Turkey, and the United States. Her research interests are in reading assessment, diagnosis and remediation, comparative studies, and reading and writing across the curriculum.

El-Said Badawi ("An Opinion on the Meanings of i'rab in Classical Arabic: The State of the Nominal Sentence") is professor of Arabic linguistics at AUC. He is the Director of the Arabic Language Institute and Codirector of the Center for Arabic Study Abroad (CASA). His book, Levels of Contemporary Arabic, is a landmark in the field of Arabic linguistics.

Huda M. M. Ghaly ("The Syntax of Colloquial Egyptian Proverbs") is an associate professor in the Department of English of the Faculty of Arts at ' Ain Shams University in Cairo. She received her Ph.D. in theoretical linguistics from the School of Oriental and African Studies, University of London, in 1988. Her dissertation was titled A Syntactic Study of the Nominal Piece and Its Temporals in dar'eyya Arabic Based on the Theory of Government and Binding. She is the author of several articles that focus on linguistic issues.

Mona Kamel Hassan ("Relativization in English and Arabic: A Bidirectional Study") is an Arabic language instructor in the Arabic Language Institute of the American University in Cairo and has done

viii

Contributors

research in the area of pragmatics and cross-cultural communication. She holds an M.A. in teaching Arabic as a foreign language from AUC. Christopher W. Horger ("Dialectal Analysis of Freshman Writing Students' Attitudes toward American and British Dialects") teaches in the Freshmen Writing Program at AUC. His field of interest is rhetoric and composition theory; he is a recent graduate of the TEFL program at AUC. His paper in this collection grew out of a research project for a sociolinguistics class. Nancy Hottel-Burkhart ("The Canons of Aristotelian Rhetoric: Their Place in Contrastive Arabic-English Studies"), an assistant profesor in humanities and social sciences at al-Akhawayn University in Ifrane, Morocco, teaches comparative rhetoric and orality/literacy in the M.A. program in applied humanities. Her Ph.D. in applied linguistics is from the University of Texas at Austin. She has taught and researched second-language writing and rhetoric since 1974, including a stint at AUC in the M.A. TEFL Program from 1987-1991. Zeinab M. Ibrahim (editor) is the Executive Director of the Center for Arabic Study Abroad. She received her Ph.D. from Georgetown University. Her research is in the fields of sociolinguistics and comparative studies.

Salwa A. Kamel ("Categories of Comprehension in Argumentative Discourse: A Cross-Linguistic Study") is a professor of linguistics in the department of English at Cairo University. Her area of specialization is syntax, and her other interests include stylistics and translation. She is the editor of Cairo Studies in English and The Symposium on Comparative Literature Proceedings. Abdel-Hakeem Kasem ("The Acquisition of the English Copula by Native Speakers of Lebanese Arabic: A Developmental Perspective") is a lecturer in Arabic language and culture studies in the School of Australian and International Studies at Deakin University, Melbourne. He is the editor of the Journal of Arabic, Islamic and Middle Eastern Studies (JAIMES), an academic, refereed journal published by the Faculty of Arts of Deakin University. Abdel-Hakeem is also currently working toward his Ph.D. in applied linguistics at La Trobe University in Melbourne.

Contributors

ix

Nagwa Kassabgy ("Relativization in English and Arabic: A Bidi-rectional Study"; editor) received her M.S. in teaching English as a foreign language (TEFL) from the American University in Cairo and is an English language instructor in the English Language Institute at AUC. She is also involved in teacher training and has done research on EFL vocabulary acquisition and teaching grammar. She is a founding member of EGYPTESOL, an affiliate of the international organization of Teachers of English to Speakers of Other Languages (TESOL). Mohammed Al-Khwalda ("The Expression of Futurity in the Arabic and English Languages") earned his M.A. in linguistics from the University of Jordan in 1990. He received his Ph.D., also in linguistics from Essex University in England in 1997. Currently he is n assistant professor at Mu'tah University in Amman, Jordan. His primary interests are syntax, tense, aspect, and temporal reference. Nabila el-Taher Makhlouf ("The Impact of Arabic on ESL Expository Writing") received her M.A. in English Literature from Brown University and her M.A. in teaching English as a foreign language (TEFL) from the American University in Cairo. She has taught English at a variety of levels, from kindergarten, primary, and secondary school, as well as at 'Ain Shams University in Cairo. She is currently an instructor in the English Language Institute at AUC. Maha El-Seidi ("Metadiscourse in English and Arabic Argumentative Writing: A Cross-Linguistic Study of Texts Written by American and Egyptian University Students") received her M.A. and Ph.D. in linguistics from Cairo University in 1996. Currently she is a lecturer in linguistics in the department of English at Minufiya University in Minufiya, Egypt. Cynthia May Sheikholeslami ("The Impact of Arabic on ESL Expository Writing") received her M.A. in teaching English as a second language (MATESL) from the University of Washington. Currently she is an instructor in the Intensive English Program of the English Language Institute at AUC. She studied contrastive rhetoric with Ulla Connor at the TESOL Summer Institute in Bratislava. Devin Stewart ("Understanding the Quran in English: Notes on Translation, Form, and Prophetic Typology") received his Ph.D. in Arabic and Islamic studies from the University of Pennsylvania in 1991, and received the Malcolm Kerr Award for the best dissertation in Middle East Studies in 1992. He is currently an associate professor of Arabic and

x

Contributors

Islamic studies in the Middle East Studies Department at Emory University in Atlanta, Georgia, where he teaches courses in Arabic and Islamic studies. His published works include Islamic Legal Orthodoxy (1998), and articles on Shi'ite Islam, Islamic law, and Arabic dialectology. Loubna Abdel-Tawab Youssef ("Teaching "Form" in English Verse to Arabic Poetry Readers") received her Ph.D. from the department of English at Cairo University, and her M.A. from St. John's College in Santa Fe, New Mexico. An assistant professor at Cairo University, she teaches English literature and rhetoric, translates books for children and writes articles on literary criticism, poetry, and travel writing. Currently she is teaching at AUC.

A Note on Transliteration and Transcription of Arabic Words Because the papers in this volume encompass a broad swath of issues dealing with Arabic linguistics, the editors have found it necessary to implement two distinct systems for representing Arabic: one a more or less standard transliteration (in italic type), and the second a simplified, American phonemic transcription (marked by oblique slashes). The latter system has been used in particular in cases that involve the representation of Egyptian colloquial Arabic (ECA) and in papers that treat the language from a phonemic perspective; in all other cases the former system of transliteration has been used.

Symbols used in transliteration Symbol ' (ordinary apostrophe) b t

th J h

kh d

dh

r z s sh ' (reversed apostrophe)

gh

Arabic equivalent

xii

A Note on Transliteration and Transcription

Geminated consonants are shown double. Short vowels are represented by a, i, and M; long vowels by a, e, i, and u; diphthongs by ay, and aw. Tamarbuta is indicated by either a or at, and a may also represent a final ya.

Symbols used in transcription /Symbol/

Phonemic description/Arabic equivalent

' b t th or s j or g H x d dh or z r z s sh S D T DH or Z

Voiceless glottal stop: fVoiced bilabial stop: M Voiceless alveolar stop: ° Voiceless interdental fricative: ° Voiced palato-alveolar fricative/voiced velar stop: £ Voiceless pharyngeal fricative: C Voiceless velar fricative: C Voiced dento-alveolar stop: j Voiced interdental/dento-alveolar fricative: J Voiced alveolar flap: J Voiced dento-alveolar fricative: -> Voiceles dento-alveolar fricative: o* Voiceless palato-alveolar fricative: o^ Velarized voiceless dento-alveolar fricative: o^ Velarized voiced dento-alveolar stop: o^ Velarized voiceless dento-alveolar stop: la Velarized voiced interdental/dento-alveolar fricative:

A Note on Transliteration and Transcription 9 gh f q k 1 m n h w y

Voiced pharyngeal fricative: Voiced velar fricative: Voiceless labio-dental fricative: Voiceless uvular stop: Voiceless velar stop: Voiced dento-alveolar lateral: Voiced bilabial nasal continuant: Voiceless dento-alveolar nasal continuant: Voiceless glottal fricative: Voiced labio-velar semivowel: Voiced palatal semivowel:

a a: u u: i i:

Low front vowel: Low front long vowel High back vowel: High back long vowel High front vowel: High front long vowel

Geminated consonants are shown doubled.

xiii

Abbreviations adj., adjective, adjectival AFL, Arabic as a foreign language AH, after the hijra (emigration of Prophet Muhammad to Medina) AUC, American University in Cairo BCE, before the common era (i.e., before birth of Christ) CA, classical Arabic CE, common era (i.e., from birth of Christ) cop., copula ECA, Egyptian colloquial Arabic EFL, English as a foreign language ELT, English language teacher ESL, English as a second language EoM, English-only movement FonF, focus on form L1, first language L2, second language MSA, modern standard Arabic MSJA, modern standard Jordanian Arabic NA, native Arabic NNA, nonnative Arabic NNS, nonnative speaker nom., nominative NP, noun or nominal phrase NS, native speaker Qur., Quran Rh., Aristotle's Rhetoric UG, universal grammar

An Opinion on the Meanings of i'rab in Classical Arabic: The State of the Nominal Sentence Summary in English El-Said Badawi The classification of Arabic sentences by classical Arab grammarians into nominal and verbal types on the basis of the one beginning with a noun and the other with a verb has recently been met with resistance by Western Arabists. They have argued that the semantic differences between a nominal sentence such as hamidun jalasa and a similar, but verbal sentence, jalasa hamidun, are so slight as to make such major structural differentiation between the two unjustifiable. For their part, Arab grammarians not only made the distinction, but they went so far in their differentiation between the two types that they not only assigned them to separate classes (i.e., nominal and verbal), but they also relegated different terms to each of their two basic constituents: the mubtadi' and khabar (subject and predicate) in the case of the nominal sentence and fi'l and fa'il) in the case of the verbal sentence. This paper, which is a part of a larger, ongoing research project, argues that in fact there is a—hitherto untreated—sufficient structural and semantic basis for maintaining the distinction.

Semantic Analysis Semantically, the argument is based on the fact that the contrast between the nominal and verbal sentences is not merely, as Arab grammarians have maintained, that the former denotes greater emphasis on the topic, but mainly that the verbal sentence denotes that a process or event is taking place along a grammatically framed time axis, whereas the nominal sentence denotes a static condition—an absoluteness. Since the elements of verbality and nominality (thanks to the derivational system of Arabic) are present in varying measurements in each of

2

An Opinion on the Meanings of i'rab in Classical Arabic

the morphological derivatives of the language (e.g., verb, verbal noun, noun of instrument, active participle, concrete noun, etc.) and because nouns (and indeed any of the other derivatives) are not limited to occurring in the initial position in the sentence or occurring in a certain number, it follows that it would be a semantic oversimplification to classify Arabic sentences into just verbal and nominal sentences. It is argued here, based on evidence of actual language usage, that nominal sentences versus verbal ones express two structural extremities, each of which stands in semantic opposition to the other. Between these two extreme boundary lines, there exist many sentence varieties, each of which is differentiated from the rest according to the particular mixture of verbal and nominal features that it exhibits. These relations can be expressed schematically, as in Figure 1. Starting from the extreme nominal boundary (NB), all the sentence varieties can thus be arranged on a scale of gradually diminishing nominal features until they reach the verbal boundary line (VB), where maximum verbal features are present. The opposite is also true of the verbal boundary line.

Figure 1. Relation of nominal and verbal features in sentence varieties of Arabic.

An Opinion on the Meanings of i'rab in Classical Arabic

Structural Analysis Structurally, the paper points out the parallel and supporting features of the grammatical cases al-halat al-i'rabiya, not merely as regards their mere grammatical values, but also (and for the first time, according to the best of my knowledge) as regards the contribution they have been discovered to be making to the total semantic value of the sentence as sketched above. Because of the complexity of the total picture and because what is presented here is only one part of a larger research project, we present here only the case of the nominal sentence, designated by Western Arabists as the equational sentence, which consists of noun + noun. Of the four cases (al-raf', al-nasb, al-jarr, and al-jazm), only al-raf and al-nasb operate within the two basic parts of the equational sentence. Examination of various occurrences of these sentence types reveals that halat al-raf is associated with static, immovable, assured, and absolute value in the sentence, whereas the opposite is true of halat al-nasb. The structural theme of noun + noun sentences is subjected to alterations by the association of one of three verbs/particles known as alnawasikh. These are kana and its sisters, inna and its sisters, and zanna and its sisters. Each of these groups when associated with the noun + noun structure bring with it semantic and grammatical changes, reflecting in its totality a degree of absoluteness (or lack of it) commensurate with the totality of these interrelated grammatical and semantic features. The grammatical changes are the function of the computation of the two cases of al-raf and al-nasb over the two positions of the equational sentences (i.e., subject + predicate). This computation yields four unique structure types as regards the distribution of the indicative and subjunctive cases. These are the following: 1) Pure nominal: noun (al-raf} + noun (al-raf) 2) Kana and its sisters: noun (al-raf) + noun (al-nasb) 3) 'Inna and its sisters: noun (al-nasb) + noun (al-raf) 4) Zanna and its sisters: noun (al-nasb) + noun (subjunctive) Semantically, the absoluteness of the equational sentence is gradually eroded in a form parallel with the grammatical one above. Kana and its

4

An Opinion on the Meanings of i'rab in Classical Arabic

sisters bring in time qualification (with al-raf in first position); 'inna and its sisters question, in varying degrees, the association between subject and predicate (with al-raf only in second position); and finally, zanna and its sisters, through varying degrees of doubt, question the plausibility of the sentence altogether (with no raf in either position).

Figure 2. Comparison of the nominal and verbal features of certain varieties of Arabic sentences.

The Syntax of Colloquial Egyptian Proverbs Huda M. M. Ghaly In the syntactic structure of the proverbs of Egyptian colloquial Arabic (ECA) as cited in Ahmad Taymur's book of Colloquial Proverbs (1986), there is always a phrasal or clausal category prior to the phrase with a finite verb or predicate, or matrix IP. In accounting for this syntactic behavior within the framework of the Minimalist Program of Chomsky (1995), this paper will show that the word order pertaining to these declarative sentences is not really free, because it is motivated by syntactic and semantic considerations. Since the proverbial declarative sentence requires a certain element of focus (i.e., information that is "new" and has the highest degree of communicative dynamism, i.e., the rheme, as distinct from topic, or theme), there is a strong feature in the complement (C; the head of the preIP position that determines whether the sentence is declarative or interrogative) of these declarative sentences. In other words, the presence of this strong feature in the C of these declarative sentences activates a rhetorical operation that necessitates the overt insertion of a base-generated phrasal or clausal category in that position. This, in turn, enables us to distinguish syntactically between these proverbial declarative sentences from declarative sentences of the same dialect that are not proverbial. The former sentences always have the structure of a complement phrase (CP; a declarative sentence that has an IP as a complement of its head and also a specifier to that head), because it has a strong feature in its C, but the latter declarative sentences may have the structure of either a CP or an IP.

Theoretical Background and Review of Relevant Literature Chomsky's Minimalist Program According to Chomsky's Minimalist Program (MP), operations of the computational system for human language (CHL) for constructing a sen-

6

The Syntax of Colloquial Egyptian Proverbs

tence recursively construct syntactic objects that are rearrangements of properties of the lexical items. The first operation of this computation, select, is a procedure that takes a lexical item from the numeration (N; the items in the lexicon) and introduces it into the derivation (the set of operations performed on the lexical items to produce the relevant structure). This process of derivation involves the operation merge, which takes a pair of syntactic objects and replaces them with a new combined syntactic object so that it may be interpreted at the logical form (LF; the semantic component of the string) interface. At some point in the computation to LF, there is an operation spellout, which strips away elements that are not relevant to LF, i.e., those elements that belong to the phonological component (Chomsky, 1995, p. 229). Whereas pre-spellout is overt (i.e., the constituents have overt phonetic form), the computation to LF after spellout is covert. Since "there is no clear evidence that order plays a role at LF or in the computation from N to LF" (p. 335), it is assumed by Chomsky (1995) that ordering applies to the output of morphology, which assigns a linear (temporal, left-to-right) order to the elements (p. 334), all of which are words or morphemes (X° categories) though not necessarily lexical items (p. 335). Accordingly, he regards ordering as "surface effects" on interpretation, and he feels that they "seem to involve some additional level or levels internal to the phonological component" that is "postmorphology, but prephonetic" (p. 220). In other words, "the distinction made in early transformational grammar between 'stylistic' rules and others" is still maintained by Chomsky (p. 324). Furthermore, he maintains that the scrambled element (the word or phrase that has been reordered and moved further to the front of the clause) is "a kind of adjunct, external to the major syntactic structure, [and] associated with an internal position that determines its semantic interpretation" (p. 324). As a result, full reconstruction, which is the formation of operator-variable constructions driven by full interpretation (FI; LF plus phonetic form, PF) that leaves part of a trace intact at LF and deletes only its operator, is restricted to the special case of an adjunct position (A') movement that involves operators (p. 326). The reason is that "reconstruction in the A-chains does not take place, so it appears" (p. 327). This in turn demonstrates that on strictly minimalist assumptions the only possibilities for adjunction are word formation and that the order assumed in the adjunction of a head to another head "seems rather obscure and may have no general answer" (p. 340). Chomsky still maintains that the CHL has move a (an operation that allows movement of anything, e.g., a word or phrase, anywhere, provided the movement is not prevented by other constraints). This is indicated by the fact that the "output conditions reveal that items commonly appear 'displaced' from the position in which the interpretation they receive is

The Syntax of Colloquial Egyptian Proverbs

7

otherwise represented at the LF interface" (p. 316). However, he now also holds the view that any displacement in language is basically reducible to morphology-driven movement and that the problems related to variable phrase (XP) adjunction do not really belong to the minimalist framework. It follows that "the primary and perhaps only case is a-adjunction" (the process by which any word is adjoined to any other, larger word) "to X , a a feature [i.e., features in words or morphemes] or (if the operation is overt) an X°" (p. 323). The reason for this restriction on a-adjunction is that this framework is concerned with last resort movement driven by feature checking within the computation (p. 319). However, it may be the case that by the strict merger of two elements or by the raising of an element, forming a chain with both elements then merging (p. 322), there are two terms but only one LF role, since "each of these is a category that is visible at the interface, where it must receive some interpretation, satisfying FI" (p. 322). But for Chomsky, such a structure is permissible only "if a is an adjunct that is deleted at LF, leaving just one term" (p. 322), such as when we have a case of "full reconstruction at LF, eliminating the adjunct entirely." Accordingly, the structure "[YP XP [yp-t...]]] (i.e., a-adjunction) is only interpreted at the trace" (p. 323). In such a case, "scrambling [is] interpreted by reconstruction" (p. 323), where the two-segment category, formed by adjunction, will be interpreted as a word by morphology. It follows then for Chomsky that "adverbials cannot be adjoined by merge to phrases that are 0-related [i.e., arguments or predicates]" (p. 330), because the adjunction of an adverbial to an XP that has a 6-role at LF to form the two-segment category [XP,XP], projecting from X, is barred when an XP is an adjective phrase (AP) or verb phrase (VP) (p. 329). This is why Chomsky believes that adverbs can "be 'base-adjoined' only to X or phrases headed by v (i.e., a verb form that has had affixes adjoined to it) or functional categories" (p. 330). He feels that apart from the fact that "adverbs seem to have no morphological properties that require XP-adjunction," there is no empirical evidence that adverbs form chains by XP-adjunctions (p. 329). In other words, "an adverb in pre-IP position cannot be interpreted as if raised from some lower position" (p. 330) and "the problem of optional raising" of the adverb can be solved by the Larsonian solution, in which a is incorporated without raising since it "appears in some higher position" (Chomsky, 1995, p. 331). Unlike the "adjunction of YP to XP" (p. 323), which does not fit easily into this general approach, the notion of a strong feature (a feature that can trigger movement) plays an important role in the Minimalist Program. The strong features are nonsubstantives that call for a category in their checking domains. In the lexicon, there are substantive elements such as nouns, verbs,

8

The Syntax of Colloquial Egyptian Proverbs

etc., with their idiosyncratic properties and some of the functional categories, such as the "complementizer (C)" (p. 240). Other functional categories that have semantic properties include tense (T) and determiner (D). When the functional category C is questions (Q; for interrogative sentences), it is interpretable (i.e., it has semantic content at the level of LF), in which case it need not be checked unless it is strong. And when it is strong, it must be checked by merge or by move by substitution or adjunction before spellout. If, on the other hand, a language has weak Q, it will remain in situ at phonetic form (PF). In referring to the discourse properties of English, Chomsky (1995) says that there is a null variant of the declarative C that must have been introduced covertly and must be weak since strength is motivated only by PF manifestation. However, despite the fact that "covert insertion of strong features is indeed barred," he still maintains that it "is not barred" if this "covert insertion of complementizers has an LF effect" (p. 294).

Other relevant literature Arguing against the assumption that word order in languages such as Japanese is strictly optional, Miyagawa (1997) provides evidence that its apparently flexible word order of indirect object-direct object (IO-DO) and DO-IO is base-generated (i.e., a lexical analysis), rather than involving optional VP-adjunction scrambling, since scrambling is a strictly optional movement operation. He also provides evidence that these two word orders involve argument positions (A-positions; e.g., a subject position or that of the complement of a verb, adjective, or noun), since they have properties such as binding, which can take place only in an A-position. As for the IP adjunction in Japanese, Miyagawa says that it involves A movement and PC movement. In the A movement, VP-internal materials such as the object appears to the left of the subject for case-agreement features. But the A' movement is motivated by focus. Concentrating on the PC movement, Miyagawa says that the accusative case, which is inflected for agreement (I), is licensed by the same functional category. Following Chomsky, Miyagawa assumes that languages like Japanese allow multiple specifier positions for a single head. Accordingly, he assumes that the functional head Agro (that is, the head in which there is object-verb agreement) incorporates in Agrs (that is, the head in which there is subject-verb agreement). Due to this fusion, we have a unitary functional head that checks both the nominative subject, in the lower IP, and the accusative object, in the higher IP node creadted by adjunction, forming [IPObj-acc [IPSubj-nom...Agro-Agrs]]. The notions of focus and topic have an acknowledged status in Universal Grammar (UG). Focus may be analyzed by analogy with quantifier phrases in the sense that it operates a quantification, effecting a partition of the uni-

The Syntax of Colloquial Egyptian Proverbs

9

verse (May, 1985), and it can occur either in overt syntax or in LF. Accordingly, focus can be realized both fronted and in situ. Phonologically, a focus constituent has always been associated with a prominence-leading accent (Chomsky, 1971). On the other hand, a topic is deaccented and separated from the sentence by an intonational break, i.e., in slow rates of speech speakers generally make a short pause between the topic and the phrase adjacent to it. As far as the syntactic analysis of the topic is concerned, Frascarelli (1997) maintains that there is no general agreement among authors whether a topic is extracted by movement from its argumentposition (Rochemont, 1989) or base-generated as an extrasentential constituent, coindexed with a predicate internal gap or clitic (Cinque, 1990). Frascarelli (1997) adds that one point that is generally agreed on makes a critical distinction between a topic and a focus: a topic in extraposed position is either an adjunct or a base-generated construction, while a focus is neither. Moreover, there can be only one focus while multiple topics are allowed. A focus cannot be resumed by a pronominal clitic, and cannot enter into coreference relations. A focus can only bind a pronominal provided it c-commands it, because in this case it is a syntactic operator. Another consideration from theoretical work that relates to the complementizers seen in proverbial declarative sentences in ECA has to do with the so-called CP hypothesis. This theory assumes that finite subordinate clauses in English that lack an overt complementizer (that-less clauses) should be analyzed as CPs with a null head, whether by adopting a rule of "that deletion" or through the lexical insertion of a null C° (a complementizer on a word, rather than a phrasal level; Chomsky and Lasnik, 1977). This hypothesis that finite subordinate clauses (with or without complementizers) share a common syntactic structure has been refuted by Doherty (1997), who has shown that there are significant differences between that and that-less clauses with respect to adjunction possibilities. He has provided evidence from adverbial adjunction, analyzing finite subordinate clauses in English without an overt C as finite IP complements, rather than as CPs with a null head.

Description of the Data There are basically six types of proverbial declarative sentences in ECA. The first type has a CP that has an embedded IP that is introduced by a subordinator such as /'in/ or /ba9dima/ generated prior to the matrix IP. The second type has a CP that has an embedded IP that is introduced by an NP operator such as the relative pronominal /'illi/ or the interrogative pronominal /min/ generated prior to the matrix IP. The third type has an

10

The Syntax of Colloquial Egyptian Proverbs

CP that has an embedded IP that has an imperative verb generated prior to the matrix IP. The fourth type consists of three subclasses of these proverbial declarative sentences, but all have a CP with an NP that is generated prior to the matrix IP. The first subclass has an NP that may have overt case and it is also in an embedded IP generated prior to the matrix IP. The second subclass of type four has an NP that is a nominal construct generated prior to the matrix IP. The third subclass of type four has an NP with a strong pronominal form that does not have deictic function generated prior to the matrix IP. The fifth type of these proverbial declarative sentences has a CP with an NP that is introduced by the vocative particle generated prior to the matrix IP. The sixth type of these proverbial declarative sentences has a CP with a PP generated prior to the matrix IP. Type 1: CP with an embedded IP introduced by a subordinator The first type of these proverbial declarative sentences has a CP with an embedded IP that is introduced by a subordinator, such as /'in/ or /ba9dima/, generated prior to the matrix IP (see sentences and their respective trees below). It should be noted that in these proposed syntactic configurations that have been designated in the light of the Minimalist Program, the "Larsonian solution" has been used, i.e., the elements of the internal domain (whether as arguments or not) appear in some higher position (Chomsky, 1995, p. 331). This is due to the fact that "there should be no adjunction to a 6-related phrase (a 0-role assigner or an argument, a predicate or the XP of which it is predicated)" (p. 323). These configurations have also made use of the simple transitive verb construction of Chomsky (1995) before tense (T) is added to form TP. Introduced by /'in/: (1) /'in fa:tak il-mi:ri 'itmarragh fi Tura:buh/ conditional + pron. infl. (3rd per., masc., sing.) + perfective verb + pron. infl. (2rd per., masc., sing.) + def. art. + noun + pron. infl (2nd per., masc., sing.) + imperative verb + prep. + noun + pron. infl. (3rd per., masc., sing.) Lit., "If the governmental job leaves you behind, roll yourself in its dust," meaning there is nothing better than a job in the public sector. Introduced by /ba9dima/: (2) /ba9dima sha:b waddu: il-kutta:b/ temporal + pron. + pron. infl. (3rd per., sing., masc.) + perfective verb + pron. infl. (3rd per., pi.) + perfective verb + pron. infl. (3rd per., sing., masc.) + def. art. + noun + pron. infl. (pi.) Lit., "After his hair became gray, they sent him to school," i.e., he has been asked to do something that is inappropriate for him.

The Syntax of Colloquial Egyptian Proverbs

11

12

The Syntax of Colloquial Egyptian Proverbs

Type 2: CP with an embedded IP introduced by an NP operator The second type has a CP that has an embedded IP that is introduced by an NP operator such as the relative pronominal /'illi/ or the interrogative pronominal /min/ generated prior to the matrix IP. This proverbial structure is shown in the sentences and their respective trees below. Introduced by the relative pronominal /'illi/ (3) /'illi yistoro rabbu ma yifDaHu:sh maxlu:'/ relative pron. + pron. infl. (3rd per., masc., sing.) + imperfective verb + pron. infl. (3rd per., masc., sing.) + noun + pron. infl. (3rd per., masc., sing.) + negative particle + pron. infl. (3rd per., masc., sing.) + imperfective verb + cont. of the negative particle + pron. infl. (3rd per., masc., sing.) + noun Lit., "Whosoever God shelters, nobody can expose (him)." Introduced by the interrogative pronominal /min/ (4) /min tarak 'adi:mu ta:h/ interrogative pron. + pron. infl. (3rd per., masc., sing.) + perfective verb + noun + pron. infl. (3rd per., masc., sing.) + pron. infl. (3rd per., masc., sing.) + perfective verb Lit., "Whosoever leaves his old (friend) is lost." Type 2: labeled tree diagrams With /'illi/ (proverb 3)

The Syntax of Colloquial Egyptian Proverbs

13

Type 3: CP with an embedded IP that has an imperative verb The third type has a CP that has an embedded IP that has an imperative verb generated prior to the matrix IP. This is exemplified by the sentence and its tree below. (5) /'imshi dughri yiHta:r 9aduwwak fi:k / pron. infl. (2nd per., masc., sing.) + imperative verb + adv. + pron. infl. (3rd per., masc., sing.) + imperfective verb + noun + pron. infl. (2nd per., masc., sing.) + prep. + pron. infl. (2nd per., masc., sing.) Lit., "(If you) follow the straight path, your enemy will not know how to attack you." Type 3: labeled tree diagram

14

The Syntax of Colloquial Egyptian Proverbs

Type 4: CP with an NP operator : There are three basic subclasses of type four of these proverbial declarative sentences. The first subclass is exemplified by sentence 6, in which there is a focused embedded CP that has itself a focused NP with overt case. The second subclass is exemplified by sentences 7-10, in which there are focused NPs with different internal structures. The third subclass is exemplified by sentences 11 and 12, in which the focused NPs have the internal structure of strong pronominal forms that have lost their deictic force. (6) /xayrin ti9mil sharran til'a / noun + genitive case + nunation + pron. infl. (2nd per., masc., sing.) + imperfective verb + noun + accusative case + nunation + pron. infl. (2nd per., masc., sing.) + imperfective verb Lit., "Good (being) done, evil returned." That is, instead of a reward for doing good, you get evil in return. Type 4, subclass 1: labeled tree diagram (proverb 6)

(7) /xi:r 'ir-rigga:la yiba:n 9ashshabbah7 noun + def. art. + noun + pron. infl. (masc., sing.) + pron. infl. (3rd per., masc., sing.) + imperfective verb + prep + def. art. + noun Lit, "When a woman's husband is rich, it is evident from her appearance." (8) /di:l 'il-kalb 9umru ma yin9idil/ noun + def. art. + noun + adv. + negative particle + pron. infl. (3rd per., masc., sing.) + imperfective verb Lit., "The tail of the dog, it is never upright," that is, old habits die hard. (9) /da waghak wala Dayyi 'il-'amar/ dem. pron. (masc., sing) + noun + pron. infl (2nd per., masc., sing.) +

The Syntax of Colloquial Egyptian Proverbs

15

conj. + emphatic particle + noun + def. art. + noun Lit., "This, your face and not the glitter of the moon (is its equal)." This is a very cordial way of complimenting someone on his appearance. (10) /dabbu:r wi zan 9ala xara:b 9ishshu/ noun + conj. + pron. infl (3rd per., masc., sing.) + perfective verb + prep. + noun + noun + pron. infl. (3rd per., masc., sing.) Lit., "A wasp, and it kept on buzzing to destroy its nest." That is, some people harm themselves. (11) /hiyya l-Hidda:ya tirammi kataki:t/ strong pron. (3rd per., fern., sing.) + def. art. + noun + pron. infl. (fern., sing.) + pron. infl. (3rd per., fern., sing.) + imperfective verb + noun + pron. infl. (pi) Lit, "She-the kite throws away chicks (that she has caught to eat)?!" That is, is it possible that the kite would throw away the chicks that she has caught for herself? Type 4, subclass 3: labeled tree diagram (proverb 11)

(12) /huwwa 1-kalb yi9ud widn 'axu:h/ strong pron. (3rd per., masc., sing.) + def. art. + noun + pron. infl. (masc., sing.) + pron. infl. (3rd per., masc., sing.) + imperfective verb + noun + noun + pron. infl. (3rd per., masc., sing.) Lit., "He-the dog bites his nephew's ear?!" Is it possible that people would really hurt others of their own race or kind?

16

The Syntax of Colloquial Egyptian Proverbs

Type 5: CP with an NP operator introduced by a vocative particle The fifth type of these proverbial declarative sentences has a CP that has an NP that is introduced by the vocative particle generated prior to the matrix IP. Several examples are given below as well as a tree diagramming this type. (13) /ya 'arD 'insha:'i wi-bla9i:ni/ vocative particle + noun + imperative verb + pron. infl (2nd per., fern., sing.) + conj. + imperative verb + pron. infl (2nd per., fern., sing.) + pron. infl (1st per., sing.) Lit, "You earth, crack up and swallow me." That is, I was so ashamed that I wished I could hide anywhere even it meant my being devoured by the earth. Type 5: labeled tree diagram (proverb 13)

(14) /ya 'arD ma 9ali:ki 'illa-na/ vocative particle + noun + negative particle + prep. + pron. infl. (2nd per., fern., sing.) + prep. + strong pron. (1st per., sing.) Lit., "You earth, no one is on you but myself." A description of an arrogant and conceited person. (15) /ya baxt min 'idir wi-9ifi/ vocative particle + noun + interrogative pron. + pron. infl. (3rd per.,

The Syntax of Colloquial Egyptian Proverbs

17

masc., sing.) + perfective verb + conj. + pron. infl. (3rd per., masc., sing.) + perfective verb Lit., "Oh lucky one, who has the ability to punish his wrongdoer, and yet he forgives." (16) /ya 'alb yakata:kit ya ma fi:k w-inta sa:kit/ vocative particle + noun + vocative particle + noun + vocative particle + relative pronoun + prep. + pron. infl. (2nd per., masc., sing.) + conj. + strong pron. (2nd per., masc., sing.) + active participial predicate Lit., "Oh heart, oh poor young chick, oh what is in you, and you are silent?" That is, my poor little heart is overcome with sadness. (17) /ya ma taHt 'is-sawa:hi dawa:hi/ vocative particle + relative pronominal + prep. + def. art. + noun + noun + pron. infl. (pi.) Lit., "Oh whatever is underneath this inadvertence, [you are] misdeeds," said of anyone whose behavior in reality is different from its appearance. Type 6: CP with an NP operator introduced by a preposition The sixth type of these proverbial declarative sentences has a CP that has a PP generated prior to the matrix IP. This type of proverbial structure is demonstrated by the sentences and their respective trees below. NP introduced by /bi:n/ (18) /bi:n 'il-ba:yi9 wi-shsha:ri yiftaH 'allah/ prep. + def. art. + noun + conj. + def. art. + noun + pron. infl. (3rd per., masc., sing.) + imperfective verb + def. art. + noun Lit., "Between the seller and the buyer, God is the Provider." God may provide for the seller and the buyer if they do not agree with one another to conclude the transaction. NP introduced by /9ala/ (19) /9ala lisaini wa-la tinsa:ni/ prep. + noun + pron. infl. (1st per., sing) + conj. + negative particle + pron. infl. (2nd per., sing., masc.) + imperfective verb + pron. infl. (1st per., sing.) Lit., "On my tongue, and do not forget me," that is, do not forget me as I have not forgotten you. NP introduced by /ba9d/ (20) /ba9d il-9i:d ma yinfitilshi 1-kaHk/ temporal particle + def. art. + noun + negative particle + pron. infl.

18

The Syntax of Colloquial Egyptian Proverbs

(3rd per., sing., masc.) + imperfective verb + def. art. + noun Lit., "After the feast, the cookies are not made," that is, there is a time for everything. NP introduced by /fi:/ (21) /fi:ha l('a)xfi:ha/ prep + pron. infl. (3rd per., sing., fern.) + emphatic particle + pron. infl. (1st per., sing.) + imperfective verb + pron. infl. (3rd per., sing., fern.) Lit., "(I am) in it (else) I will dispose of it," i.e., if I am not part of it, I will put an end to it. NP introduced by /zayy/ (22) /zayy 'il-magazi:b kulli sa:9a f(i) Ha:l/ prep. + def. art. + noun + pron. infl. (pi.) + universal quantifier + noun + pron. infl. (fern., sing.) + prep. + noun Lit., "As with crazy people, each hour [they are] in a different condition," i.e., he is very moody like a lunatic. (23) /zayy 'il-marakbeyya ma yiftikiru:sh rabbina 'ilia wa't 'il-ghara'/ prep. + def. art. + noun + pron. infl. (pi.) + negative particle + pron. infl. (3rd per., masc., pi.) + imperfective verb + discontinous negative particle + noun + pron. infl. (1st per., pi.) + prep. + noun + def. art. + noun Lit., "As with the sailors, they remember God only at the time of drowning." Type 6: labeled tree diagram With /bi:n/ (proverb 18)

The Syntax of Colloquial Egyptian Proverbs

19

Analysis of Each Proverbial Structure Type 1. If strength is motivated only by phonetic form (PF) manifestation (Chomsky, 1995), then it may be said that the strength of the C of the matrix CPs in sentences of type 1 is overtly manifested by the base-gen-

20

The Syntax of Colloquial Egyptian Proverbs

eration of the embedded CPs as the focused element. These embedded CPs are adjunct CPs: in the tree of proverb 1, the embedded CP is an adverbial of condition, which is introduced by the conditioner /'in/, and in tree of sentence 2, the embedded CP is an adverbial of time, which is introduced by the temporal /ba9dima/. I have adopted the view that these embedded CPs are "baseadjoined" (Chomsky, 1995, p. 330) in the C of these matrix CPs due to the strong feature in the C of these matrix CPs. This is because adjunction of maximal projections headed by a word category (e.g., YP and XP, where Y and X represent variables) does not "fit easily into this general approach." (p. 323). Moreover, "adverbials cannot be adjoined by merge" to phrases that are 0-related, i.e., those phrases that play a semantic role either as an argument or a predicate (p. 330). Stated another way, this strong feature in the C of these matrix CPs is eliminated by having the focused element (i.e., the embedded CPs) base-generated in its checking domain, rather than by overt movement. Furthermore, the lexical analysis has been assumed here because there is no specific categorial feature involved in this operation. Type 2. As with the sentences of type 1, the strength of the C of the matrix CPs in the second type is overtly manifested by embedded CPs that are assumed to be base-generated in the C of their matrix CPs due to the strong feature in their C. But the embedded CPs in sentences of type 2, unlike those of type 1, have coreferential small pros (a small pro is a covert pronoun that is the subject or object of a finite clause) within their matrix CPs, and this reminds us of Chomsky's (1995) assumption that in some languages "arguments [are] attached as adjuncts associated with internal elements" (p. 324). Accordingly, in sentence 3, this embedded CP is the internal argument of the matrix verb /yifDaHu:/; this is indicated by the fact that it has a coreferential object small pro in its matrix CP. In sentence 4 the embedded CP generated in its C has a coreferential subject small pro in its matrix CP, making it the external argument of the matrix verb /ta:h/. This is demonstrated by the respective trees of sentences 3 and 4. The difference between the embedded CP like that in type 2, as represented in sentence 3, and that in type 1, as represented in sentence 1 arises from the fact that the focused CPs in the sentences of the latter type are adjuncts, whereas those of the former are arguments. When the focused CPs are adjuncts, they are not associated with internal elements within their major syntactic structures, i.e., the embedded CPs do not have coreferential small pros in their matrix CPs. But when the focused CP is an argument, it does require an internal element within its major syntactic structure for its semantic interpretation. We may, accordingly, assume that the focused embedded CPs in proverbs like sentence 3 must be base-generated in an A-position in the C of their matrix CPs; whereas the focused

The Syntax of Colloquial Egyptian Proverbs

21

embedded CPs in proverbs like sentence 1 must be base-generated in an Apposition in the C of their matrix CPs. Another basic difference between the embedded CPs of type 2 and those of type 1 is that the former's embedded CPs are introduced by pronominals: a relative pronominal in sentence 3 (i.e., /'illi/) and an interrogative pronominal in sentence 4 (i.e., /mm/). As the relative pronominal (i.e., /'illi/) has Aproperties such as binding (i.e., in sentence 3 it binds the object small pro2 in its major syntactic structure), this demonstrates that in ECA there is also an A-position in the C of these embedded CPs. In sentence 4, it is the whole embedded CP that is the external argument of the matrix verb, as indicated by its subject small pro within the matrix CP. In either case, the pronominal in the embedded CP, whether it is relative or interrogative, functions as an operator in relation to the embedded IP it heads; accordingly, it may be regarded as the focused NP within these embedded CPs. In sentence 3 we have the noun /rabbu/, which also has A-properties such as binding. It binds the subject small pro in its major syntactic structure. But the noun /rabbu/ is a topic NP, rather than a focused NP in this embedded CP. This assumption is built on two premises: (1) this NP does not function as an operator in relation to the embedded IP it heads; and (2) it displays a different syntactic behavior from that characterizing focused NPs. Concentrating on the distinct syntactic behavior of the topic NP, we notice that it displays a flexibility of the movement that is not available to the focused NP. It has been extraposed from its pre-IP position, where it is assumed to be base-generated, to a post-IP position. Not being part of the focused element, this topic NP has been moved to the post-IP position. This movement of the topic NP /rabbu/ may be described as "not belonging] at all within [this] framework of principles" (Chomsky, 1995, p. 333) since it is a stylistic variation, which is not applicable to the focused NP. And in trying to account for this syntactic behavior of the topic NP without an overt complementizer, as distinct from the focused argument, within the framework of the Minimalist Program, we could maintain that this type of NP is base-generated as a multiple specifier of I, along with the subject small pro. Since both specifiers (i.e., the topic NP and the subject small pro) are checked by a single head (i.e., I or T), they may be regarded as multiple specifiers.3 This indicates that the A-position for the base-generation of a topic NP is distinct from that A-position in which a focused argument is base-generated be it a CP or an NP. Accordingly, it is only the topic NP that may extraposed, producing a stylistic variation. Type 3. The IP generated in the C of type 3 is similar to the embedded CPs of type 1 in that they are adjuncts, rather than arguments. Their status as adjuncts is indicated by the fact that they do not have coreferential small

22

The Syntax of Colloquial Egyptian Proverbs

pros, within their matrix CPs. However, the basic difference between both types of proverbial sentences is that in those of type 3 the focused category is an IP, rather than a CP, and that the strong feature found in the C of this type of proverbial structure is given overt manifestation by the obligatory presence of the imperative verb form in the embedded IP, rather than the imperfective or the perfective verb forms. In other words, the strength of the C of the matrix CP in type 3 proverbs is overtly manifested by the requirement that this embedded IP have an imperative verb only. Type 4. The fourth type of these proverbial declarative sentences in this variety of Arabic is characterized by their having an NP generated in their C as illustrated by sentences 6-12; and this in turn provides us with more evidence that there is an A-position in the C of these declarative sentences for the focused argument be it an CP or an NP. The strong feature in the C of this type of the ECA proverbs is overtly manifested in different ways, leading to their subclassification into different subtypes. The first subtype is exemplified by sentence 6. This subtype is composed of an embedded CP that itself has a focused NP, but one that has overt case manifested on its nouns. The second subtype is exemplified by sentences 7-10. These sentences have focused NPs that may be internally composed of nominal constructs, as in sentences 7 and 8, or complex NP structures, as in sentences 9 and 10. The third subtype is exemplified by sentences 11 and 12, in which the focused NP has the internal structure of a strong pronominal form that has lost its deictic force. Analyzing sentence 6 first, we find that there are overt case markers and the overt indefinite marker (i.e., the nunation) in the nouns /xayrin/ and /sharran/, both features of which are marked phenomena because there are no overt case markers nor an overt indefinite marker associated with nouns in ECA. The vowel /i/ in the noun /xayrin/ is the overt case marker of the genitive and the vowel /a/ in the noun /sharran/ is the overt case marker of the accusative with the final /n/ after the overt case markers in both nouns being the nunation marker. It is the presence of the overt case marker carried by the noun /xayrin/ and the fact that it is genitive that enables us to maintain that this noun is base-generated in a complementizer A-position, rather than in an A-position within the IP. Despite the fact that both nouns in 6 (i.e., /xayrin/ and /sharran/) are the internal arguments of their respective verbs (i.e., /ti9mil/ and /til'a/), it is only the former noun that has genitive case, indicating that it differs in its base-generation from the noun /sharran/ and providing us with evidence that these two nouns cannot be base-generated in the same A-position. The noun /xayrin/ acquires its genitive case as a specifier to the head C (the spec-head relation) of the matrix CP. It is to be noted that /xayrin/ is

The Syntax of Colloquial Egyptian Proverbs

23

generated in the focused CP, which is itself base-generated in the C of the matrix CP. The noun /sharran/, on the other hand, acquires its accusative case marker by being in a spec-head relation to the V of the matrix CP. Not being part of the focused category, the noun /sharran/ in sentence 6 has been extraposed to a position before its verb (i.e., /til'a/) as the result of "surface effects" mentioned above. This extraposition is a kind of stylistic variation that makes the word order in the matrix CP apparently similar to that found in the focused CP, making the proverb more harmonious. Sentence 6 also demonstrates that when the scrambled element is an argument whose semantic interpretation is determined by its overt case marker, there is no need for it to be associated with an internal element within the major syntactic structure. This is probably why the NP with the noun /xayrin/ does not have a coreferential small pro within its major syntactic structure, nor does the NP with the noun /sharran/ have a coreferential small pro within its major syntactic structure. In other words, the presence of the overt case marker alleviates the need for an internal element within the major syntactic structure of each of these nouns. Sentences 7-10 differ from sentence 6 in that it is only the latter sentence that has overt case markers. However, it may be said that the strength in the C of the former sentences, which are also proverbial declarative sentences, is nonetheless given PF manifestation. In sentences 7 and 8, there is a nominal construct base-generated in these sentences' Cs; in 8 there is also the adverbial particle /9umru / and the negative particle /ma/, both of which provide further evidence that this nominal construct must be in a position external to the major syntactic structure. In sentence 9, the strength of its C is indicated by the emphatic particle /la/, the demonstrative pronominal /da/, and the conjunction /wa/, all of which are generated in order to focus the noun /waghak/. In other words, the demonstrative pronominal in this sentence does not have a deictic function nor does the conjunction have a coordinating function. Similarly, in sentence 10, the presence of the conjunction /wi/, which has lost its coordinating function, focuses the noun /dabbu:r/. It is to be noted that since the NPs generated in the C of sentences 7-10 do not have overt case markers they have coreferential small pros within their major syntactic structures. It is only in sentence 9 that there is no internal element that determines the semantic interpretation of the NP that is base-generated in its C. This is because sentence 9 is a nominal sentence, in which case it is not possible to have a subject small pro. The nominal predicate, unlike the verbal predicate in Arabic, does not have inflections heavy enough for the local determination of a subject small pro A In sentence 11, the strong feature of its C is basically indicated by the base-generation of the strong personal pronominal (i.e., /hiyya/) in its C. Having lost its deictic force, this strong personal pronominal simply

24

The Syntax of Colloquial Egyptian Proverbs

recasts the whole sentence so that it rhetorically questions the possibility of a kite ever letting go of its prey. This strong personal pronominal is base-generated prior to another noun that is also assumed to be base-generated in a pre-IP position: it is /l-Hidda:ya/. As with the tree of sentence 3, sentence 11 likewise provides evidence that in ECA there is not only an A-position in C for the base-generation of these focused NPs but also another A-position in a pre-IP position for the the base-generation of a topic NP without an overt complementizer. But this NP may be regarded as a multiple specifier, along with the subject small pro since both specifiers are checked by a single head (i.e., I). As for the focused NP, it is checked by a distinct head from that which checks the subject small pro. It is checked by a declarative C with a strong feature; and it has an operatorlike function. Despite the distinctness of these two pre-IP base-generated NPs in sentence 11, they nonetheless constitute one NP in relation to the remainder of the sentence, and this is indicated by the agreement in gender and number between them (i.e. the pronominal /hiyya/ and the noun /l-Hidda:ya/). As both of them represent the contrastive element in this sentence, together they constitute the focused element since there can be only one focus (Frascarelli, 1997), forming one phonological unit associated with a prominence-leading accent (Chomsky, 1971). Both NPs together rhetorically ask whether the proposition within its major syntactic structure is true, i.e., they constitute a syntactic operator analyzed by analogy with quantifier phrases, effecting a partition of the universe (May, 1985). Type 5. In sentences of type 5 (13-17), the strong feature in their C is given PF manifestation by the base-generation of a focused NP that is introduced by a vocative particle. As the vocative NP in ECA does not have an overt case marker, it is likewise associated with an internal element that determines its semantic interpretation within the major syntactic structure. In sentence 13 (as shown by its tree) these internal elements are the coreferential subject small pros in both CPs, while in sentence 14, the vocative NP is the internal argument of the preposition /9ali:ki/ in the underlying nominal sentence /'ana 9ali:ki/ ("I am on you"), in which there is a prepositional predicate and an object small pro. Sentence 14 illustrates an important characteristic of this type of proverbial declarative sentence: the fronting of the prepositional predicate, placing it next to the vocative NP and changing the assumed underlying structure /'ana 9ali:ki/ ("I am on you") to /ma 9ali:ki 'illa-na/ ("Not on you except me"). This fronting emphasizes that predicate, and by addressing the internal argument of the prepositional predicate and making it the

The Syntax of Colloquial Egyptian Proverbs

25

vocative NP (i.e., /ya 'arD/,"Oh, you Earth"), the emphatic effect is even greater. In other words, with the base-generation of this focused element (i.e., the vocative NP) in the C of this type of the nominal sentence (i.e., with a prepositional predicate), the fronting of the prepositional predicate is obligatory. But this obligatory fronting of the prepositional predicate, which takes place in the nominal sentence with a vocative NP base-generated in its C, can only be regarded as falling within the domain of the rules that have been referred to by Chomsky (1995) as the "surface effects" (p. 220). This is because this fronting is contingent only on the presence of a vocative NP in a sentence with a prepositional predicate, and not on the presence of a strong feature in a nonsubstantive category. This obligatory fronting of the prepositional predicate in the nominal sentence with a vocative NP base-generated in its C is found not only in sentence 14 but also in sentence 17, in which /taHt 'issawa:hi/ is fronted due to the generation of the vocative NP (i.e., /ya ma:/). Another important characteristic of these proverbial sentences that have a vocative NP is demonstrated by sentences 15 and 16: it is the recursiveness of the focused element. Looking first at sentence 15 as an example, if we assume that this sentence is derived from the underlying structure /'illi 'idir wi 9afa baxtu kwayyis/ ("whoever has the ability to punish and yet forgives has good luck"), then the vocative NP (i.e., /ya baxt/) refers to the one who is in possession of this good fortune because he has the above qualities. The other focused element in this sentence is the compound verbal clauses (i.e., /min 'idir wi 9ifi/), which describes the qualifications of the one who is in possession of this good fortune. Sentence 16 provides us with further evidence of this recursiveness of the focused element in ECA. It has three focused elements: the first vocative element is the NP (i.e., /ya 'alb/), which describes the object she is addressing. Being the first vocative, it is base-generated prior to the other vocatives in this sentence's C. The second vocative NP, which is /yakata:kit/, is base-generated adjacent to the first vocative NP, describing the heart as a little chick. This provides an even more focalizing effect to the first vocative NP. The third vocative element is a nominal clause (i.e., /ya ma fi:k/), which describes the second vocative NP as full despite of its small size. In other words, the first vocative addresses the heart, the second describes the size of that heart, and the third vocative states the full capacity of that heart with the conjunction /wi-/, focalizing these focused elements. It is to be noted that the proposition in this sentence's major syntactic structure is not deleted: it is /inta sa:kit/ ("you [masc. sing.] are quiet and tolerant"), while in sentence 15 the proposition in its major syntactic structure is deleted. The proposition in the major syntactic structure of sentence 16 is not deleted because

26

The Syntax of Colloquial Egyptian Proverbs

it is emphasized, as shown by its having the strong pronominal form (i.e., /inta/) as the subject NP. In other words, the meaning of this sentence would not be complete had the proposition in its major syntactic structure been deleted. Type 6. The prepositional phrases (PPs) in sentences of type 6 should also be regarded as being base-generated in the C of their sentences because adverbs do not form chains by XP-adjunction and because the adjunction of an adverbial to an XP that has a 6-role is barred when an XP is an adjectival phrase or a verbal phrase (see above, "Theoretical background and review of relevant literature"). The PPs in these sentences are also similar to the embedded CPs in sentences of type 1 in that they are focused elements that are adjuncts, and this is indicated by the fact that they do not have coreferential small pros within their major syntactic structures. Accordingly, these PPs are generated in an A'-position in their C. As with the adjunct CPs in type 1 sentences, these PPs of type 6 have been regarded as adjunct operators that are base-generated in the specifier position of CP when they modify that IP (Rizzi, 1990). It is to be noted that the topic NP (i.e., /'alla:h/) in sentence 18 and its tree is not part of the focused element, and this is probably why it has undergone a "surface effect" rule, moving it from its base-generated preIP position to a post-IP position. That the topic NP (i.e., the noun /'alla:h/) has been moved from a pre-IP position is indicated by the fact that it is assigned an external thematic role and nominative case, rather than accusative case by the verb adjacent to it. Being a topic NP with no overt complementizer, it has been regarded as the multiple specifier of the I head, along with the subject small pro. Moreover, the lack of coindexation between the topic NP and the focused element in sentence 18 indicates that they do not constitute one unit, which is the focus in this sentence.6

Conclusion In this study, it has been assumed that the proverbial declarative sentence in EGA is syntactically distinct from the nonproverbial declarative sentence. The former declarative sentence requires an obligatory focused element in a pre-IP position as a rhetorical device. This pre-IP position has been regarded as a position in C because the focused element functions as a syntactic operator as regards the IP it heads. It follows that the nonproverbial declarative sentence that does not have an overt head with an overt complementizer in EGA may be regarded as having an IP structure, rather than a CP, because it has a weak feature in its C, rather than a strong feature. In other words, the nonproverbial declarative sentence

The Syntax of Colloquial Egyptian Proverbs

27

that does not have an overt head with an overt complementizer in ECA does not require an obligatory focused element in a pre-IP position as a rhetorical device. It is in this respect that we may say that there are two types of declarative sentences in ECA: one for the nonproverbial declarative sentence that does not have an overt head with an overt complementizer, and another for the proverbial declarative sentence. It is the former type of declarative sentence that supports Chomsky's (1995) belief that there is a null variant of the declarative C and that this null variant of the declarative C is introduced covertly because it has a weak feature. In other words, the notion that there are strong and weak features not only distinguishes between interrogative sentences and declarative sentences in English but also between the different types of declarative sentences in ECA. The focused elements in these proverbial declarative sentences have been divided into focused arguments and focused adjuncts. The focused arguments (be they CPs or NPs) are associated with internal positions within their major syntactic structures that determine their semantic interpretation. As for the focused adjuncts (be they PPs, IPs, or CPs), these are not associated with internal positions within their major syntactic structures for the determination of their semantic interpretations. Accordingly, the focused arguments are base-generated in an A-position, while the focused adjuncts are base-generated in an A'-position. But both positions (i.e., the A-position for focused arguments and the A'-position for focused adjuncts) are external to the major syntactic structure, i.e., in C. This not only highlights the importance of this pre-IP position in the syntactic configuration of these proverbial declarative sentences in ECA, but also the distinction between arguments and adjuncts. The difference in the syntactic behavior between the focused NP and the topic NP also warrants the assumption that they are base-generated in different A-positions in the pre-IP position. Accordingly, the topic NP without an overt complementizer has been regarded as a multiple specifier of the head I or T, allowing it to be extraposed (i.e., it has the ability to subsume to rules at the phonological component, leading to its extraposition from its base-generated position, and in turn display some flexibility in the word order of these proverbial declarative sentences). This distinctness of the focused NP and the topic NP is also indicated by the fact that each type of NP is assigned a different case, providing further evidence that they must be base-generated in two different A-positions. Being the specifier of C, the focused NP is assigned genitive case, rather than nominative case, and this is overtly manifested when there is an overt case marker carried by the focused NP. As for the topic NP that does not have an overt complementizer, it is assigned the nominative case

28

The Syntax of Colloquial Egyptian Proverbs

because it is the multiple specifier of I or T of Chomsky (1995). Thus, it is only when the declarative C is strong that it has the capacity to assign a distinct case to the NP it holds a spec-head relation with, i.e., it assigns genitive case to the focused element and not nominative case. The lexical type of analysis assumed for these focused elements is in keeping with Chomsky's (1995) assumption that 9-role assignment is the property of the base. It has enabled us to differentiate between the domain in which the focused argument is assigned a 0-role and that in which the topic NP is assigned a 0-role. Being in a spec-head relation with its C in its base-generated position, the focused argument be it a CP or an NP is assigned a 0-role by its head. Likewise, the topic NP without an overt complementizer is assigned a 0-role by its T because it is base-generated as its multiple specifier. Finally, the lexical type of analysis has been maintained in this study of the proverbial declarative sentences because the word order of these focused elements is not really free: these proverbial declarative sentences do not involve strictly optional movement operations since the focused element must be in pre-IP position, unlike the topic NP.

Notes 1 As adverbials of time such as /ba9dima/) incorporate a relative pronominal (indicated by the boldface part of this temporal), it seems more exacting to refer to them as "temporals" rather than as simply adverbs. For a more detailed discussion of the temporals in one of the Arabic dialects, see Ghaly (1988). 2 Shigeru Miyagawa (1997) has stated that binding can take place only in an A-position. 3 Shigeru Miyagawa (1997) states in accordance with Chomsky (1995) that specifiers count as multiple specifiers if and only if elements in these specifiers are checked by the same head. 4 Sentence 8 is a nominal sentence, which has been defined as sentence with a nonverbal predicate. This includes sentences with nominal predicates, with adjectival predicates, or prepositional predicates. See Ghaly (1988 ) for a discussion of the nominal sentences in one of the dialects of Arabic. 5 The feminine form of the third person, singular pronoun is used here because the use of this proverbial sentence is found mostly in female speech. But this is a separate study that would be interesting to pursue. 6 Cf. sentence 11 in which both the topic NP (i.e., /l-Hidda:ya/) and the focused element (i.e., the strong personal pronominal, /hiyya/) constitute one unit that is the focus in that sentence.

The Syntax of Colloquial Egyptian Proverbs

29

References Authier, J.-M. 1992. Iterated CPs and embedded topicalization. Linguistic inquiry, 23:329-336. Badawi, al-Said Muhammad. 1973. Mustawayat al-'arabiya al-mu'asira fi-misr. Cairo: Dar al-Ma'arif. Badawi, S. M., and Hinds, M. 1986. A dictionary of Egyptian Arabic: Arabic-English. Beirut: Libraire du Liban. Baker, M. 1988. Incorporation: a theory of grammatical function changing. Chicago: University of Chicago Press. Baker, M. 1995. The polysynthesis parameter. Oxford: Oxford University Press. Burton, S., and Grimshaw, J. 1992. Coordination and VP-internal subjects. Linguistic inquiry, 23:305-313. Chomsky, N. 1971. Deep structure, surface structure and semantic interpretation. In D. D. Steinberg and L. A. Jakobovitz (eds.), Semantics: an interdisciplinary reader, Cambridge: Cambridge University Press, pp. 183-216. . 1982a. Lectures on government and binding: the Pisa lectures. Dordrecht, Netherlands: Foris Publications. . 1982b. Some concepts and consequences of the theory of government and binding, Cambridge, MA: MIT. Press. . 1986a. Knowledge of language: its nature, origin, and its use. New York: Praeger. . 1986b. Barriers. Cambridge, MA: MIT. Press. . 1989. Some notes on economy of derivation and representation. In I. Laka and A. Mahajan (eds.), Functional heads and clause structure: MIT. working papers in linguistics, 10:43-75. . 1995. The minimalist program. Cambridge, MA: MIT. Press. and H. Lasnik. 1977. Filters and control. Linguistic inquiry,

8:425-504. Cinque, G. 1990. Types of A'-dependencies. Cambridge, MA: MIT Press. Cowan, D. 1982. Modern literary Arabic. Cambridge: Cambridge University Press,. Doherty, C. 1997. Clauses without complementizers: finite IP-complementation in English. The linguistic review, 14:197-220. Emonds, J. 1980. Word order in generative grammar. Journal of linguistic research, 1:33-54. Frascarelli, M. 1997. The phonology of focus and topic in Italian. The linguistic review, 14:221-248. Greenberg, J.1963. Some universals of grammar with particular reference

30

The Syntax of Colloquial Egyptian Proverbs

to the order of meaningful elements. In J. Greenberg (ed.), Universals of language. Cambridge, MA: MIT Press. May, R. 1985. Logical Form. Cambridge, MA: MIT Press. Miyagawa, S. 1997. Against optional scrambling. Linguistic inquiry,, 28:1-25. Pollock, J.-Y. 1989. Verb movement, universal grammar, and the structure of IP. Linguistic inquiry, 20:365-424. Rizzi, L. 1990. Relativized minimality. Cambridge, MA: MIT Press. Rochemont, M. S.1989. Topic islands and the subjacency parameters. Canadian journal of linguistics, 34:145-170. Takahashi, D. 1997. Move-F and null operator movement. The linguistic review, 14:181-196.

Understanding the Quran in English: Notes on Translation, Form, and Prophetic Typology Devin J. Stewart All too often the traditional dictum that the Quran cannot be translated, in recognition of its inimitable eloquence and doctrinal status as God's eternal speech and the primary miracle of the Prophet Muhammad's mission, is not attended by any explanation why this might be the case, or what elements one might miss when reading a translation. Such explanations would in fact help further an understanding of the Quranic text, yet the repeated moratorium not only hinders an informed awareness of Islam and its scripture among non-Muslims, but also runs the risk of alienating nonArab Muslims from their sacred text. Teaching the Quran in English translation to American students, both Muslims and non-Muslims, I have struggled with the problems associated with producing and reading an English rendition of the Quran. I have found, overall, that they reside less in difficult grammatical constructions or recherche vocabulary than in issues of form, genre, and rhetoric. The following remarks touch on some representative problems of translation and examine prophetic typology, a crucial rhetorical strategy in the Quran.

The "Genre" of the Quran and the Problem of Accurate Translation It is widely agreed that the Quran is a beautiful text. 'Umar 'ibn alKhattab, later the second Caliph, vehemently opposed the Prophet's early preaching in Mecca but was so moved upon hearing Td Ha (sura 20) recited that he converted on the spot. What is it that makes the Quran so beautiful and that renders any translation a pale shadow of the original? Rhyme and rhythm are certainly the most outstanding elements lost in translation. Doctrinal restrictions—the idea that the Quran is miraculous and therefore should not be likened to human literary artifacts—often discourage Muslims from saying this directly, but the Quran is a profoundly artistic

32

Understanding the Quran in English

and indeed poetic text. Comparison with poetry or with the statements of pre-Islamic soothsayers is explicitly denied in the Quran itself, and thus it is generally seen as heretical to call the Quran poetry or claim that it contains poetry.1 Nevertheless, a very large percentage (roughly 85%) of the verses in the Quran rhyme; a somewhat smaller proportion of the text exhibits rhyming, rhythmically parallel phrases, sometimes continuing throughout entire suras. Sura 55, The Beneficent, for example, contains 78 verses all rhyming in -an/-am that fall into groups of rhythmically parallel cola. The Arabic text of the opening verses shows the strong rhyme and rhythmical pattern, among the main reasons this sura is renowned as especially beautiful: al-rahman 'allama al-qur'an khalaqa al-insan 'allamahu al-bayan al-shamsu wa-al-qamaru bi-husban wa-al-najmu wa-al-shajaru yasjudan... (55:1-6)

This and other suras closely follow the compositional patterns of saj'—a type of writing in Arabic generally translated as "rhymed prose" or "rhymed and rhythmical prose." Though traditional Muslim exegetes have often, but by no means always, shied away from doing so, it is fair to label large sections of the Quran saj'.2 In recognition of this type of composition's poetic nature, one might even go so far as to define saj' as accent poetry, where word accents determine the number of feet or beats per line, distinguishing it from the quantitative poetry of the classical qasida tradition, where more strict combinations of short and long syllables make up each foot or beat.3 This and a number of other features having to do with verbal form are lost in translation, but nonetheless, the student of the Quran who knows no Arabic may develop a good understanding of many features of the Quran by concentrating on aspects that are less dependent on linguistic form. Many problems one faces in translating the Quran or in approaching it in translation have to do with expectations, and expectations are shaped largely by genre. The Quran is a sacred text and is approached on those terms not only by devout Muslims but also by non-Muslims and, perhaps surprisingly, even by atheists. One effect this has on translators is to make them use what they believe to be high, sacred, and often archaic language. Pickthall's translation—the translation I use for class and one of the best available, though a number of others are in print4— often uses an archaic English vaguely reminiscent of the King James

Understanding the Quran in English

33

translation of the Bible, with the archaic distinction between singular and plural second person pronouns (thou, thine, thee vs. you, your, ye) and verb forms (thou thinkest, etc.), the regular use of the preposition "unto" for "to," and other similar features. This serves to impart to the text a somewhat more holy ring, but at the same time removes it from ordinary, contemporary language and renders comprehension slightly more difficult for the average student. Translators of sacred texts tend to stick more closely to the original than translators of other types of composition. This phenomenon is of course not limited to the Quran but is clear in such texts as the Septuagint, the Latin Vulgate, and the genre of sharh—Judaeo-Arabic translations or renditions of Biblical texts. This tendency is so strong in some cases as to result in translations where every word in the original is represented by a corresponding word in the target language; the original syntax is reproduced in the target language at the expense of an idiomatic rendering. One need only peruse Wycliffe's translation of the Vulgate, for example, to see how awkward and how unpoetic the results of such methods may be. Pickthall's rendition of the Quran shows many examples of this types of translation. One glaring example is his translation of the Arabic particle 'inna, which he renders regularly "lo!" This is simply a bad translation, because "lo!" indicates surprise whereas 'inna indicates emphasis, even leaving aside the fact that "lo!" occurs very rarely in contemporary English and never in every third sentence. 'Inna generally lends a slight emphasis to the sentence it introduces: The question then becomes how to represent this emphasis idiomatically in English. In many cases, it seems best left untranslated. Lo! those who disbelieve, among the People of the Scripture and the idolaters, will abide in fire of hell. They are the worst of created beings. (And) lo! those who believe and do good works are the best of created beings. (98:6-7) This could better be translated as follows: »Those who disbelieve.... »Those who believe.... The latter translation is a more accurate rendering of the Arabic and adequately brings out the intended contrast between the two groups mentioned. In other contexts, one might use italics, exclamation points, "truly," "verily," or "indeed" to convey emphasis.

34

Understanding the Quran in English Lo! man is an ingrate unto his Lord (100:6) » Man is indeed ungrateful to his Lord! Lo! it is thy insulter (and not thou) who is without posterity. (108:3) » Your insulter is the one without progeny!

The particle wa- ("and") presents similar problems in translation because of differences between Arabic and English style. In Arabic, a series is listed as A and B and C, whereas idiomatic English requires A, B, and C. Pickthall and others, however, striving to stick as close to the original text as possible, usually preserve the extra "ands" in lists. ...[We] cause the grain to grow therein And grapes and green fodder And olive-trees and palm-trees And garden-closes of thick foliage And fruits and grasses: Provisions for you and your cattle. (80:27-32)

Only the last "and" here would appear in an idiomatic English rendition. In addition, wa- is often used in Arabic to begin a sentence, indicating that the previous topic has ended and marking the beginning of a new topic. This type of wa- in particular should be left untranslated, but Pickthall and other translators often leave it in. And how many a community revolted against the ordinance of its Lord and His messengers... (65:8) » How many a community...

This feature is so common in the Quran that in the second sura, for example, the longest in the Quran, 91 out of 286 verses, or just under one third, begin with "And" in Pickthall's translation. Many of these "ands" do not belong in an idiomatic rendering. The verb qala ("to say") presents similar problems, although again its meaning is apparent in a general way. While rendered regularly in translations as "say," the verb qala clearly means also "ask," "answer, respond, reply," or "command, order," depending on context, since it may be followed by a question, response, or imperative. Just as one might describe a conversational exchange in colloquial English using the verb "go" ("He goes..., so she goes...then he goes..."), the verb qala serves as an all-purpose speech introducer. Translating the verb with a more specific term according to context would improve the flow of the text and increase comprehension on the part of the reader.

Understanding the Quran in English

35

When he said unto his father and his folk: What is it that ye worship? (37:85) » When he asked his father and his people: What do you worship? And they say: When (will) this promise (be fulfilled), if ye are truthful? Say: The knowledge is with Allah only, and I am but a plain warner. (67:25-26) »They ask: When...? »Answer:... More difficult to render and more disconcerting to the average reader are the uses of the command qul ("say") to introduce various passages in the Quran that do not represent a human conversational exchange but rather the transmission of revelation to the Prophet. This occurs, among many other passages, in the last three suras in the Quran (112-114), which are clearly meant to be repeated or recited as prayers. The command "say" seems to operate here as an equivalent of quotation marks, setting off a particularly important passage the Prophet has been commanded to repeat. In order to convey this idea one might add in parentheses "Say (O Muhammad)" as Pickthall does on occasion elsewhere in his translation, but not in these suras. The functional sense behind the term, however, might be represented by stating, "Repeat (after Me):...." or "Recite (the following):...."

Form and Content If a less slavish translation of elements like those discussed above helps the reader of the Quranic text in English understand the relationship of sentences to one another within a passage or follow the flow of the text more easily, knowledge of the context of entire passages often proves crucial for an understanding of the text. Two elements are especially relevant here: immediate context of the revelation and the genre to which the passage belongs. Context has been treated extensively in the Islamic tradition, often in general exegeses but more particularly in works designated asbab al-nuzul ("the occasions of revelation"), the most famous of which are those of al-Wahidi al-Nisaburi (d. 468 AH/1075 CE) and Jalal al-Din al-Suyuti (d. 911 AH/1505 CE). Translations often present brief notes at the beginning of suras that explain something about the original context, drawing on material from the asbab alnuzul. The amount and specificity of the information provided varies and often leaves something to be desired. Pickthall explains in the short intro-

36

Understanding the Quran in English

duction to sura 108, Abundance, "The disbelievers used to taunt the Prophet with the fact that he had no son, and therefore none to uphold his religion after him." The introduction to sura 111, Palm Fibre, reads in part, It is the only passage in the whole Qur'an where an opponent of the Prophet is denounced by name. Abu Lahab (the Father of Flame), whose real name was Abdul 'Uzza, was a first cousin of the Prophet's grandfather and was the only member of his own clan who bitterly opposed the Prophet. He made it his business to torment the Prophet, and his wife took a pleasure in carrying thorn bushes and strewing them in the sand where she knew that the Prophet was sure to walk barefooted.

Both of these statements provide some vital information for an understanding of the suras in question, yet they could both be improved by a more close reliance on the material found in asbab al-nuzul. The other problem is one of genre, which has figured much less prominently in traditional exegeses. Translators generally fail to explain that both suras are essentially retorts, answers to specific insults directed at the Prophet that use linguistic forms based on the original insult. Recognition of the genre of the suras helps one understand the text better. Concerning sura 111, al-Wahidi reports: The Apostle of God ascended one day upon al-Safa and called out, "Oh ill-fated morning!" so the Quraysh gathered around him. They asked him, "What's the matter?" He asked, "Do you not see, were I to inform you that the enemy was about to attack you in the morning or in the evening, you would believe me?" They answered "Yes, of course." Then he said, "Then I am warning you of a painful torment I see approaching." Abu Lahab said, "Perdition to you (tabban laka)! Did you gather us together merely for this!"5

Given this background, it makes sense to understand sura 111 as a retort to the curse tabban laka, a curse responding directly to another curse. For this reason Pickthall's translation of the first verse of Palm Fibre, although it follows an interpretation commonly found in the traditional commentaries, is almost certainly wrong: "The power of Abu Lahab will perish, and he will perish" (111:1). Pickthall renders the verb here, in the perfect in the Arabic text, as future tense. This in itself is not surprising, for the perfect indicates future events in other passages of the Quran,

Understanding the Quran in English

37

but here, the verb is a cognate retort to a curse and thus should probably be rendered as a curse also, with the Arabic perfect serving as an optative: "May the hands of Abu Lahab perish, and may he (himself) perish!" With regard to sura 108, al-Wahidi reports, It was revealed concerning al-'As. He saw the Apostle of God coming out of the mosque when he was going in, and they met at the gate of the Banu Sahm clan and spoke. Some of the notables of Quraysh were inside the mosque, sitting. When al-'As entered, they asked him, "To whom were you speaking?" He replied, "That cut-off man (alabtar)" meaning the Prophet.6

Al-Wahidi provides another account: Al-'As ibn Wa'il al-Sahmi, whenever the Apostle of God was mentioned, would say, "Let him be. He is only a cut-off man (abtar) who has no progeny. If he were to die, mention of him would come to an end, and you would be rid of him."7

Al-Suyuti gives yet another version: "When the son of the Prophet died, al-'As ibn Wa'il said, 'Muhammad has become cut off (abtar),'' and the sura was revealed."8 Despite the differences in the versions, it is clear that the sura responds to a specific insult directed against the Prophet by al-'As ibn Wa'il al-Sahmi. The insult was the term abtar, meaning "devoid of progeny." In context, the insult was quite a grave one, not only because the Prophet had no sons except one, Ibrahim, who died very young, but also because of the enormous importance attached to sons in Arab culture. There was a prevalent idea that having sons was the only, or at least the surest means to carrying on one's legacy, so that mention of one's name— and with it praises of one's virtues and great deeds—should not die out. Knowledge of this background leads to a better understanding of sura 108, which throws the same insult back at the man who uttered it. We have granted you abundance, So pray and sacrifice to your Lord. Your insulter is the one with no progeny! (108:1-3) [my translation ]

The sura is meant to console the Prophet by turning the table on al-'As, arguing that although he will have no progeny, the Prophet will have abundant progeny. Given the nature of the insult and the fact that the sura as a

38

Understanding the Quran in English

whole is a retort, kawthar ("abundance") must be interpreted as "abundant progeny" in particular. It seems most plausible that this progeny is the Muslims or the believers, who are in a sense tantamount to the Prophet's family. Arguments found elsewhere in the Quran corroborate this interpretation, such as the statement that Noah's son, who drowns in the flood, is not actually part of Noah's family since he is a disbeliever (11:45-46), or the claim that the Jews and Christians are not true descendants of Abraham because they do not follow his religious legacy correctly (16:117-123; 57:26-29). There too, religious affiliation stands in place of blood relations. Other formal and rhetorical features act as a much greater stumbling block in the way of the average reader trying to understand the Quran in translation. When approaching the Quran, the English speaker's expectations are most likely to be shaped by knowledge—however limited—of the Bible. In a way this is helpful, because the Quran deals with a great deal of Biblical material. It may nonetheless lead to confusion with regard to form. A large portion of the Hebrew Bible, including the books Genesis through Ezra and Nehemiah, is presented chronologically, and this includes the parts of the Bible read most often. God creates the world in the opening verses of the first book, and the narrative flows, occasional interruptions aside, through antediluvian history, the age of the patriarchs, the invasion of Canaan, the period of the kingdoms up to the Babylonian captivity, and, taking up the thread again after the return from Babylon, postexilic history until ca. 400 BCE. Large parts of the text are couched in plain historical narrative, concentrating on the presentation of stories and events and making little or no commentary outside the narrative itself. The Quran, however, does not proceed chronologically. Its 114 suras or chapters are not presented in historical order, either in terms of the stories they contain or in terms of their revelation. Rather, with the exception of the short opening sura, they are presented roughly in descending order of length, an order that seems to be nearly completely arbitrary with regard to content. Moreover, individual suras often include sections describing various periods, such as sura 27, The Ant, which includes sections on Moses, Solomon, the pre-Islamic Arabian prophet Salih, and Lot. Often such material is presented in the order of Biblical history—or pseudoBiblical history, including the prophets Hud and Salih, for example—as in sura 54, The Moon, which presents sections on Noah, Hud, Salih, Lot, and Pharaoh. Elsewhere it does not keep to this order: Lot obviously comes before Moses and Solomon historically, though his story is presented after theirs in sura 27. Very few suras are devoted to a single historical narrative, the main exceptions being sura 12, Joseph, and sura 71, Noah. In addition, many of the suras which contain this type of material belonging

Understanding the Quran in English

39

to sacred history also include additional material, outside of the historical framework, referring to the present, the recent past, or the future. This leaves out the many suras and sections of suras that are devoid of narratives of sacred or prophetic history and focus instead on legal topics, exhortation, apocalyptic predictions, descriptions of heaven and hell, descriptions of the natural world, and so on. All this variety, over and above the lack of adherence to chronological order, makes the Quran a difficult and confusing book to read for an audience conditioned by expectations to look at the Quran through the lens of the Hebrew Bible or Old Testament, that is, as a relatively continuous narrative of sacred history. It is not surprising that one of the suras most often read by Western scholars and most widely used in Arabic chrestomathies is Joseph, despite the fact that it is rather atypical, precisely because it fits these expectations more closely than any other sura of the Quran. Another expectation that Christians or readers in predominantly Christian societies may entertain when approaching the text of the Quran derives from their knowledge—again, however superficial—of the New Testament. It is altogether reasonable for a Christian reader to expect that just as the Gospels narrate the life and miracles of Jesus, the Quran should narrate the life and miracles of the Prophet Muhammad. This goes along with the mistaken notion that Muhammad's status in Islam is equivalent to that of Jesus in Christianity, an idea responsible for the term Mohammedanism, used formerly to denote Islam, though it has since fallen out of usage.9 The reader of the Quran is struck by the fact that Muhammad appears by name very seldom, only four times in the entire sacred text (3:144; 33:40; 47:2; 48:49). Instead, the protagonist of the Quran, the character mentioned far and away most often, is Moses, whose name appears 136 times. The Quran does not describe major sections of Muhammad's life and works in plain narratives, although it does include many references to events in the history of the prophetic mission and the early Muslim community. One example is the description of the apparitions of Gabriel to the Prophet (53:5-18), but it is worth noting that this passage is neither embedded in a general narrative of the Prophet's life nor situated temporally as having occurred before or after other specific events. The overall effect is that the reader, whether expecting an Islamic counterpart to the Gospels or even conditioned by familiarity with the form of the modern novel, is surprised not to find an extensive or connected account of the Prophet Muhammad's mission.

Prophetic Typology in the Quran Given the fact that the Quran fails to conform to many of the English reader's expectations, thereby engendering cognitive dissonance, what can one

40

Understanding the Quran in English

do to ease his or her introduction to this difficult text? One approach that has proved successful in classes and enabled students to grasp a significant portion of the texts included in the Quran is to explain what I believe to be one of the Quran's most important rhetorical strategies: the use of the pattern of Biblical prophecies in order to comment on or serve as a model for the prophecy of Muhammad. This strategy is in fact a sort of typology, familiar to analysts of the New Testament, whereby characters from the Hebrew Bible are taken as models for or precursors of Christ or are used to make specific arguments concerning the nature of his life and works. An understanding of this rhetorical strategy helps explain the form and content of many suras of the Quran, particularly those containing series of stories of earlier prophets. Essential for a grasp of this rhetorical strategy is the basic understanding that the prophecy of Muhammad was a Biblical prophecy. It was Biblical not in the sense that it appears in the Bible, but in the sense that Biblical history was the fundamental framework within which the Islamic revelation unfolded and according to which the Prophet interpreted events and undertook his mission. As mentioned above, Moses is the main character in the Quran. This is so for a simple reason: that he serves as a model, indeed, apparently the main model, for the Prophet Muhammad. While scholars have long recognized the important connections between the Quran and the Bible, they have not sufficiently stressed the point that Muhammad's prophecy was formulated in Biblical terms, instead showing that elements were borrowed from Judaism and Christianity or comparing Quranic material with Biblical accounts.10 Furthermore, translations of the Quran often fail to recognize this Biblical framework sufficiently. In order to render it transparent, translations of key terms in the Quran and early Islamic history should reflect the Biblical connections that the original terms were based on or meant to bring out. The most obvious of these terms is, of course, "God." Pickthall and others leave Arabic "Allah" untranslated. Despite the fact that Allah was one of many gods in the pre-Islamic pantheon, Allah in the Quran and in Islam in general is the Biblical God, the same God who delivered the Hebrews from slavery in Egypt and gave wisdom and prosperity to King Solomon. It makes sense, therefore, to render "Allah" regularly as "God" in the English translation. To retain Allah is like retaining Latin Deus or Greek theos in English translations of the Bible, for Allah is simply Arabic for God, and, indeed, is used as such by Jewish and Christian speakers of Arabic as well as Muslims. The term ahl al-kitab, used many times in the Quran in reference to Jews, Christians, or both, is usually rendered in English as "the people of the Book" or "the people of the Scripture." It would be more fitting to translate the term as "the people of the Bible," in order to stress the point that the book in question here is a specific one and not just any member of the category "scripture." This term glosses over the fact that Jews and

Understanding the Quran in English

41

Christians have different ideas about what constitutes the Bible, but nonetheless serves to refer to it as a recognizable unit. The hijra, the Prophet's "emigration" or "flight" from Mecca to Medina, should be rendered "Exodus," since, in all probability, it reflects a comparison between Muhammad and Moses, between the early Muslims and the Hebrews, between the tyrannical chiefs of Quraysh and Pharaoh. The term ansar, "helpers," designating the Medinan converts to Islam, is based on an analogy with the disciples of Jesus. This analogy, unlike that of the hijra, is made explicitly in the Quran itself: But when Jesus became conscious of their disbelief, he cried: Who will be my helpers in the cause of Allah? The disciples said: We will be Allah's helpers. We believe in Allah, and bear thou witness that we have surrendered [are Muslims] (unto Him). (3:52) The term ansar, plural of nasir ("helper," " ally"), intentionally puns here on nasara "Christians," which derives from al-Nasira, "Nazareth." These two terms, hijra and ansar, crucial for the history of the early Muslim community and of great importance in the Quran, are only two among many indications of the centrality of Biblical models in Muhammad's prophecy. Elements of pre-Islamic Arabian religious tradition that were incorporated into Islam and appear in the Quran are legitimated by being presented in Biblical guise. The pre-Islamic shrine of the Ka'ba is thus reinterpreted as a Biblical temple built by Abraham. For this reason, the term bayt ("house"), which appears in several passages in reference to the Ka'ba, should instead be translated "temple" (e.g., 106:3). The pre-Islamic Arabian prophets Hud and Salih are likened in the Quran to Biblical prophets. Furthermore, the pagan enemies of the Biblical prophets are conflated with the pagan Arabs. Noah's opponents, for example, worship preIslamic Arabian gods: "And they have said: Forsake not your gods. Forsake not Wadd, nor Suwa', nor Yaghuth and Ya'uq and Nasr" (71:23). In sum, Biblical prophecy is a controlling idea or theme in the Quran, one which shapes the incorporation of various religious concepts and elements into the Quranic text and Islam in general. An awareness of this Biblical framework is fundamental for an understanding of many aspects of the Quran. Biblical prophecy in the Quran involves two main ideas. First is the recognition that the careers of prophets are similar to one another in their major, important features; by and large they follow the same pattern, though they might differ in detail. The Quran provides many examples of these patterned prophetic careers, presenting a number of Biblical characters such as Noah, Abraham, and Lot as prophets even though they are not

42

Understanding the Quran in English

termed prophets or portrayed as such in Genesis. For the purposes of typology, it matters little whether these figures are termed "prophet" (nabi), "messenger, apostle" (rasul, mursaf), "warner" (nadhir, mundhir), or "bearer of glad tidings" (bashir, mubashshir) in the Quranic text; they all perform essentially the same functions and the terms describing them appear in nearly identical contexts. Second is that the prophecy of Muhammad follows the pattern evident in earlier prophetic missions. On occasion, the comparison between Muhammad's prophecy and that of earlier prophets is explicit, while in other instances it is merely understood. The comparison is so pervasive, though, that it should always be present in the mind of the reader when interpreting Quranic accounts of Biblical figures. Sura 66, The Banning, provides a clear instance of such a comparison. Here, the text scolds the Prophet's wives for some transgression on their part involving a jealous plot or breach of confidence.11 The wives of Noah and Lot are held up as examples of bad women who suffered damnation for their evil behavior despite their close connection with men of God. The marriage bond was not enough to save them from punishment. The comparison and its implications are evident. Muhammad, as a prophet, is analogous to Noah and Lot; his wives are analogous to their wives. The punishment in Biblical history implies that, should they persist in their misbehavior, Muhammad's wives will not escape divine punishment, despite their special connection with the Prophet himself. In other passages, the comparison is perhaps less obvious, but nevertheless crucial. An analysis of sura 54, The Moon, an early Meccan sura, demonstrates how this rhetorical strategy works in many suras of the Quran. The Moon is of medium length as far as Quranic suras go: it is composed of 55 verses, all of which rhyme in -ar/-ir/-ur. As in many of the suras that contain stories about prophets, a series of accounts of various prophets occurs in the middle section of the sura (verses 9-42), sandwiched between an introductory section (1-8) and a final section (43-55), which stand outside the historical narrative. The middle section contains five subsections, each dealing with one prophet and his audience: Noah and his people, Hud and the tribe of 'Ad, Salih and the tribe of Thamud, Lot and his people, and Moses and the house of Pharaoh. The organization of the sura may be represented as follows: Introduction (1-8) Noah and his people (9-17) Hud and 'Ad (18-22) Salih and Thamud (23-32) Lot and his people (33—40) Moses and Pharaoh (41—42) Epilogue (43-51)

Understanding the Quran in English

43

Deeds recorded in Scripture (52-53) Promise of heaven (54-55) Perhaps the most striking feature of this sura is the strong parallelism between the individual prophet subsections. These passages do not present exhaustive narratives of the prophecies, but rather highly compact and stylized summaries of their events. Their parallelism is emphasized by the repetition of specific phrases. The five subsections all begin with a statement that a people of the past, the audience of a particular prophet, denied or rejected (Ar. kadhdhabai) the prophet or the warnings he conveyed. The folk of Noah denied... (9) (The tribe of) A'ad rejected warnings... (18) (The tribe of) Thamud rejected warnings (23) The folk of Lot rejected warnings (33) And warnings came in truth unto the house of Pharaoh Who denied Our revelations, every one... (41-2) The subsections end with two successive statements that serve as comments outside the narrative proper, presenting the moral of the story, as it were; the first of these stresses God's punishments of the earlier peoples. Then see how (dreadful) was My punishment after My warnings! (16) Then see how (dreadful) was My punishment after My warnings! (21) Then see how (dreadful) was My punishment after My warnings! (30) ...Taste now My punishment after My warnings! (37) Now taste My punishment after My warnings! (39) The subsections end with a reference to the Quran itself, stressing that these punishment stories are meant to serve as instructive examples for posterity: And verily We left it as a token; but is there any that remembereth? (15) And in truth We have made the Qur'an easy to remember; but is there any that remembereth? (17, 22, 32, 40) The key terms in these last verses are the words translated as "remember" and derived from the Arabic root dh-k-r: wa-la-qad yassarna alqur'ana li-dhikri fa-hal min muddakir, which Pickthall translates, "And in

44

Understanding the Quran in English

truth We have made the Qur'an easy to remember; but is there any that remembereth?" These statements are in fact puns of a sort, the first clause of the English translation (in verses 17, 22, 32, 40, etc.) apparently meaning that God has made the Quran easy to memorize, but the second clause asking whether anyone will take heed or learn from the punishments inflicted on past peoples. These three parallel phrases, framing each of the subsections, stress the close parallelism between the prophecy accounts themselves. Examination of the accounts in parallel shows that the prophecies all follow a predictable pattern. The following general steps emerge as belonging to the generic prophetic pattern: 1. God selects a prophet (implied) 2. The prophet addresses his people (implied) 3. The prophet warns his people of God's wrath 4. They reject the warnings 5. God annihilates the rejecters, saving only prophet and believers The middle sections of the sura therefore serve mainly to present what may be called punishment stories. Prophets have been sent to various communities throughout sacred history. The prophets warn them of God's impending punishment should they not heed the messages that the prophet relays. They reject these warnings and are punished accordingly, annihilated by God through some cataclysmic event: Noah's people by the flood, 'Ad by a raging wind, Thamud by a tremendous shout (= earthquake? eruption?), and Lot's people by a deluge of stones. The punishment of Pharaoh and his people is described in vague terms: "We grasped them with the grasp of the Mighty, the Powerful" (v. 42). Prophecy follows the predictable steps listed above, and all prophetic careers follow this same pattern. The narratives are present not as mere histories but as didactic examples to serve as a warning to a contemporary audience, that of the Prophet Muhammad. Both the introductory and the final sections are set in the present and relate the punishment stories to the contemporary situation. The introduction shows parallelism with the prophet narratives, implying that the steps occurring in the earlier punishment stories also occur in the present career of the Prophet. The sura opens with a miraculous sign, the splitting of the moon: "The hour drew nigh and the moon was rent in twain." It is referred to as a sign or "portent" (ayd) in the next verse, and is parallel to the "warnings" (nudhuf) mentioned in the punishment stories. Indeed, there is a reference to "warnings" in verse 5. "They," apparently the unbelievers in the Prophet Muhammad's audience, then deny the sign. Muhammad him-

Understanding the Quran in English

45

self is not mentioned directly, but is commanded in the imperative in verse 6: "So withdraw from them (O Muhammad)...." It is clear that "they" in verses 2-7 are contemporary with him, but they are not identified as the disbelievers (of Quraysh) until verse 8: "Hastening toward the Summoner; the disbelievers say: This is a hard day." Thus far, the introductory section follows the prophetic pattern: 1. God selects a prophet, (implied) 2. The prophet addresses his people, (implied) 3. The prophet warns his people of God's wrath. 4. They reject the warnings.

The problem comes with the last step, God's punishment and annihilation of the disbelievers through a cataclysmic event that, in the case of the Prophet Muhammad's mission, obviously cannot have taken place yet. The text aims to resolve this discrepancy, preserving the tight parallelism of prophetic careers despite the fact that the Prophet Muhammad's career is still in progress and the annihilation of his enemies has not yet occurred. It does so by replacing the actual historical punishments of the past with a description of a future punishment awaiting contemporary disbelievers. So withdraw from them (O Muhammad) on the day when the Summoner summoneth them unto a painful thing. With downcast eyes, they came forth from the graves as they were locusts spread abroad, Hastening toward the Summoner; the disbelievers say: This is a hard day. (54:6-8)

This punishment is not a typical annihilation but rather part of the apocalyptic events of the Resurrection and the Day of Judgment. The punishments meted out to earlier peoples who denied their prophets' warnings are replaced, for the Prophet Muhammad's deniers, with the threat of Judgment. Similarly, the Prophet's withdrawal stands in place of the earlier prophets' actual escape from annihilation. A number of elements within the punishment stories also call attention to the Prophet Muhammad. In verse 9, Noah is called a madman by his audience. The same accusation was, of course, directed at Muhammad (52:29, 68:2, 81:22, etc.). The tribe of Thamud balked at following Salih on the grounds that he was a mere mortal like themselves (v. 24). Muhammad met with similar remonstrance from Quraysh (21:3). Thamud called Salih a liar (v. 25), just as the Quraysh taunted Muhammad (38:4).

46

Understanding the Quran in English

That the earlier stories are linked to the contemporary situation is made explicit in verse 9, the beginning of Noah's story, "The folk of Noah denied before them..." The pronoun "them" here refers to the disbelievers of Muhammad's time who appear in verses 2-7 and are defined in verse 8, just preceding this verse. Verse 9 shows that the punishment stories which follow are intended as commentaries on the present situation and didactic examples. Likewise, the epilogue section reverts to the present time, comparing between the earlier peoples and Muhammad's contemporary audience: "Are your disbelievers better than those...?" (v. 43). The demonstrative pronoun "those" here denotes the unbelievers annihilated in the punishment stories presented in verses 9-42, while the possessive adjective "your" refers to Muhammad's audience: the tribe of Quraysh. This implied threat of annihilation is developed in the subsequent verses, again replacing an actual destruction with images of Judgment Day: On the day when they are dragged into the Fire upon their faces (it is said unto them): Feel the touch of hell. (54:48)

The epilogue section (verses 43-51) also parallels the punishment stories of the middle section, a fact emphasized by the repetition of the closing phrase in verse 51: "And verily We have destroyed your fellows; but is there any that remembereth?" The phrase "your fellows" refers back to the earlier destroyed people, establishing the comparison between them and contemporary disbelievers explicitly yet again. The phrase "but is there any that remembereth" of course recalls the same phrase that occurred in verses 15, 17, 22, 32, and 40. The fundamental rhetorical strategy of The Moon is typological. It comments on the situation the Prophet Muhammad faces using characters from sacred history who are analogous to him and in whose place he stands in effect. Earlier prophets preached to their peoples, and their peoples denied them. The prophets then warned them, yet they persisted in ignoring these warnings until God fulfilled his threat and punished them, inflicting destruction upon them but saving the prophets and the small groups of believers. The implication for the Quraysh is that they have been warned by Muhammad of God's impending wrath, yet they continue to ignore these warnings even when they are as obvious as the splitting of the moon in the sky. Alternatively, one might interpret the sign of the split moon as referring not to the past but to a future event. The meaning of the text might be that even were this miraculous event to occur, the disbelievers of Quraysh would still insist on their stubborn resistance to God's messages. In either case, Muhammad's audience is supposed to understand

Understanding the Quran in English

47

that continued denial will lead to their perdition, just as it led to the perdition of earlier recalcitrant nations. Rather than threatening punishment in this world, the sura stresses the impending punishment of Muhammad's opponents on Judgment Day: they will taste the torments of hellfire. The threat is made all the more urgent through repeated mention of annihilation and emphasis on the stories' unassailable, authoritative source: scripture. This explains the mention of the Quran itself in verses 17, 22, 32, and 40 and the mention of the Scripture in verses 52-53, just before the end of the sura. And every thing they did is in the scriptures, And every small and great thing is recorded. (54: 52-53) The sura closes with mention of the abode of the righteous in the gardens of heaven (verses 54-55). This type of ending, found elsewhere in the Quran, leaves the audience with a ray of hope. Should they change their ways, they may yet avoid the terrible doom that the rest of the sura portrays so vividly. Analysis of The Moon shows that while the Quran does not narrate the life and works of Muhammad directly, the way the Gospels narrate the life and works of Jesus, it nevertheless speaks of his prophetic mission using typology, a rhetorical strategy based on model and analogy. Characters of Biblical or sacred history are defined as prophets whose careers follow a predictable pattern, reflecting the fact that they have all been sent by the one God and therefore represent the same boss, so to speak. The prophets' messages are essentially the same because God's truth is eternal, and the events of their prophetic missions are the same because neither human nature nor God's customary manner of dealing with humanity (sunnat Allah) has changed. Since Muhammad is indeed a prophet, his prophecy and his contemporaries' reactions to it can be explained, interpreted, and even to some extent predicted by analogy, using the stories of earlier prophets as a basis for comparison. This strategy is by no means limited to The Moon, but is rather quite prominent in the Quran as a whole, occurring quite clearly in suras 7, 10, 11, 14, 15, 21, 23, 26, 29, 37, 38, 40, 41, 50, 51, 66, and 69, 89, and in somewhat more diffuse form in a number of other suras. Even in suras in which there is no explicit reference to the Prophet Muhammad, such as sura 91, The Sun, which presents a summary of the story of Salih and Thamud, accounts of former prophets are in all probability intended to be understood in the typological sense. Perhaps more than any other single mode, this rhetorical strategy shapes the discourse of the Quran. Teaching this type of analysis helps students grasp the ideas behind a great

48

Understanding the Quran in English

deal of the Quranic text, elucidating the ways in which much Biblical material is used in the Quran, defining the Quranic relationship between Judaism, Christianity, and Islam, and bringing into relief the Quran's portrayal of Muhammad's prophetic mission.

Notes 1 A classic statement of this position is given by Abu Bakr Muhammad ibn al-Tayyib al-Baqillani (d. 403 AH/1013 CE), I'jaz al-qur'an, edited by Muhammad 'Abd al-Mun'im al-Khafaji (Beirut: Dar al-Jil, 1991), pp. 103-109. 2 For a statement that the Quran does not contain saj' per se, see alBaqillani, I'jaz al-qur'an, pp. 110-119. 3 See Devin J. Stewart, "Saj' in the Qur'an: prosody and structure," Journal of Arabic literature, 21 (1990): 101-139. 4 Unless otherwise noted, all Quranic quotations in the following essay are from Muhammad Marmaduke Pickthall's translation, The meaning of the glorious Qur'an. Beirut: Dar al-kitab Allubnani, 1973. 5 Al-Wahidi, Asbab al-nuzul (Cairo: Matba'a Hindiya, 1316 AH), p. 344; al-Suyuti, Asbab al-nuzul (Cairo: Maktabat Nusayr, 1983), p. 308. 6 Al-Wahidi, Asbab al-nuzul, p. 343. 7 Al-Wahidi, Asbab al-nuzul, p. 343. 8 Al-Suyuti, Asbab al-nuzul, p. 306. 9 See, e.g., D. S. Margoliouth, Mohammedanism (London: Williams and Norgate, 1911) and H. A. R. Gibb, Mohammedanism (London: Oxford University Press, 1949). 10 See G. Geiger, Was hat Mohammed aus den Judentums aufgenommen? (Bonn: F. Baaden, 1835); H. Hirschfeld, Beitrdge zur Erklarung des Koran (Leipzig: O. Schulze, 1886); I. Schapiro, Die haggadischen Elemente im erzdhlenden Teile des Korans (Leipzig: G. Fock, 1907); W. Rudolph, Die Abhdngigkeit des Qorans von Judentum und Christentum (Stuttgart: Kohlhammer, 1922); J. Horovitz, Koranische Untersuchungen (Berlin: Gruyter, 1926); Ch. Torrey, The Jewish foundation of Islam (New York: Jewish Institute of Religion Press, 1933); Heinrich Speyer, Die Biblischen Erzdhlungen im Qoran (Hildesheim: Georg Olms, 1961); Jacques Jomier, The Bible and the Koran, trans. P. Arbez (New York: Desclee, 1964); M. S. Scale, Qur'an and Bible: studies in interpretation and dialogue (London: Croom Helm, 1978). A recent exception that takes a typological approach is David Marshall's God, Muhammad and the unbelievers: a Quranic study (Surrey, England: Curzon, 1999). 11 See al-Wahidi, Asbab al-nuzul, pp. 325-327.

Relativization in English and Arabic: A Bidirectional Study Nagwa Kassabgy and Mona Kamel Hassan Research has established the impact of a number of factors on foreign-language learning. Motivation, age, aptitude, exposure, attitude, and input are among those factors (Larsen-Freeman and Long, 1991). Another significant and at the same time controversial factor identified by research is the impact of cross-linguistic differences and the role of language transfer. According to Doughty and Williams (1998, p. 226), "a learner's previous linguistic knowledge influences the acquisition of a new language in a principled, if not straightforward, contrastive way." Syntactic transfer, in particular, has long been controversial; research has shown evidence of absence of transfer, of both positive and negative transfer, and of the interaction of transfer with other factors in acquisition (Odlin, 1990). There are several reasons for language teachers to consider the problem of transfer, the most significant of which is that teaching may be more effective when teachers are aware of differences between languages and between cultures. An English teacher aware of Arabic-based transfer errors and an Arabic teacher aware of English-based transfer errors will not only be able to pinpoint learners' problems better, but will also better understand what may be difficult or easy for that group of learners. One of the problematic structures for most foreign-language learners identified by both researchers and teachers is relative clauses. Considerable research has been done on the acquisition of relative clauses. Keenan and Conmrie's (1977) cross-linguistic survey of 50 languages indicates that there is considerable variation in relative clause structures, which may lead to language transfer. Characteristics of relative clause structure, specifically pronoun retention, pose an interesting problem for both teachers and researchers. Research on the acquisition of English and Swedish, languages that do not use resumptive pronouns, suggests that transfer plays a role in the use of these pronouns. Gass (1979, 1983) also found that the native language does influence pronoun-retention errors in relative clauses. Hyltenstam (1984) found that speakers of Greek and Farsi, which allow pronominal retention, produced many more resumptive

50

Relativization in English and Arabic

pronoun errors than did speakers of Finnish and Spanish, which do not allow pronoun retention. On the other hand, some researchers (loup and Kruse, 1977; Tarallo and Myhill, 1983; Birdsong, Johnson, and McMinn, 1984) provided evidence for universalist arguments that the erroneous use of resumptive pronouns does not indicate transfer, since learners of foreign languages that do not use resumptive pronouns often considered acceptable sentences that contained resumptive pronouns. A universalist explanation, therefore, seems plausible since all speakers of foreign languages—regardless of the use or nonuse of pronoun retention in their LI—produce or accept sentences with resumptive pronouns in their L2. Kharma (1987) investigated Arabic speakers' problems in the acquisition of English relative clauses. He identified 14 types of errors, concluding that most of the errors can be attributed to negative interference from Arabic. Another point Kharma made was that all the errors were errors of form rather than use and did not affect communication. Gass (1986, cited by Odlin, 1990) compared difficulties encountered by Italian speakers learning English with those of English speakers learning Italian. According to Odlin (1990), there is a need for more bidirectional research, as such comparisons could provide a better understanding of the general structural principles that affect transfer. To the knowledge of the authors of this paper, no bidirectional study that compares difficulties in the acquisition of relative clauses by Arabic speakers of English as a foreign language (EFL) and English speakers of Arabic as a foreign language (AFL) has so far been done. Comparing the difficulties encountered by both groups of learners could shed light on whether errors can be attributed to interference, to universalist explanations, or to a combination of factors.

Relative Clauses in English and Arabic Schachter (1974, cited in Larsen-Freeman and Long, 1991) discussed the work of Keenan and Comrie (1972) and identified three main dimensions in which relative clauses can differ and that may pose problems for L2 learners. The first dimension relates to the position of the relative clause with respect to the head noun. Relative clauses in English and Arabic— Egyptian colloquial Arabic (ECA) as well as modern standard Arabic (MSA)—follow the head noun, i.e., the syntactic pattern of word order of English and Arabic is similar. The second dimension relates to how relative clauses are marked. English uses a relative pronoun that agrees with the noun it replaces, i.e., who for subject-case human, whom for object-case human, which for non-

Relativization in English and Arabic

51

human, that for both human and nonhuman, and whose for human and nonhuman possessive determiner. ECA uses one relative pronoun (/illi/) for the different structures. MSA uses relative pronouns that need to agree with the head noun in case, i.e., nominative, genitive, or accusative, e.g., /allata:n/ vs. /allatayn/; gender, i.e., feminine or masculine, e.g., /alladhi/, /allati/, and number, i.e., singular, dual, or plural, e.g., /alladhi/, /alladha:n/, /alladhirn/. In other words, English and both EC]A and MSA differ in this dimension. The third dimension identified by Schachter relates to the presence or absence of a pronominal reflex. English does not allow pronoun retention as an object noun or as object of a preposition. Both ECA and MSA retain the object noun in the relative clause in a pronominal form and a pronominal reflex as object of a preposition, e.g.: /hiyya di il-mudarrisa illi shuftaha imba:riH/ (ECA) /hadhihi hiya al-mudarrisa allati ra'aytuha 'ams/ (MSA) (lit., "This is the teacher whom you [masc. sing.] saw her yesterday") /huwwa da il-kita:b illi iddaitu lu imba:riH/ (ECA) /hadha huwa al-kita:b alladhi 'a9Taytuhu lahu 'ams/ (MSA) (lit., "This is the book that I gave it to him yesterday")

Again, English differs from EGA and MSA in this dimension.

The Current Study Research Questions. The current study attempted to answer the following questions: • What errors in the acquisition of English relatives by Arabic speakers and Arabic relatives by English speakers may be attributed to L1 interference? • What errors may be explained as developmental or universal? Subjects. The study included 86 subjects, 39 of whom were native speakers of Arabic and 47 of whom were native speakers of English. Of the 39 Arabic speakers, who were all enrolled in intensive English as a foreign language (EFL) programs in the English Language Institute and in the Center for Adult and Continuing Education of the American University in Cairo (AUC), 13 subjects were at the beginner's level, 17 were at the inter-

52

Relativization in English and Arabic

mediate level, and nine were at the advanced level. Of the 47 Englishspeaking subjects, who were enrolled in intensive Arabic as a foreign language (AFL) programs in the Arabic Language Institute at AUC, for both ECA and MSA at three different levels, 18 were beginners, 17 were intermediate, and 12 were advanced. According to the AUC catalog, AFL beginners are students "who work to develop the fundamentals of language through reading drills, within a framework of the essentials of syntax and morphology." AFL intermediate students are students "who emphasize the acquisition of vocabulary and increase the command of grammatical and syntactical structures and further develop reading, writing, listening, and speaking skills." AFL advanced students are students "who, through the reading and analysis of selected texts, are exposed to a wide range of vocabulary, idiom, and style, while reviewing the major topics of grammar." Study instrument. The instrument designed for this study was a sentencecombining task in English'that contained 10 items. The same items were translated into Arabic (Please see Appendix 1 for the English version, Appendix 2 for the ECA version, and Appendix 3 for the MSA version). The students were instructed to combine the sentences, using the second sentence as an adjective/relative clause. The test was administered to the EFL students after they had received instruction and practice in the use of relative clauses in English and to the AFL students also after they had received instruction and practice in the use of relative clauses in both ECA and MSA. In other words, the study investigated relative clause production errors made by the same AFL learners in both ECA and MSA. The students were familiar with all the words, i.e., the vocabulary, used in the sentences. The study did not examine Keenan and Connie's accessibility hierarchy with regard to the ease or difficulty of the acquisition of relativization of various types of relatives. Rather, it sought to investigate production errors on the sentence-combining task, which included a variety of relative clause structures and one relative structure not included in the accessibility hierarchy, namely, using expressions of quantity in the relative clause as in sentences 9 and 10.

Data Analysis Following is an item-by-item presentation of the results. Tables 1-8 illustrate the type and number of errors made by both EFL and AFL learners on each item. Each table is followed by a discussion, giving examples of errors made by EFL and AFL students at the three levels of proficiency. Results were not converted into percentages because this was a small-scale error analysis study

Relativization in English and Arabic

53

that attempted to investigate and analyze errors in the production of relative clauses. Ungrammatical sentences are marked with an asterisk (*) and, whenever a specific error (based on the above list) was identified, it was underlined. Table 1: Comparative error analysis of Item 1. Type of error

AFL

EFL

MSA

ECA B

I

n=13 n=17 Cop

A n=9

B

I

n=18 n=17

B

A

n=12

I

n=18 n=17

A

n=12

2 1

WO

Cop = omission of copula; WO = word order. B = beginners; I = intermediate-level students; A = advanced students. EFL/AFL (Item 1) The student is from China. She sits next to me. For both EFL and AFL learners, this type of relative seems to be the easiest to acquire, which agrees with Keenan and Comrie's hierarchy. Table 2: Comparative error analysis of Item 2. AFL

Type of error

EFL B I n=13 n=17

RP

RPAg. N/Rel.pr. AgAv.

3

9

ECA A n=9

B n=18

3

14

I A n=17 n=12 2

MSA B I n=18 n=17 13

A n=12

1

1 1

3

RP = resumptive pronoun/pronoun retention; RPAg. = resumptive pronoun agreement (in ECA and MSA the pronoun is retained and agrees with the noun); N/Rel.pr. Ag. = noun-relative pronoun agreement (in MSA the relative pronoun agrees with the noun in case, gender, and number); Av. = avoidance (producing a correct sentence but avoiding the relative construction). B = beginners; I = intermediate-level students; A = advanced students

Relativization in English and Arabic

54

EFL (Item 2) I knew the woman. I met her at the party yesterday. The persistent error was retention of the pronoun (RP). This can indicate transfer from L1 since Arabic is a language that retains the pronoun. Other errors relate to the use of the wrong relative pronoun, i.e., which instead of whom, and errors of avoidance, i.e., avoiding the relative and producing a correct sentence: / met the woman at the party yesterday. The use of the wrong relative pronoun can also be attributed to Arabic since in ECA, only one relative pronoun (/illi/) is used for the different structures. AFL (Item 2) /'ana kunt 9a:rif is-sitt. 'ana 'abilt is-sitt fi-1-Hafla imba:riH./ (EGA) /kunt 'a9rif al-mar'a. qarbalt al-mar'a fi-l-Hafla 'ams./ (MSA) ("I knew the woman. I met the woman at the party yesterday.") Unlike the EFL learners, the RP error, which may be attributed to L1 transfer, improved across levels, which could mean it is developmental. The 14 beginners and two intermediate students produced: * /'ana kunt 9a:rif is-sitt illi 'abilt fi-1-Hafla imba:riH/ Similarly, in MSA, the error made by 13 beginners and one intermediate student was RP, which may indicate transfer only at the lower level of proficiency, i.e., a developmental error: * /'ana kunt 9a:rif al-mar'a allati qa:balt fi-1-Hafla 'ams/ Table 3: Gomparative error analysis of Item 3. Type of error

B

I

n=13 n=17

N/Rel.pr. Ag.

Av. WO

AFL

EFL

8

B A n=9 n=18 n=17

5 2 4

A

B

MSA I

A

n=12

n=18

n=17

n=12

10

3 4

ECA I

4

16 1

2

9

6

14

N/Rel.pr. Ag. = noun-relative pronoun agreement (in MSA the relative pronoun agrees with the noun in case, gender, and number); Av. = avoidance (producing a correct sentence but avoiding the relative construction); WO = word order. B = beginners; I = intermediate-level students; A = advanced students.

Relativization in English and Arabic

55

EFL (Item 3) The man is Japanese. His wallet is new. All the errors made by both the beginners and intermediate students relate to the relative whose. Instead of whose, students used *who his, *which his, *who's, and *whose his. The error made by the advanced learner was an error of avoidance; the student produced a correct relative clause, avoiding the possessive: The man who has a new wallet is Japanese. Again, there is evidence of transfer in this item since the errors are a direct translation from Arabic. The relative whose is problematic, as in Arabic the equivalent is the relative pronoun with a pronominal suffix attached to the noun. AFL (Item 3) /ir-ra:gil yabatni. maHfaDHit ir-ra:gil gidirda./ (ECA; MSA differs only in pronunciation) It is significant that none of the beginners got this item correct. Errors here seem to indicate that the possessive structure is one of the most difficult to acquire, which compares with the EFL learners. There is evidence of the use of avoidance strategies by intermediate and advanced learners. Word order seems to pose problems for both beginners and advanced MSA learners. There is evidence of the use of coping strategies and avoidance at the higher levels, which compares with the EFL learners. Table 4: Comparative error analysis of Item 4. Type of error

B

I

n=13 n=17

RP RPAg. N/Rel.pr.

WO

AFL

EFL

7

A n=9

6

B

ECA I

n=18

n=17

13

6

A

B

MSA I

n=12 n=18 n=17

1

14

6 5

5 1

A n=12

14

7

5 1

RP = resumptive pronoun/pronoun retention; RP Ag. = resumptive pronoun agreement (in ECA and MSA the pronoun is retained and agrees with the noun); N/Rel.pr. Ag. = noun-relative pronoun agreement (in MSA the relative pronoun agrees with the noun in case, gender, and number); WO = word order. B = beginners; I = intermediate-level students; A = advanced students.

56

Relativization in English and Arabic

EFL (Item 4) She read the two lessons. The teacher explained the two lessons. Errors made by both beginners and intermediate students were pronoun retention and use of the wrong relative pronoun. Unlike item 2, there is evidence of improvement across levels. The only error made by an advanced student was noun-relative pronoun agreement. Again, there is evidence of L1 transfer since the students were translating the Arabic equivalent: * She read the two lessons which the teacher explained them. AFL (Item 4) /'arit id-darsain. il-mudarris sharaH id-darsain./ (ECA) Errors made by 13 beginners and six intermediate students provided evidence of L1 transfer, i.e., translating the L1 equivalent and dropping the pronoun: * /'arit id-darsain illi il-mudarris sharaH/ The errors made by five advanced learners indicated fewer problems with pronoun retention but difficulty with appropriate suffixes: * /'arit id-darsain illi sharaHha il-mudarris/ /qara'at ad-darsayn. al-mudarris sharaH ad-darsayn./ (MSA) Errors made here relate to case, use of inappropriate pronominal suffixes and pronoun retention. Fourteen beginners and six intermediate students produced the direct translation from English: * /qara'at ad-darsayn alladhaan al-mudarris sharaH/ Five advanced learners produced: * /qara'at ad-darsayn alladhi:n sharaHha al-mudarris/ * /sharaH al-mudarris ad-darsayn allata:n qara'athuma/ Error analysis of Item 5. EFL (Item 5) The doctor examined the two patients. We spoke to them. RP errors persisted across the three levels: four at the beginner's, five at the intermediate, and two at the advanced level. Other errors were use of the wrong relative pronoun (by beginners), word order and preposition omission (by intermediate students): * The two patients we spoke to them the doctor examined them. * The doctor examined the two patients which we spoke. * The doctor who spoke to them examined the two patients.

Relativization in English and Arabic

57

Table 5: Comparative error analysis of Item 5. Type of error

RP RPAg. N/Rel.pr. Ag-

WO Om.Prep Av.

AFL

EFL

ECA

B

I

n=13

n=17

4

5

A n=9 2

2

3

2

B

I

n=18

n=17

12

5

A

B

MSA I

n=12 n=18 n=17

2

16

5

3

5

A n=12

3

2 1 1

1

RP = resumptive pronoun/pronoun retention; RP Ag. = resumptive pronoun agreement (in ECA and MSA the pronoun is retained and agrees with the noun); N/Rel.pr. Ag. = noun-relative pronoun agreement (in MSA the relative pronoun agrees with the noun in case, gender, and number); WO = word order; Om.Prep. = omission of preposition; Av. = avoidance (producing a correct sentence but avoiding the relative construction). B = beginners; I = intermediate-level students; A = advanced students. AFL (Item 5) /id-duktu:r kashaf 9ala il-9ayyantain. iHna itkallimna ma9a il9ayyantain./ (ECA) Problems here relate to the difficulty of attaching the pronominal suffix to the preposition rather than to the verb. * /iHna itkallimna ma9a il-9ayyantain illi kashafhum id-duktu:r/ Five intermediate students wrote the English equivalent: * /iHna itkallimna ma9a il-9ayyantain illi id-duktu:r kashaf/ It is interesting to note that whether or not a verb requires a preposition is different not only in both languages but also in ECA and MSA. Thus, this could pose problems for learners. For example, "examined (the patients)" requires a preposition in ECA (/kashaf 9ala il-mari:D/) but not in MSA (/faHaS al-mari:D/) or in English; the preposition after "speak (to)" is optional in ECA and MSA, but required in English; the verb "go (to)" requires a preposition in English and in both ECA and MSA. Items 5, 6, and 8 provide evidence that prepositions in relative clauses are problematic for learners.

Relativization in English and Arabic

58

/aT-Tabi:b faHaS al-mari:Datayn. naHnu takallamna ma9a almari:Datayn./ (MSA) This item posed a bigger problem for beginners. Only one student got it correct. Sixteen students produced the English direct translation, dropping the RP with the pronominal suffix, and making a noun-relative pronoun agreement error: * /naHnu takallamna ma9a al-mari:Datayn alla:ti faHaS aT-Tabi:b/

Errors made by the intermediate and advanced students relate to use of the inappropriate relative pronoun, dropping the pronominal suffix and making a noun-relative pronoun agreement error as seen in the underlined words: * /naHnu takallamna ma9a al-mari:Datayn alla:ti faHaS aT-Tabi:b/ * /naHnu takallamna ma9a al-mari:Datayn allata:n faHaS aT-Tabi:b * /naHnu takallamna ma9a al-mari:Datayn alladhi:n faHaS aT-Tabi:b/ * /naHnu takallamna ma9a al-mari:Datayn alladha:n faHaShuma aTTabi:b/ It is very interesting to note here that in spite of the erroneous sentences produced by the advanced MSA learners, they used the correct word order in the relative clause, i.e., the verb followed by the subject. Table 6: Comparative error analysis of Item 6. Type of error

RP N/Rel.pr. Ag. Om.Prep Av.

AFL

EFL B I n=13 n=17 7

A n=9 1

ECA MSA A B A B I I n=18 n=17 n=12 n=18 n=17 n=12 8 11 2 5 4

12

1 2

2

4

3 4

RP = resumptive pronoun/pronoun retention; N/Rel.pr. Ag. = noun-relative pronoun agreement (in MSA the relative pronoun agrees with the noun in case, gender, and number); Om.Prep. = omission of preposition; Av. = avoidance (producing a correct sentence but avoiding the relative construction). B = beginners; I = intermediate-level students; A = advanced students.

Relativization in English and Arabic

59

EFL (Item 6) The picture is beautiful. She is looking at it. Surprisingly, no RP errors were made by the beginners, whereas seven errors were made by the intermediate students and one by an advanced student. Other errors were omission of the preposition, word order, and avoidance. AFL (Item 6) /iS-Su:ra gamiila. hiyya bitbuSS 9ala iS-Su:ray (ECA) /aS-Su:ra jamkla. hiya tanDHur 'ila aS-Su:ra./ (MSA) Unlike the EFL learners, this item seemed to pose a bigger problem for beginners. Problems relate to dropping the preposition with the pronominal suffix and separating the relative clause from the head noun: * /iS-Su:ra gami:la illi hiyya bitbuSS/ * /iS-Su:ra gami:la illi hiyya bitbuSSaha/ * /iS-Su:ra gami:la illi hiyya bitbuSS 9alaiha/ MSA students produced similar errors, such as attaching the pronominal suffix to the verb instead of the preposition, and several students avoided the relative clause entirely. * /aS-Su:ra jami:la allati tanDHurha/ * /hiya tanDHur 'ila aS-Su:ra al-jami:la/ Four intermediate ECA students avoided the relative clause and produced correct sentences: /bitbuSS 9ala iS-Su:ra il-gami:l/ Comparative error analysis of item 7. Type of error

RP RP.Ag.

N/Rel.pr. Ag-

AFL

EFL B n=13 4

I n=17 10

A n=9

B n=18 8

ECA I

n=17 3

A n=12 2

B n=18 3 9

MSA I A n=17 n=12 1 3

3

RP = resumptive pronoun/pronoun retention; RP Ag. = resumptive pronoun agreement (in ECA and MSA the pronoun is retained and agrees with the noun); N/Rel.pr. Ag. = noun-relative pronoun agreement (in MSA the relative pronoun agrees with the noun in case, gender, and number). B = beginners; I = intermediate-level students; A = advanced students.

Relativization in English and Arabic

60

EFL (Item 7) These are the two newspapers. I bought them. All of the errors made by the beginners and intermediate students involved pronoun retention, four errors by the beginners and 10 by the intermediate group. It is interesting that although RP improved at the advanced level as indicated by this item, more intermediate students than beginners made the error. AFL (Item 7) /du:l humma ig-gurna:lain. 'ana ishtarait ig-gurna:lain./ (EGA) /hata:n huma: al-jari:data:n. 'ana ishtarayt al-jaridatayn./ (MSA) This item was problematic for beginners, seven of whom did not answer. Errors related to dropping the pronominal suffix, word order, and inappropriate suffixes (MSA): * /ig-gurna:lain illi ishtarait du:l/ * /'ana ishtrait ig-gurna:lain illi du:l humma/ * /du:l ig-gurna:lain illi ishtaraitha/ * /'ana ishtarayt al-jari:data:n allatarn huma/ * /hata:n huma al-jari:data:n allata:n ishtarayt/ * /hata:n huma al-jari:dattan allahdi:n ishtaraytaha/

Table 8: Comparative error analysis of Item 8. Type of error

RP Om.Prep.

Av. N/Rel.pr. Ag-

AFL

EFL B

I

n=13

n=17

3 4 2

6 2 1

1

2

ECA A n=9

B

I

n=18 n=17

MSA A

9 7

B

n=12 n=18

1

I

A

n=17 n=12

2

2

7

6

1

1

RP = resumptive pronoun/pronoun retention; Om.Prep. = omission of preposition; Av. = avoidance (producing a correct sentence but avoiding the relative construction); N/Rel.pr. Ag. = noun-relative pronoun agreement (in MSA the relative pronoun agrees with the noun in case, gender, and number). B = beginners; I = intermediate-level students; A = advanced students.

Relativization in English and Arabic

61

Item 8

EFL The meeting was interesting. I went to it. This item shows some improvement across levels. RP errors were made by beginners and intermediate students. Other errors included use of the wrong relative pronoun and avoidance of the relative.

AFL /il-igtima:9 ka:n mumti9. 'ana ruHt l-igtima:9./ (ECA) This was another problematic item. Seven beginners left it blank. Nine students made a pronominal suffix and word order error: * /il-igtima:9 ka:n mumti9 illi ruHt/ Seven intermediate students and one advanced student changed the word order to avoid using the pronominal suffix and produced a correct relative clause, i.e., evidence of the use of coping strategies: * /ruHt l-igtima:9 illi ka:n mumti9/ /ka:n al-igtima:9 mumti9an. 'ana dhahabt 'ila al-ijtima:9./ (MSA) Similarly, eight beginners left this item blank. However, seven beginners, six intermediate students, and one advanced student avoided a structure that requires a pronominal suffix and produced a correct relative clause: /'ana dhahabt 'ila al-ijtima:9 alladhi ka:n mumti9an/ It is interesting to note that although the verb "go" requires a preposition in both English and Arabic, two beginners and two intermediate students attached the suffix to the verb instead of the preposition which was dropped: */ka:n il-ijtima:9 mumti9an alladhi dhahabtu/ Item 9

EFL The members of the band came from all parts of the city. The majority of them were amateurs. None of the 13 beginners, none of the intermediate students, and only one of the advanced students got this item correct. Results here obviously indicate that this is the most difficult relative structure to acquire. Some of the errors were a direct translation from the Arabic equivalent: * The members of the band whom the majority of them were amateurs came from all parts of the city.

62

Relativization in English and Arabic

* The members of the band who came from all parts of the city the majority of them were amateurs. Other errors were pronoun retention and wrong relative pronoun: * ...which the majority of them... * ...who the majority of them...

AFL /il-9a:zafi:n fi-1-fir'a il-musiqiyya gum min kull makarn fi-1madi:na. mu9DHam 9a:zifi:n il-fir'a il-musiqiyya karnu huwa:./ (ECA) Like the EFL students, none of the beginners, none of the intermediate, and only one of the advanced students got this item correct. Almost half of the students in the three levels left this item blank. Nine beginners dropped the pronominal suffix: * /il-9a:zifi:n fi-1-fir'a il-musiqiyya illi gum min kull maka:n fi-1madi:na ka:nu mu9DHam huwa:/ Interestingly, nine intermediate and two advanced students changed the word order to avoid attaching the pronominal suffix to the expression of quantity, i.e., used a coping strategy and produced: /mu9DHam 'a9Da:' il-fir'a il-musiqiyya illi gum min kull maka:n ka:nu huwa:/ /mu9DHam 9a:zifi:n il-fir'a illi gum min kull maka:n ka:nu huwa:/

MSA /ja:' 'a9Da:' al-firqa al-musiqiyya min kull 'anHa:' al-madi:na. ka:n mu9DHam 'a9Da:' al-firqa al-musiqiyya huwa:./ Again only one advanced student produced an accurate sentence and many students in all levels left the item blank. Errors relate to word order and relative pronoun agreement. Item 10

EFL The residents were given help. All of their homes had been damaged by the flood. None of the students in any of the three levels got this item correct. However, there is evidence of the use of coping strategies. Beginners produced correct sentences like: The residents whose homes had been damaged were given help. The residents were given help because their homes had been damaged by the flood.

Relativization in English and Arabic

63

The residents were given help when all of their homes had been damaged by the flood. Errors made by intermediate students were: * ...whose their homes... * ...who their homes... * ...whom all their homes... * ...that all of their homes... Errors made by the advanced students were: * ...whose all homes... * ...which all of their homes... * The residents were given help, all of whose homes had been damaged by the flood.

AFL /ka:n fi: musa:9da li-s-sukka:n. kull biyu:t is-sukka:n iddammarit bisabab il-fayaDa:n./ (ECA) This item was as problematic for the AFL learners as it was for the EFL learners. Fourteen beginners wrote: * /ka:n fii musa:9da is-sukka:n illi iddammarit bi-sabab il-fayaDa:n/ Three intermediate students wrote: * /iddammarit kull biyu:t is-sukka:n illi fi: musa:9da bi-sabab ilfayaDa:n/ One advanced student wrote: * /ka:n fii musa:9da li-kull il-biyu:t illi iddammarit/ /laqad tamma musa:9adat as-sukka:n. dummirat kull biyurt assukka:n bi-sabab al-fayaDa:n./ (MSA) Only one advanced student produced a correct sentence. Most of the students (17 of 18 beginners, and 13 of 17 intermediate, and 11 of 12 advanced learners) left the item blank. Obviously this is the most difficult relative structure to acquire, even for advanced learners.

Discussion Results of this small-scale study cannot be conclusive. Errors that persisted across the three different levels for both EFL and AFL learners involved mostly pronoun retention, use of the wrong relative pronoun, incorrect use of the possessive whose, omission of prepositions, and problems with structures with expressions of quantity. This agrees with

64

Relativization in English and Arabic

Gass (1979, 1983) and Hyltenstam (1984) who found that speakers of languages that allow pronominal retention produce more resumptive pronoun errors. Although with some items there is evidence of improvement with level, with other items errors persisted. For both groups of learners, there is evidence of transfer errors, of developmental errors, and of the use of coping strategies to overcome difficulties, i.e., avoidance errors. The findings of this study agree with Kharma's (1983) conclusion that the errors made by Arabic-speaking EFL learners are errors of form that do not affect communication. ECA relative clauses did not seem to pose as many problems as MSA for the AFL learners. This is probably due to the fact that only one relative pronoun (/illi/) is used for the different structures. In MSA relative pronouns are more problematic for AFL learners, as they need to agree with nouns in case, gender, and number, e.g., /alladha:n/, /allatayn/, alladhi:n/, etc. Pronoun suffixes that are attached to verbs and prepositions, e.g., /katabatha/, /naDHarat 'ilayha/, etc., posed problems for the learners in both ECA and MSA, yet learners made more errors in MSA. Relative structures with expressions of quantity (items 9 and 10) were obviously the most problematic for both EFL and AFL learners. This structure seems to be the most difficult to acquire.

Conclusion and Suggestions for Teaching Findings of this study support DeKeyser's (1998) claim that "although the applied linguistics literature of the 1980s was characterized by a debate over whether or not second language instruction should make students attend to form, the vast majority of publications since the early 1990s support the idea that some kind of form is useful to some extent, for some forms, for some students, at some point in the learning process" (p. 42). Research has established the role of "consciousness" and "noticing" in foreign-language learning. Communicative focus-on-form (FonF) activities, consciousness-raising tasks, and grammar problem-solving activities have been suggested as effective techniques in the grammar class (Fotos, 1994; Ellis, 1991). "Consciousness-raising tasks suggest that learners should be deliberately directed to attend to form. Teachers and materials writers seek to make students aware of new target language items, rules, or irregularities by highlighting them in the input" (Ellis, 1991, cited by Long and Robinson, 1998, p. 17). There is evidence of L1 transfer in the errors both EFL and AFL learners in this study made. If students are made consciously aware of differences between their L1 and the L2 they are learning, they may be able to avoid transfer errors. Error-identification tasks and grammaticality-judgment tasks

Relativization in English and Arabic

65

can also help learners notice differences between the two languages. Following are examples of communicative FonF activities that students can do in class: Activity 1: Matching, a card game with incomplete sentences on cards. The students are asked to find the person who has the card with the part of the sentence that begins with the correct relative pronoun to complete their sentence. When students have found the correct completion and formed a complete sentence, they are asked to stick the cards on the blackboard for peer correction. This also serves as a visual aid. This can be done in both English and Arabic since both groups of learners have problems with relative pronouns. Example sentences: This is the man This is the restaurant

whose car I bought. where we had lunch yesterday.

/dhahabt 'ila al-maTa:9im/ /sharibt ash-sha:y/ qara'na al-garidatayn/ /ishtarayt al-qalamayn/

/allati 'akal fiiha yu:suf/ /allahdi ishtara:hu li/ /allatayn Talabahuma al-mudarris/ /alladhayn ba:9ahuma al-9a:mil/

Activity 2: Students are given structures with relative clauses of different types. In groups, students are asked to figure out the rules and to provide more example sentences. Activity 3: Students are given a list of sentences containing correct and incorrect relative clauses. In pairs or groups, students identify the errors and correct them. Activity 4: General knowledge/vocabulary quiz. Students are divided into two groups. The teacher asks each group a general knowledge/vocabulary question, e.g., What do we call a person who rules with absolute power? /man huwa ar-ra'i:s al-miSriyy alladhi ughti:1 fi-9a:m 'alf tus9uma:'a wa:Hid wathama:ni:n/ A group gets one point for identifying the relative clause and one point for answering the question. The cooperative (within groups) and competitive (between groups) nature of this activity is stimulating and enhances communication. To sum up, the findings of this study support the claim that transfer is an important factor in second language acquisition. Given the existence of that influence, teachers need to become familiar with both the language and the culture of their foreign-language learners. This awareness should help not only in pinpointing learners' problems, but also in being better able to work on them.

66

Relativization in English and Arabic

Appendix 1: ESL Combine the sentences, using the second sentence as an adjective clause: 1. The student is from China. She sits next to me. 2.1 knew the woman. I met her at the party yesterday. 3. The man is Japanese. His wallet is new. 4. She read the two lessons. The teacher explained the two lessons. 5. The doctor examined the two patients. We spoke to them. 6. The picture is beautiful. She is looking at it. 7. These are the two newspapers. I bought them. 8. The meeting was interesting. I went to it. 9. The members of the band came from all parts of the city. The majority of them were amateurs. 10. The residents were given help. All of their homes had been damaged by the flood.

Relativization in English and Arabic

67

Appendix 2: ECA /urbuT kull gumlitain bistixda:m ism al-mawSu:l al-muna:sib ma9a taghyi:r alla:zim/ 1. /iT-Ta:liba min issi:n. iT-Ta:liba 'a:9da gambi./ 2. /'ana kunt 9a:rif is-sitt. 'ana 'arbilt is-sitt fi-1-Hafla imba:riH./ 3. /ir-ra:gil yabarni. maHfaDHit ir-ra:gil gidi:da./ 4. /'arit id-darsain. il-mudarris sharaH id-darsain./ 5. /id-duktu:r kashaf 9ala il-9ayyantain. iHna itkallimna ma9a il-9ayyantain./ 6. /iS-Su:ra gami:la. hiyya bitbuSS 9ala iS-Su:ra./ 7. /du:l humma ig-gurna:lain. 'ana ishtarait ig-gurna:lain./ 8. /il-igtima:9 ka:n mumti9. 'ana ruHt l-igtima:9./ 9. /il-9a:zifi:n fi-1-fir'a il-musiqiyya gum min kull maka:n fi-l-madi:na. mu9DHam 9a:zifi:n il-fir'a il-musiqiyya ka:nu huwa:./ 10. /ka:n fi: musa:9da li-s-sukka:n. kull biyu:t is-sukka:n iddammarit bi-sabab il-fayaDa:n./

Appendix 3: MSA /urbuT kull jumlatayn bi-stixdaan ism al-mawSu:l al-muna:sib ma9a taghirr al-la:zim7 1. /aT-Ta:liba min aS-Si:n. aT-Ta:liba tajlis bi-ja:nibi./ 2. /kunt 'a9rif al-mar'a. qa:balt al-mar'a fi-1-Hafla 'ams./ 3. /ar-rajul yaba:ni. maHfaDHat ar-rajul jadi:da./ 4. /qara'at ad-darsayn. al-mudarris sharaH ad-darsayn./ 5. /aT-Tabi:b faHaS al-mari:Datayn. naHnu takallamna ma9a al-mari:Datayn./ 6. /aS-Su:ra jami:la. hiya tanDHur 'ila aS-Su:ra./ 7. /hata:n huma: al-jari:data:n. 'ana ishtarayt al-jari:datayn./ 8. /ka:n al-ijtima:9 mumti9an. 'ana dhahabt 'ila al-ijtima:9./ 9. /ja:' 'a9Da:' al-firqa al-musiqiyya min kull 'anHa:' il-madi:na. ka:n mu9Dham 'a9Da:' al-firqa al-musiqiyya huwa:./ 10. /laqad tamma musa:9adat as-sukka:n. dummirat kull biyu:t as-sukka:n bisabab al-fayaDa:n./

68

Relativization in English and Arabic

References Birdsong, D., Johnson, C., and McMinn, J. 1984. Universals versus transfer revisited. Paper presented at the 9th Boston University Language Development Conference. Celce-Murcia, M., and Larsen-Freeman, D. 1983. The grammar book: an ESL/EFL teacher's course. Rowley, MA: Newbury House. DeKeyser, R. 1998. Beyond focus on form: cognitive perspectives on learning and practicing second language grammar. In C. Doughty and J. Williams (eds.), Focus on form in classroom second language acquisition. Cambridge: Cambridge University Press. Doughty, C. and Williams, J. 1998. Pedagogical choices in focus on form. In C. Doughty and J. Williams (eds.), Focus on form in classroom second language acquisition. Cambridge: Cambridge University Press. Ellis, R. 1992. Grammar teaching practice or consciousness-raising? In R. Ellis (ed.), Second language acquisition and second language pedagogy. Philadelphia, PA: Multilingual Matters, pp. 232-241. Fotos, S. 1994. Integrating grammar instruction and communicative language use through grammar consciousness-raising tasks. TESOL quarterly, 28:323-351. Gass, S. 1983. Second language universals. In R. Di Pietro, W. Frawley, and A. Wedel (eds.), The first Delaware symposium on language studies. Newark, Delaware: University of Delaware Press. . (1979). Language transfer and universal grammatical relations. Language Learning, 29:327-344. Hyltenstam, K. 1984. The use of typological markedness conditions as predictors in second language acquisition. In R. Anderson (ed.), Second languages: a cross-linguistic perspective. Rowley, MA: Newbury House. loup, G., and Kruse, A. 1977. Interference and structural complexity as a predictor of second language relative clause acquisition. In C. Henning (ed.), Proceedings of the Los Angeles second language acquisition research forum. Los Angeles: Department of English, University of California at Los Angeles. Keenan, E., and Comrie, B. 1977. Noun phrase accessibility and universal grammar. Linguistic Inquiry, 8:63-99. Kharma, N. 1987. Arab students' problems with the English relative clauses. International review of applied linguistics, 25:257-266.

Relativization in English and Arabic

69

Larsen-Freeman, D., and Long, M. 1991. An introduction to second language acquisition research. London: Longman. Odlin, T. 1990. Language transfer: cross-linguistic influence in language learning. New York: Cambridge University Press. Tarallo, R, and Myhill, J. 1983. Interference and natural language in second language acquisition. Language learning, 33:55-76.

The Expression of Futurity In the Arabic and English Languages Mohammad Al-Khawalda From a linguistic point of view, the future tense is a controversial issue. Many grammarians (Leech, 1971; Quirk et al., 1972, 1985; Palmer, 1988; Pennington, 1988) have argued that the future is not a true tense, although more recently, it has been accepted as such by many grammarians (e.g., Dahl, 1985; Maslove, 1985; Comrie, 1985, 1989; Hornstein, 1990; Declerck, 1991). For the Arabic language, the situation is more complicated. It has been claimed that tense itself does not exist in Arabic (see AlKhawalda, 1997). Even when authorities accept the existence of tense in Arabic, futurity and the future form /sayaf9alu/ ("he will do") are ignored. Many scholars, following the traditional analysis of English, claim that the present form /'af9alu/ ("I do") is used to express futurity (Hassan, 1990; Comrie, 1991; Fehri, 1993). Because opinion on the definition of Modern Standard Arabic (MSA) is not universal, Modern Standard Jordanian Arabic (MSJA), the language of official matters in Jordan, was used as a reference point for this paper, although the author believes there is no difference between the MSA and MSJA.

Data for the Study The data base for this study was collected from three sources. To investigate the use of futurity in classical Arabic (CA), the Quran was scanned. Although CA is not the focus, the aim of investigating the use of futurity in the Quran was for comparative purposes. To investigate the use of futurity in MSJA, two sources were used. The first was the Jordanian Prime Minister's policy statement to the lower House of Parliament on September 18, 1999. The reason the researcher selected this text was that such statements frequently include a large number of promises, which means intensive usages of futurity. The second source was the Arabic translation of an English television series, Acapulco Bay (recording was made of four episodes for a total of three hours). Expressions of futurity in the series were examined in both English and Arabic. Because the aim

The Expression of Futurity In the Arabic and English Languages

71

of this study was to test the hypothesis that the future in Arabic is a true tense and that /sayaf9alu/ is basically a future form (Al-Khawalda, 1997) and to avoid complex statistical computations, the data were left as simple totals and percentages .

Analysis of the Data It must be noted that the use of the negative particles such as /lan/ and /la:/, and the use of modal verbs such as /'ara:da/ ("he wanted") were excluded from the data. In general, three expressions used to express futurity were taken into consideration: the use of the present form /yaf9alu/ ("he does/is doing") to refer to the future, the use of the prefix /sa-/ with the present form /sayaf9alu/ ("he will do"), and the use of the particle /sawfa/ with the present form /sawfa yaf9alu/ ("he will do"). In this paper, appropriate forms of/fa9ala/ will refer to the verbs in Arabic, e.g., /fa9ala/ for the past form, /yaf9alu/ for the present form, and /sayaf9alu/ for the future form.

Futurity in the Quran Table 1 tabulates the distribution of expressions of futurity in the Quran. Table 1. Distribution of futurity expressions in the Quran. sawfa yaf9alu yaf9alu sayaf9alu TOTAL

42 53 114 209

20.1% 25.4% 54.5% 100%

As the table shows, the number of occurrences of futurity is around 209 times. Out of the total 209, /sawfa yaf9alu/ ("he will do") is used 42 times, which represent 20% of total occurrences. Seventeen of the occurrences (40.4%) of /sawfa/ are with the verb /ta91amu:n/ ("you (pl.) know"): /sawfa ta91amu:n/ ("you (pl.) will know"). The present form /yaf9alu/ ("he does") is used 53 times to express the future ("he is going to do"; "he will do"), which represents 25.3% of total occurrences. However, it must be noted that 12 of these occurrences appear in compound structures coordinated by /wa/ and/or /thumma/ ("then") in which /sayaf9alu/ or /sawfa yaf9alu/, ("he will do") is used in the coordinate clause. In other words, /yaf9alu/ in separate structures is used 41 times, representing about 20% of the total. The other important issue is that /yaf9alu/ ("he does"), when used to express future time, is

72

The Expression of Futurity In the Arabic and English Languages

generally accompanied by a future temporal adverb, such as yawm alqiyama ("the day of resurrection"), al-janna ("paradise"), al-naar ("hellfire"), al-Hashr ("the gathering day"), etc. In other words, /yaf9alu/ in itself, without a future temporal adverb, does not express futurity. /sayaf9alu/ ("he will do") accounts for 114 future expressions, the most frequent construction of the total: 54.5%. It seems to this researcher that the difference between /sayaf9alu/ and /sawfa yaf9alu/ is not significant. That is, it is difficult to identify any syntactic or semantic reason why either of them is used since they are used interchangeably in several situations. For example, it is mentioned above that /sawfa ta91amu:n/ ("you (pl.) will know") is repeated 17 times, at the same time the verb /ya91amu:n/ ("they know") is used with /sa/ in different places to express the same idea, e.g., / saya91amu:na/ ("they will know") /sata91amu:na/ ("you (pl.) will know"), etc. (see, e.g., Qur., 20:135; 67:17,29). Futurity in the speech of the Jordanian Prime Minister Table 2 summarizes the occurrences of futurity in the speech of the Prime Minister. Table 2. Distribution of futurity expressions in the Prime Minister's speech. sawfa yaf9alu yaf9alu sayaf9alu TOTAL

4 2 67 73

5.5% 2.7% 91.8% 100%

As the table indicates, the total number of occurrences of futurity is 73, with /sawfa yaf9alu/ ("he will do") occurring four times, or 5.4% of the total. Again, /sawfa yaf9alu/ appears interchangeably with /sayaf9alu/ ("he will do") for example, /sawfa yakuunu/ ("he will be") alongside /sayaku:nu/ ("he will be") and /sawfa nuwa:Sl/ ("we will continue") as well as /sanuwa:Sl/ ("we will continue"). /yaf9alu/ appears twice, for 2.7% of the total. The two cases in which /yaf9alu/ is used to express futurity are found in result clauses in the same sentence: /min xhila:li hadha al-nahj tata9amaq al-musharaka'al-ka:mila fi:T'iTa:r mumara:sat alHurriya:t wifqa mabad' seya:dat 'l-qa:nu:n mimma: yakfalu 'ija:d altawazun al-maTlu:b/ ("Through this way [of democracy] full participation is strengthened within the framework of freedom according to the principle of the sovereignty of the law, which guarantees the required balance"). It seems that the use of /sawfa yaf9alu/ ("he will do") and /yaf9alu/ ("he does") to express futurity is accidental. When we compare the num-

The Expression of Futurity In the Arabic and English Languages

73

her of their occurrences with the number of the occurrences of /sayaf9alu/ ("he will do"), we find that 67, or 91.7%, of the total expressions of futurity are /sayaf9alu/.

The translation of Acapulco Bay Table 3 summarizes the usage of futurity in the Arabic translation of the English-language television series.

Table 3. Distribution of futurity expressions in an Arabic translation of an American TV series. sawfa yaf9alu yaf9alu sayaf9alu TOTAL

2 1 131 134

1.5% 0.8% 97.7% 100%

As the table indicates, a total of 134 expressions of futurity occur. Of this total, /sawfa yaf9alu/ ("he will do") is used twice (1.4%). It seems that /sawfa yaf9alu/ is used randomly and there is no explanation why it is selected rather than /sayaf9alu/ ("he will do"). The present form /yaf9alu/ ("he does") is used once (0.8% of the total), where it is a structure coordinated by /'aw/ ("or...."), e.g., /'aw 'aqtuluka/ ("or I kill you"). As can be noted, the use of /sawfa yaf9alu/ and the present form /yaf9alu/ to express futurity is very low in comparison with the use of /sayaf9alu/, which appears 131 times, or 97.7%, of the total.

Futurity in English As mentioned above, a three-hour recorded videotape from the Englishlanguage series Acapulco Bay is used to analyze how futurity is expressed in English. Table 4 summarizes the ways in which futurity is expressed in Acapulco Bay.

Table 4. Distribution of futurity expressions in an American TV series. "Will"

42

30.9%

present from "be going to"

31 27

22.8% 19.9%

19

14.0%

present progressive form modals TOTAL

17

12.5%

136

100%

74

The Expression of Futurity In the Arabic and English Languages

As can be noted, unlike Arabic in which futurity is expressed in three ways (/yaf9alu, sayaf9alu/, or /sawfa yaf9alu/), five ways are used to express futurity in English. Of 136 cases, will appears 42 times, or 30.8%, of the total usage of futurity, which is the highest percentage. The present form is used to express futurity 31 times, or 22.7%; be going to is used 27 times, or 19.8%. The present progressive/from_appears 19 times, for 14% of the total. The rest, such as the use of would, have to, etc., are classified under modals, which are used 17 times, or 12.5% of the total.

Discussion and Conclusions The difference between the two languages, Arabic and English, in expressing futurity is significant. The dispute over whether the future is a tense in English, mentioned above, cannot be applied to Arabic. In other words, the arguments that are used to show that English lacks a future tense are not applicable to Arabic. For instance, it is contended that will indicates a general tendency, probability, willingness, etc., rather than future time, e.g., "Boys will be boys" (Pennington, 1988, p. 71). The data demonstrate that such usages typical of English (the nontemporal reference of the future morpheme) are not found either in CA or MSJA. Furthermore, /sayaf9alu/, ("he will do") has one and only one usage: locating the situation sometime after the moment of speech. That is, it truly expresses future temporal reference. Another argument against treating the future as a tense and that will is the future morpheme is that there are many ways to express futurity in English. The data support this argument. Table 4 shows the five general ways to express futurity in English, and none of them could be considered marginal. The situation in Arabic and in particular MSJA, which is the focus of this paper, is different. The data indicate that futurity in MSJA is expressed basically by the future verb form /sayaf9alu/. It scores 91.7% in the speech of the Jordanian Prime Minister and 97.7% in the translation of Acapulco Bay, whereas /sawfa yaf9alu/ and /yaf9alu/ are used marginally in both. The other important argument is that the use of "will" "be going to" the present form, the present progressive form, or modals is not arbitrary. Each of them has its own semantic meaning in addition to futurity. In other words, the differences among them is significant. For example: •It will rain (general future) •It is going to rain (there is an evidence for that) •*It is raining (ungrammatical for expressing the future since the progressive indicates arrangement) •I will visit him (sudden decision)

The Expression of Futurity In the Arabic and English Languages

75

•I am going to visit him (the decision has already been made ) •I am visiting him tomorrow (arrangement) In Arabic, such distinctions are not found. It seems that the selection of/sawfa yaf9alu/ and /sayaf9alu/ is arbitrary, since as mentioned above it is difficult to find any semantic or syntactic reason for selecting one expression over the other to express futurity. Moreover, unlike the simple present and past forms, future in English is expressed in a periphrastic way, that is, will is not an inflection (Quirk el al., 1985; Palmer, 1988) The future (i.e., /sayaf9alu/) in Arabic is not periphrastic. It is an orthographic whole and must be considered a morphological whole. In spite of these arguments against treating the future as a tense, many scholars argue that the future in English is a tense and that will is the future morpheme. They state that it is possible for a given morpheme to have more than one reading, and it is possible for will to express modal meaning in addition to the temporal one (for the discussion of this issue and why will is treated as a future morpheme, see Comrie, 1985; Hornstein, 1991; Declerck, 1991). Consequently, there is no reason to explain why futurity in Arabic is ignored. It is clear that the number of occurrences of /sawfa yaf9alu/ ("he will do") and the present form /yaf9alu/ ("he does") to express futurity is higher in CA than in MSJA. Nevertheless, this cannot be taken as an argument against an Arabic future tense since nowadays we are talking about MSA and that the use of the present form to express futurity is a universal phenomenon (Comrie, 1985, pp. 44-45; Declerck, 1991, p. 10; for further discussion, see Al-Khawalda, 1998). To sum up, according to our data and the above discussion, the future in Arabic is a true tense that is expressed primarily by the verb form /sayaf9alu/.

References Binnk, R. I. 1991. Time and the verb: a guide to tense and aspect. New York: Oxford University Press. Comrie, B. 1985. Tense. Cambridge: Cambridge University Press. . 1991. On the importance of Arabic to the general linguistic theory. Perspectives on Arabic linguistics, 3:3-30. Dahl, O. 1985. Tense and aspect system. Oxford: Blackwell. Declerck, R. 1991. Tense in English: its structure and use in discourse. London: Routledge. Declerck, R. 1995. Is there a relative past tense in English? Lingua, 97:1-36. Depraeteve, I. 1994. Some observations on the expression of temporal relation in future-time relative clauses. Linguistics, 32:459-473. Fassi, F. 1993. Issues in the structure of Arabic clauses and words.

76

The Expression of Futurity In the Arabic and English Languages

Dordrecht: Kluwer. Hassan, H. M. 1990. A contrastive study of tense and aspect in English and Arabic with special reference to translation. Ph.D. diss., University of Bath, UK. Hornstein, N. 1990. As time goes by. Cambridge, MA: MIT Press. Al-Khawalda, M. 1997. Tense, aspect and temporal reference. Ph.D. diss., Essex University, UK. . 1998. Are "fa9ala" and "yaf9alu" temporal or aspectual forms? Paper presented at the 14th International Conference on Language and Linguistics, Literature, and Translation, Yarmouk University, Irbid, Jordan, April 28-30, 1998. Klein, W. 1994. Time in language. London: Routledge. Leech, G. 1971. Meaning and the English verb. London: Longman. Maslov, Y. 1985. Outline of contrastive aspectuality. In Y. Maslov (ed.), Contrastive studies in verbal aspect. Heidelberg: Julius Groos. Quirk, R.G., Leech, S., Greenbaum, S., and Svatvik, J. (1972). A university grammar of English. London: Longman. . 1985. A comprehensive grammar of the English language. London: Longman. Palmer, F. 1988. The English verb (2nd ed.) London: Longman. Pennington, M. 1988. Context and meaning of English simple tenses. RELC journal, 19:49-74.

A Sociolinguistic Study on the Use of Color Terminology in Egyptian Colloquial and Classical Arabic Jehan Allam The six basic color categories (white, black, red, green, yellow, and blue) belong to the physical domain and can be defined and measured in terms of their wavelengths and their relative heat intensity (Omar, 1982, p. 91). Languages seem to differ in how they treat colors. Thus, the idea of "turning yellow" indicates one is "becoming ill" in Egyptian colloquial Arabic (ECA), but in English it suggests one is "lying" or "cowardly"; to become "white-faced" means to "become ill" in English, but to be "pleased or proud" or to "emerge as victorious" in ECA and in classical Arabic (CA). According to the dictionary Lisan al-Arab, the Arabic words /lawn/ ("color") and /'alwa:n/ ("colors") are from the root /1/w/n/, which denotes a state such as blackness or redness. The color of anything is what distinguishes it from other things. Thus, /'alwa:n/ ("colors") also can refer to personality types. For example, a person is described as /mutalawwin/ ("changing his color") if he is not of a strong or steady character. /'allawn/ ("the color") is also used in CA to denote a "palm tree," on the grounds that dates grow through a series of colors (green, yellow or red, black or brown) similar to the colors that appear during sunset to darkness. The focus of this paper is on the sociolinguistic aspects of/'al-lawn/ as used in the Arabic language. Psychological aspects will also be included to examine color usage in the language. Colors have their place in many customs and traditions. It is known that green and white, for instance, are optimistic colors for Egyptians in general. Green is a sign of prosperity and white is a sign of happiness. Black and blue, on the other hand, are colors of mourning, pessimism, and sadness (Amin, 1952, p. 57). In other societies, this is not necessarily the case. In some societies, black is a color of happiness, respect, and elegance. Therefore, it is important to recognize the connotations of various color words as they are used in each language. Due to the psychological effects of colors, people tend to associate them with different aspects of their lives. Historically in the

78

A Sociolinguistic Study on the Use of Color Terminology

West, blue was associated with law and the king's court, because in Roman times, prefects wore blue cloaks. "Royal" purple is just as ancient. Many of these connections have a more rational basis: blue with the sea, summer, and coolness; green with the countryside, spring, and coolness; yellow with sunlight, heat, and cheerfulness; red with warmth, passion, excitement, and fire; white with cleanliness and purity (Danger, 1968, p. 29). Aside from the psychological associations of colors, color terms often acquire, in certain fixed allocations, a range that goes beyond what they normally possess, e.g., "white" in the expression "white coffee" refers to a deep shade of brown, and a "white" (person) may be anything from an off-pink color (natural pigmentation) to golden brown (sun-tanned). In the latter example, "white" contrasts with "black," the range of which is similarly extended to cover various shades of dark brown (Crawford, 1982, p. 339). On the other hand, when referring to skin color, an Arabic speaker may use /'abyaD/ ("white") as a euphemism for /'aswad/ ("black"). In general, various societies and, as a result, various languages tend to treat colors in differently. Therefore, a sociolinguistic study of colors should show similarities or differences, since the vocabulary of color analysis is psychological and metaphorical rather than rigorously physical, for color is part of nature, life, and human relations.

/'abyaD/ ("white") /'abyaD/ is derived from the root /b/y/D/ ("to become whitened"); /bayDa:'/ is its feminine form; /al-'abyaD/, with the definite article /'al/ may be a noun or an adjective. As a noun it denotes the color itself generically, in opposition to the rest of the members of the color spectrum. /'abyaD/ is emotionally connected with purity and goodness; it is a symbol of optimism and happiness. "The color traditionally and universally used for wedding celebrations and brides is /'abyaD/" (Omar, 1982, p. 166), although in some societies, /'aswad/ ("black") is the color for celebrations. /'abyaD/ represents one end of the color spectrum: in psychological terms it generates positive rather than negative feelings. In CA, the sense of /'abyaD/ usually revolves around purity and light. However, there is the exceptional negative impact of white in the Quranic story of Joseph: there, /'abyaD/ occurs in connection with Jacob's eyes to describe his blindness (Qur., 12:84). A brief study of the use of color terminology in the Holy Quran found that the six basic colors (white, black, yellow, red, green, blue) were mentioned 32 times. /'abyaD/ is the color mentioned most often (10 times), probably because of its psychological connotation to goodness, purity, and heaven and its distinctive contrast with /'aswad/ ("black"), which is related to evil, sinfulness, and hell.

A Sociolinguistic Study on the Use of Color Terminology

79

Hence, reference to /'abyaD/ appears in both CA and ECA extensively. A survey of the meanings of /'abyaD/ in CA and ECA show that it is used in similar senses. Common expressions in ECA are frequently carried over from CA and may have metaphorical as well as literal meanings: /baya:Du-'al-naha:r/ ("the whiteness of the day," i.e., "daylight"), /baya:Du-'al-bi:D/ ("egg white"), /baya:Du'al-9ayn/ ("white part of the eye"). Common CA metaphors likewise are often retained in ECA: /bayyaDa wajhuhu/ (literally, "made his face white" and metaphorically "made him proud"), /kidhbun 'abyaD/ ("white lie," as in English), /yaddun bayDa:'/ ("white hand," i.e., generous). All these expressions have been carried over from CA to ECA. Figurative meanings may vary with different derivatives of the root. For example, masculine /abyaD/ and its feminine /bayDa:'/ is used, to denote among other things, a beautiful woman, but /'abyaD/ does not apply to the male parallel; in fact, it is used euphemistically to refer to a black man. The opposite again is not true, a black woman is not referred to as /bayDa:'/. Arabic /'abyaD/ and English "white" more or less agree in invoking the following meanings: sunlight, white race, silver or gray color, blank space, bleached clothes, and household items such as bed sheets and towels; in both languages the words are applied as adjectives to mean: honest and dependable, fortunate, clear or free from spots, morally pure, innocent, and harmless. Common expressions with "white" include: "white lie" (a minor lie uttered from polite, harmless motives), "white smith" (tin smith), and "white space" (Random House Dictionary, 1956). The Arabic term /'al-abyaDa:ni7 "the two whites" (said of milk and water) is a sign of beauty and seems to contrast in this sense with English, since milk and water are not referred to as the two whites and "white" denotes paleness, as from fear or other strong emotion. Other senses of white as used in English include that characterized by snow, e.g., "white Christmas"; in a political sense, white is used to describe something radically conservative; auspicious, or fortunate. Magic that harms no one is "white," whereas coffee that is "white" contains milk. To "bleed white" means to be or cause to be deprived of all one's resources. A project still "in the white" is in an unfinished state or condition; a "white alert" of military defense is signal that danger no longer exists. A "white beard" refers to an old man; a "white list" is a list of individuals or organizations with security clearance from government officials; someone who is "white livered" lacks courage or vitality; "white trash" is a slur on poor whites collectively; "white wing" is one who wears a white uniform, especially a public street cleaner. Other senses of white as used in Arabic include: /yadun bayDa:'/ ("white hand"), which means proof (whether by argument, allegation, or evidence) as well as a favor or benefit for which one is not reproached and

80

A Sociolinguistic Study on the Use of Color Terminology

which is conferred without being asked. It is a sign of generosity and good will. Similarly, we find in ECA, /bayyaD wishshuhum/ ("whitened their face," i.e., he made them proud of him as a result of his success or achievement) and /naharak 'abyaD/ ("your day is white," that is, it is a good day). Other common metamorphic expressions include: /'il-'irsh'il-'abyaD yinfa9 fi-l-yu:m-'il-'iswid/ ("a white piaster is useful on a black day," i.e., a little money saved helps when times are bad), /ma9andu:sh la:-abyaD wa-la:swid/ ("he has neither white nor black," i.e., he is penniless), and /sha:y 9ala mayya bi:Da/ ("tea on white water," i.e., tea made without being boiled). In CA the following expressions occur: /'al-'ayyamu-l-bi:D/ ("the white days," i.e., the 13th, 14th, and 15th nights of an Islamic month, when there is a full moon), /'al-'arDu-l-bayDa:V ("the white land," i.e., a waste of uncultivated land), and /'al-'abyaDa:n/ ("the two whites," i.e., milk and water).

/'aswad/ ("black") Emotionally, /'aswad/ ("black") is the symbol for grief, pain, misery, and death. It represents darkness, the unknown, suppression, nonexistence, and destruction. It is the color that prevails after fires; it colors the walls of old buildings in cities that have suffered great fires. In many languages, /'aswad/ ("black") is associated with charcoal, darkness, and night (Berlin and Kay, 1969). /'aswad/ ("black"; the root is /s/w/d/) is the color of evil, black magic, bad luck, and poverty, and it is associated with racial discrimination and slavery. For Egyptians and many others, /'aswad/ is the color of mourning and death. It is the color used in funerals and dresses of women in mourning. Although it is the color of pessimism, it is also the favorite color for beautiful eyes. This is probably because most Arabs have relatively dark complexions with which black eyes and black hair are the most suitable (Amin, 1963, p. 58). A large vocabulary describes black eyes, which shows Arab's appreciation of the beauty seen in the /sawa:d/ ("blackness") of the eyes. Poetic expressions to describe the eyes underscore that fine appreciation: /9aynun barja:'/ (i.e., eyes that have very clear white and very clear black); /'al-Hawar/, which is deep black contrasted with the clear white of the eye; and /9aynun da9ja:V, i.e., very dark black eyes (Ibn Sidah, 1898, pp. 98-99). In the Holy Quran, /'aswad/ is mentioned seven times, ranking it third after /'abyaD/ ("white") and /'axDar/ ("green"), /'aswad/ is used as a verb, /taswwadu/ ("to become black") and /'iswaddat/ ("became black"). It contrasts with /'abyaD/: it is hell opposed to heaven, misery opposed to happiness, mourning opposed to celebration, and night opposed to day. Metaphorically, it has been used to mean gloominess or misery: /'iswaddat wuju:hahum/ ("their faces became black") meaning became sad or dis-

A Sociolinguistic Study on the Use of Color Terminology

81

appointed, and /'al-xayTu-l-'aswad/ ("the black thread") representing the end of the night from the beginning of the day. In CA, the sense of /'aswad/ revolved around the meaning of pessimism, grief, darkness, sinfulness, as well as nighttime, dark complexion, and the beauty of the eyes. It also used to be associated with slavery, servants, or an inferior race. The only positive use of /'aswad/ is when describing the eyes, and usually other words are used for that purpose. In ECA, /'aswad/ is pronounced /'iswid/. As in CA, it is the color of pessimism. Just pronouncing the word is sometimes objectionable; that is why certain expressions use /'abyaD/ instead, e.g., /ya-nha:r 'abyaD/ ("what a white day," euphemistically for "what a black day") or /ya-xabar abyaD/ ("white news," which is used when hearing bad news), /'iswid/ is also associated with black magic and evil. /'uTTa su:da/ (a "black cat"), when it passes in front of someone, is believed to be a bad omen. People's pessimism from /'aswad/ is obvious both in CA and in ECA when they call a black man /ya-'abyaD/ and when they say in ECA/ya-ssmar/ ("tanned") in describing a black man to avoid saying /'iswid/. Comparing /'aswad/ in CA and ECA reveals similarities. Both refer to night as /'aswad/ and both appreciate black eyes: in ECA /9asha:n sawa:d 9uyu:nak/ ("for the sake of your black eyes") and /'il-9uyu:n-is-su:d/ ("the black eyes") and their beauty are commonly mentioned in songs as well as in poems in CA. Another common meaning shared in both CA and ECA is the use of /'aswad/ to mean a rough copy /miswadda/ or /taswi:da/ in ECA and /sawwadda/ in CA. Arabic /'aswad/ and English "black" more or less invoke the same following senses: pertaining or belonging to an ethnic group characterized by dark skin pigmentation, such as the dark-skinned peoples of Africa; characterized by the absence of light (e.g., "black night"); gloomy, pessimistic, dismal, sullen (e.g., "black looks"); and deliberately harmful, without any moral right of goodness, evil, wicked (e.g., "black hearted"). In both languages, "black" indicates censure, disgrace, or liability to punishment, and it is used to describe clothing for mourning. Finally, the expressions "black gold" (for "petroleum"') and "black market" are found in both English and Arabic. Metaphoric expressions used in CA to indicate different meanings of /'aswad/ include /'idha: kathura-l-baya:D qalla 'assawa:d/ (lit, "when whiteness increases, blackness decreases") where whiteness signifies milk and blackness dates (Al- Zubaydi, 1888); /sawwad/ (lit., "to cause blackness") is to write something in a rough or draft copy (Lane, 1865); and /sa:wadahu/ (lit., "made him black") is to meet someone in the blackness of the night. From this sense, it developed to mean "in secrecy," as in /sawwadtuhu/ ("I spoke secretly with him") and /sa:wada/ ("a secret speech"; Lane, 1865). /waT'atun sawda:'/ (lit., "a black footstep") denotes a recent footprint. Some of the Arabs' common expressions include: /kaththartu sawa:da-l-qawmi bi-

82

A Sociolinguistic Study on the Use of Color Terminology

sawa:di/ ("I increased the number of the collective body of people by my person"), in which /sawaad/ means "myself," with /'asa:wid/, the plural form, meaning "persons" (Lane, 1865); /sawa:du-l-9ira:q/ ("the blackness of Iraq") to refer to the districts, towns, or villages and cultivated lands of Iraq, so called because of the greenness and near-blackness of its trees and produce (Lane, 1865). /'al-dhahab al-'aswad/ ("black gold"), meaning "petroleum," is an introduced term; /'aswad-al-qalb/ ("black heart") refers to an unforgiving person. In ECA we find: /yu:m 'iswid/ ("black day'") for a bad day and /xabar'iswid/ ("black news") for bad news; /wa'9itak su:da/ and /naharak 'iswid/ both can be used as a threat; /maza:g sawdawi/ refers to a melancholic temperament; and /saHa:ba su:da/ ("black cloud") signifies a problem. In EGA the associations of black with the meaning of shameful or disgraceful is apparent in expressions such as: /9amlitak is-su:da/, literally, "your black doing," that is, "your shameful deed." Common proverbs that give this meaning include /ma yinu:b il-kaddab 'ilia sawa:d wishshu/, literally, "the liar gets only the blackness of his face" and /'il-di:n sawa:d-il-xaddi:n/, "being in debt means black cheeks," which means that debt causes one misery and shame (Taymour, 1956, p. 235).

/'aSfar/ ("yellow") /'aSfar/ is the color of the sun, it resembles /'abyaD/ ("white") and daylight, and is associated with energy, readiness, and activity. It is a brilliant and cheerful color. It is associated with sunshine, heat, cheerfulness, and summer. People who prefer yellow are said to have an intellectual bent (Danger, 1968, p. 27). According to Max Lusher's psychological color test, /'aSfar/ means a search for a way out of troubles. If chosen number one, it is a sign of happiness, hope, and good expectations, and progress and innovation. /'aSfar/ is mentioned five times in the Holy Quran and ranks fourth after /'abyaD/ (white"), /'axDar/ ("green"), and /'aswad/ ("black") in the number of times it occurs. It is used in the sense of the basic color yellow; bright yellow describes good, healthy, and strong animals. /'aSfar/ in CA revolved around gold and wealth—even good camels, in which Arabs took pride and considered part of their wealth, are /Sufr/ (pi., "yellow"). From this sense developed the expressions /Sifru 1-yaddayni/ and /ma li-fula:n la: Safra:' wa-la: bayDa:'/ to mean "penniless or empty handed." /'aSfar/ then is a synonym for wealth because it means gold and wealth. It is also used to express common physical examples of the color: dry plants and leaves are /Sufr/ ("yellow"); the sun is /'al-Safraa'/; and gold and saffron are referred to as /al-'aSfara:n/ (lit., "the two yellows"). In CA, /'iSfarra/ (MSA, "to become yellow") was used to mean "to become yellow and also black"), /Sufra/ is defined as "yellowness" and "blackness" (Lane, 1865), particularly when referring to black camels. The Arabs, presum-

A Sociolinguistic Study on the Use of Color Terminology

83

ably, conceived of black camels as yellow and/or black, hence the mixing of the two terms. This sense is not apparent in ECA. In ECA, in addition to the basic sense, /'aSfar/ revolves around sickness, e.g., /wishshu 'aSfar zayy-il-lamu:na/ ("his face is as yellow as a lemon," i.e., pale either of fear or bad health), and /'iS-Safra/ is jaundice. /'aSfar/, when it describes a complexion, indicates a sign of illness, while yellowness of the face could indicate a disease in the liver, anemia, or poisoning (Omar, 1982, after Birren, 1950, pp. 110-113). It also denotes something old: /waraq-ilkita:b 'iSfarr/ ("the pages of the book became yellow") and /waraq-ish-shagar 'iSfarr/ ("the leaves of the trees became yellow"). /'iS-Sara:ya-iS-Safra/ ("the yellow palace") is slang for a mental hospital; the origin of this name could not be determined for certain, but it might be related to the original color of the building. There is a common belief that yellow stones prevent misery or upsets, as expressed by the proverb /'illi yilbis 9aqi:q 9umru mayshu:f Di:q/ ("he who wears carnelian is always happy"). The sense of wealth as mentioned in CA does not exist in ECA in relation to /'aSfar/. Comparing Arabic /'aSfar/ with "yellow" in English shows a similarity in the two languages' use of the color to denote or describe jealousy and envy; the yolk of an egg; any yellow metal (naval brass, gold); sallow complexion, jaundice, and race (of persons). Metaphorically, ECA uses /'aSfar/ in the following expressions to denote spite: /fula:n da Safra:wi/ ("this person is yellow," i.e., he is spiteful), /DiHka Safra/ (a cunning or deceitful smile), /9i:nha Safra/ (lit., "her eye is yellow," i.e., she is envious), and /maza:gu Safraawi/ ("he has a choleric temperament"; Badawi and Hinds, 1986). The color is also used frequently to denote sickness: /wishshu 'aSfar zayy-il-lamu:na/ (lit., "his face is as yellow as a lemon," i.e., his face is pale either from fear or bad health), /'iS-Safra:/ (lit., "the yellow," i.e., jaundice), and /'iS-Sara:ya-iSSafra/ (slang, "mental hospital"). In short, English use of yellow signifies cowardice, fear and illness; ECA /'aSfar/ signifies distress and illness, whereas in CA it signifies wealth and richness.

/'aHmar/ ("red") Psychological studies proved that red has an evocative effect; it arouses feelings of anxiety, tension, courage, revenge and desire for attacking. It causes muscle tension and increases body temperature. It has a positive and lively effect on people. By contrast, red in food, studies have proved, is the most appetizing and tempting color (Omar, 1982, p. 154). In general, /'aHmar/ ("red") is the color of blood, danger, fire and alertness. It is the color that attracts the eye. Therefore, it is used for safety measures, e.g., red cross for hospitals, red fire extinguishers, red light of traffic

84

A Sociolinguistic Study on the Use of Color Terminology

light for stop signal, etc. It is universally associated with danger and alertness. In the Holy Quran, /'aHmar/ is mentioned only once as an adjective denoting brown color describing mountains. /'al-'aHmar/ ("the red") in CA is strongly associated with white complexion and beauty, as well as blood and violence, an apparent contradiction, for both impressions are at two extremes if put on a single scale. In the first sense, /'aHmar/ is used instead of/'abyaD/ ("white") because /'abyaD/ has another significance when related to complexion—it denotes illness. According to Lisan al-'Arab (Ibn Manzur, 1982) "red and white are antonyms." This sense occurred in the sunna: /HumayraaV was applied to Aisha, the Prophet's wife. The colors of the expression /'ata:ni kull 'aswad minhum wa-'aHmar/ ("every black and red of them came to me,") denotes every single one, and one should not use /'abyaD/ ("white") in this sense (Lane, 1865). /'al-Husnu 'aHmar/ (lit, "beauty is red") means that beauty is in fair complexion; it is also a metaphor indicating that beauty is attended by difficulty, that is, he who loves must bear difficulty or distress, or "the lover experiences from beauty what is experienced in war" (Lane, 1865). In ECA, the sense of /'aHmar/ denoting fair complexion occurs in the proverb /min barra Hamra Hamra wi min guwwa halla halla/, meaning "beautifully made up from the outside but rotten inside." /'aHmar/ in both CA and ECA is associated with death, blood, and violence. In CA, we find the following expressions: /'al-mawtu l-'aHmar/ ("red death," i.e., slaughter ) and /Hamira 9ala:yya/ ("he became red on me," i.e., he burned with anger and rage against me); /Hamira/ ("became red") also applies to a horse that is out of control (Lane, 1865). The color is found in ECA in several metaphoric expressions: /warra:lu-l-9i:n-ilHamra/ ("showed him the red eye," i.e., behaved in a severe, intimidating manner toward someone), /Hammarlu 9i:nak/ ("make your eye red to him," i.e., be severe with him; Badawi and Hinds, 1986), and /'il-bijama:1-Hamra/ ("red pyjamas," i.e., the death penalty). In CA, /'al-'aHmar wa-l-'abyaD/ ("the red and the white") denotes gold and silver. /'al-'aHmar/ ("the red") is also a metaphor referring to a sort of dates that are red in color, /sanatun Hamraa'/ ("a red year") is a year of severe drought because in such years the tracts of horizon are reddish, i.e., brown in color, due to the heat and dryness (Al-Numayri, 1976, p. 35). /Hammara/ ("he caused to become red") means either to dye a thing red or to write with red ink. /Hamira/ ("to become red") in the context of leatherworking means "he pared a thong; stripped it of its superficial part or he (a sewer of leather or of skin) pared a thong by removing its inner superficial part and then oiled it, previously to sewing with it, so that it became easy to sew with; apparently because this operation makes it to appear red or reddish color" (Lane, 1865). /Hamira/ is removal of soft hair or fur and wool and /'iHamara ma 9ala 1-jild/ means "what was upon the skin became

A Sociolinguistic Study on the Use of Color Terminology

85

removed (said of hair and of wool)." Other expressions include: /Hammarahu bi-1-SawT/ ("he excoriated him with the whip") and /waT'atun Hamra:'/ ("a red footstep," i.e., a new of recent footstep or foot print; Lane, 1865). /Humrann/ is an adverb used metaphorically in the sense of having a weapon on oneself. The sword is very important to the Arab and for one to be without his weapon or sword is comparable to removing an important part of himself. /'aHmar/ in this sense is not accounted for in ECA. In ECA, common expressions include: /Hammar/ ("to make red," i.e., to fry in oil, /'itHammar/ or /'iHmmar/ ("browned or fried, became red"), /'iHmarri:t min ish-shams/ ("my skin became red, i.e., burned, from sitting in the sun"), /baTaaTis miHammara/ (lit., "potatoes made red," i.e., French fries), /laHma miHammara/ ("browned meat"). It is notable that /'aHmar/ is used in food rather not /bumuV ("brown"); this supports the concept that red is a favorite color in food. Further support can be seen in the phrase /Hama:r wi Hala:wa/ (lit., "redness and sweetness"), which is used to praise ripe watermelons. /'aHmar/ in ECA, as in CA, is associated with death, blood, and violence. In the Egyptian culture, /'aHmar/ is also a color of youth, cheerfulness, and immodesty. It is connected with make-up, decoration, and glamour, with children, young girls, and brides; elderly ladies are strongly criticized if wearing /'aHmar/. It seems that /'aHmar/ is opposed to /'iswid/, psychologically or socially, since it is not acceptable to visit a person in mourning when wearing red or even red make-up. The metaphor /sahra Hamra:'/ ("red evening") is probably borrowed from the English expression "painting the town red," i.e., celebrating "boisterously, especially by making a round of stops at bars and nightclubs" (Random House Dictionary, 1956).

/'axDar/ ("green") Egyptians consider /'axDar/ ("green") a color of optimism and a sign of prosperity (Amin, 1953, p. 58). It is associated with the color of plants and agriculture, which represented a major source of their economy; the more green is produced, the more prosperous is their year. Ever since the pharaohs, agriculture represented a main occupation for the Egyptians. As for the Arabs in general, /'axDar/ is a precious color because of its scarcity amid the vast yellow desert lands. /'axDar/ ("green") is from the root /x/D/r/ ("to become green"). Next to /'abyaD/ ("white"), /'axDar/ is the second most frequently mentioned color in the Holy Quran (eight times). It is used to refer to plants and greenery as well as to signal richness, prosperity, and heaven. /'axDar/ in CA embraces several senses. It is associated with crops, greenery, trees, profit, prosperity, and youth. It is strongly connected with

86

A Sociolinguistic Study on the Use of Color Terminology

agriculture and plants. From this sense developed connotations such as fresh (as opposed to dry) and unripe, which are extended metaphorically in expressions such as: /'al-mawadatu baynana xaDra:'/ (lit., "affection between us is green," meaning our affection is recent or fresh) and /'ixtuDira sh-shaab/ (lit., "the young man has become green," i.e., he died in his youth). /xuDra/ ("greenery") describes trees and landscape as well as pot herbs and green vegetables. Metaphorically, we find /ma:'un 'axDar/ (lit., "green water") for crystal-clear water and /xuDara/ ("greenery") for the sea (Ibn Manzur, 1982). In the sunna, we find /man xuDDira lahu fi shay'in fa-liyalzamahu/ ("whosoever is blessed in a thing [meaning an art or a trade, or a means or subsistence] let him stick to it"; Lane, 1865) and /xuDDira/ (lit, "made green") as a metaphor for blessing. Other expressions include: /9i:shatun xaDra:'/ (lit., "green life," i.e., a mode of life soft and delicate, plentiful, and pleasant; Lane, 1865) and /'ixDDara jilduhu/ (lit., "his skin became green" from carrying the produce of the land), which is a metaphor for prosperity (Lane, 1865). In ECA, /'axDar/ also denotes prosperity as evidenced by expressions such as: /sana xaDra bi-'idhn-i-llah/ (lit., "a green year if God wills," said when wishing someone a happy and prosperous year), /'adamuh xaDra 9ali:na/ (lit., "his feet are green," i.e., he brings us good luck), and /'i:du: xaDra/ (lit., "his hand is green," i.e., he has a lucky touch or he brings profit to whatever he involves himself in; Badawi and Hinds, 1986). In CA, /'axDar/ denotes /'aswaoV ("black") when referring to complexion: /'inna-1-Harith 'ibn-al-Hakam tazawwaja 'imra'atan fa-ra'a:ha xaDra:' ('ay sawda:') fa-Tallaqaha:/ (lit., "al-Harith ibn al-Hakam married a woman and discovered her to be green, i.e., black, so he divorced her"; Lane, 1865). Similarly, blackness is understood when /'axDar/ is applied to a variety of objects: /'al-katu:ba 'al-xaDra:' (lit., "the green regiment," i.e., a group of soldiers with black shields), /'ixDarra-1-layl/ (lit., the night became green," i.e., became dark and black), and /'arD-as-sawa:d/ (lit. "the land of blackness," a euphemism for Iraq, because its fertile lands and richness in greenery and trees appear black to someone coming from far away in the desert; Lane, 1865). /'axDar/ is sometimes used to mean /'azraq/ ("blue") in CA: /ma taHta l-khaDra:'t 'akrahu minh/ (lit., "there is not under the sky more hateful than he"; Lane, 1865). The interchanging of colors, /'axDar/ ("green") for /'azraq/ ("blue") and /'aswad/ ("black") in CA is interesting, yet it seems to follow a certain logic: from a distance dark green indeed seems black (as in /'arD-assawa:d/, "the land of blackness") and since /'axDar/ is the optimistic color that stands for prosperity, it is used to describe the color of the sky instead of/'azraq/, which has bad associations and is disliked (see below).

A Sociolinguistic Study on the Use of Color Terminology

87

ECA seems to conform to the general impressions of/'axDar/ in all the senses of CA, except that in ECA another sense has developed—/'axDar/ in the sense of wet and damp: /'il-hudu:m lissa xaDra/ (lit., "the clothes are still green," i.e., damp) and /'il-'arD xaDra/ (lit., "the ground is green," i.e., wet).

/'azraq/ ("blue") Blue in Western societies stands as the color of the sky and the sea; it is associated with coolness, water, men, youth, confidence, and innocence. On the other hand, dark blue reflects darkness, night, laziness, tranquillity, and rest. It has been the color of loyalty, obedience, respect and meditation. It represents responsibility, seriousness, and faith in one's aims (Danger, 1968). According to Max Lucher's color test, people who chose blue as number one are in need of either emotional stability and a feeling of security or need physical rest and relaxation. Those who prefer blue in general like tranquillity and quietness; they seek an organized environment free of disturbance and trouble. Life goes smoothly in its normal directions and they have good relations with people (Omar, 1982, pp. 61-63). In the Egyptian culture, /'azraq/ has been associated with purity and faith, and this explains the superstition of people who pin a blue stone on a baby for protection from envy and evil eyes. Yet /'azraq/ is a pessimistic color for Egyptians; they often call it /'axDar/ for this is their optimistic color. This is why they call /'al-9ataba-al-zar'a:/ ("the blue threshold") /'al-9ataba-al-xaDra:/ (lit., "the green threshold," which is a district in Cairo); it was named after a palace of the princess in this area and its entrance was /'azraq/ (Amin, 1953, p. 30). /'azraq/ is from the root /z/r/q/ ("to become blue"). As a noun with the definite article /'al/ it denotes the color itself as opposed to the rest of the members of the color spectrum. /'al-zurqa/ ("the blue"), a verbal noun, embraces in actual usage a wide range of colors, such as /'al-baya:D/ ("whiteness") and /'al-xuDra/ ("greenery"), as well as other connotations. /'azraq/ is mentioned only once in the Holy Quran where it denotes blindness (Qur., 12:84). Consideration of the use of/'azraq/ in CA shows that the word may refer to several colors. For example, /zurqa/ is used to denote /xuDra/ ("green") in reference to the iris of the eye, which may have meant a greenness or gray color intermixed with blackness or deep ash color (Lane, 1865). In ECA /'azraq/ is used in this sense only in the expression /HuSa:n 'azraq/ ("blue horse"), which is said of a gray horse. /'azraq/ seems to be an indefinite color in Arabic. It is associated with blue eyes, enemies, arrows, and war. It seems that /'azraq/ in CA conveys similar pessimistic impressions as /'aswad/ ("black"). Metaphorically, /'al'azraq/ ("the blue") is used to mean an arrow; /zurqun/ denotes arrows.

88

A Sociolinguistic Study on the Use of Color Terminology

/zarqa:' 'al-yama:ma/ in Arab tradition, denotes shrewd eyesight; /'alzarqa:'/ ("the blue") refers to alcoholic drinks. /zarraq/ (the intensive form) means a deceitful man or a great deceiver, and /9aduwwun 'azraq/ ("blue enemy") is a fierce enemy (al-Tha'albi, 1977). According to Lane (1865), an enemy "that is vehement in hostility" because of the /zurqa/ of the eyes is "prominent in the Greeks and Deylem, between whom and the Arabs is a confirmed enmity." According to Ibn Yaqub (1911) an enemy is /'azraq/ because his hatred is as clear as water; he is a pure enemy. In ECA, /'azraq/ in the sense of the color is used more often and is a more defined color, e.g., /9i:nu mizriqqa/ ("he has a bruised eye") and /wishshu 'izraqq/ ("his face became blue" due to lack of oxygen). Metaphorically, /9aDHma zarqa/ (lit., "blue bone") is used as an abusive epithet for a Copt (Badawi and Hinds, 1986) and /'azraq zayy in-ni:la/ (lit., "as blue as indigo") is an expression to describe how ugly the shade is. Other expressions include: /HaTalla9 il-bala l-'azraq 9ala gittitu/ ("I will beat him black and blue or will give him hell") /'il-dibba:n 1-azraq mish hayi9raf makkanu/ (lit., "blue flies will not know its place," i.e., will hide it well), /'il-9afa:rit-iz-zurq bititnaTTaT fi 9i:nu/ (lit., "the blue devils are dancing crazily in his eyes," i.e., he is furious), /hatishtaghalli fi l-'azraq/ (lit., "you will work in blue," i.e., you will play tricks on me). The color is also used to mean deceitful, as in the phrase /na:bu 'azraq/ (lit., "blue fangs"), which is an adjective for a deceiver. An Egyptian proverb, /ma ti'a:minsh li-'umm-9uyu:n zarqa/ (lit., "do not trust the blue-eyed girl"), underscores the association between blue eyes and deceivers; they are not trustworthy. All senses mentioning /'azraq/ in expressions or metaphorically show negative associations with the color. A comparison of Arabic /'azraq/ with English "blue" reveals similarities only in their use as adjectives, e.g., blue sky, blue sea. One interesting similarity is in the extended meaning derived from /z/r/q/, relating to /'inzaraq/, which also exists in English as "blue streak," informally denoting something moving fast, e.g., "they traveled like a blue streak through Italy" or something continuous, vehement, or interminable. In CA this meaning appears in /zaraqat 9aynuhu naHwi/ (lit., "his eye blued toward me"), which means his eyes turned toward me so that the white thereof appeared (Lane, 1865), and in /'inzaraqa/ ("passed through and went forth on the other side") and /mizra:qan/, which is said of a camel delayed behind the rest of the caravan because it moved too slowly (Ibn Yacub, 1911). In ECA, this sense is found in the expression /zaraq-il-musma:r fi 1-xashsab/ ("the nail passed through the wood"), /fula:n 'inzaraq/ (lit., "he became blue," i.e., he went out or in quietly without anyone noticing him). It seems that /'azraq/ ("blue") is characterized by a wide range of contradictory ambiguous usages both in Arabic and English; in Arabic, the usage of/'azraq/ as a protection from the evil eye contradicts its usage to

A Sociolinguistic Study on the Use of Color Terminology

89

designate evil spirits /'il-9afa:ri:t 'il-zurq/. In the English language, "baby blues" (i.e., eyes) are used to denote innocence side by side with the use of "blue movies" to denote pornography.

Conclusion A survey of the number of times colors are used in the books of teaching Arabic as a foreign language found no emphasis given to the category of color although it represents an important aspect of the language. As vocabulary items, colors are comparatively easy to teach for they all have the same pattern (the masculine form: /'abyaD/, /'aswad/, /'aSfar/, /'aHmar/, /'axDar/ and /'azraq/; the feminine form: /bayDa:'/, /sawda:'/, /Safra:'/, /Hamra:', /xaDra: V, /zarqa: V). The use of colors in the Arabic language is interesting. Every color implies different meanings depending on the context, and it is important for the learner to understand the social connotations of these colors. By studying colors, the student can learn more about the culture and beliefs. It would be interesting to contrast the connotations of these colors in ECA to other Arabic dialects. This study revealed, among other things, that the lines of demarcation between the various colors are not as clear in CA as they are in the colloquial language. This may be due to the phenomenon, observed by various sociolinguists, that modern times bring with them more color sophistication. As societies advance, more colors are isolated and more hues of the same color are identified by separate names.

References Abboud, P., Abdel-Messih, E., S. Altoma, et al. 1971. Modern standard Arabic: intermediate level. Ann Arbor, MI: University of Michigan, Department of Near Eastern Studies. Abd al-Baqi, Muhammad Fu'ad. 1958. Al-mu'jam al-mufahras al-'alfaz al-quran al-karim. Cairo: Matabi' al-Sha'b. Abd al-Masih, E. T., Bahij, A. F., Badawi, S. M., et al. 1976. Bahth shamil fi al-lugha al-'arabiya al-misriya. Ann Arbor, MI: University of Michigan. Abdu, D. 1967. A course in modern standard Arabic. Beirut: Khayats. . 1983. The living Arabic of Cairo. Cairo: American University in Cairo Press. Abdu, K., Hassanayn, Ahmad Tahir, Rashad, N.M., et al. 1984. Al-lugha al- 'arabiya li-ghayr al-natiqin biha. Cairo: Al-Markaz al-Qawmi li-1Buhuth al-Tarbawiya.

90

A Sociolinguistic Study on the Use of Color Terminology

Al-Araby, S. 1983. Intermediate Egyptian Arabic. Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa. . 1982. Colloquial Arabic of Cairo. Tokyo: Institute for the Study of Languages and Cultures of Asia and Africa. Al-Askari, Abu Hilal. 1973. Al-furuqfi al-lugha. Beirut: Dar al-Afaq alJadida. Ali, Y. A. 1965. The holy Kuran: text. translation and commentary. Mecca: The Muslim World League. Amin, A. 1953. Qamus al- 'adat wa al-taqalid al-misriya. Cairo: Matba'at Lajnat al-Ta'lif wa al-Tarjama wa al-Nashr. Anwar, M. S., and Cadora, F. J. 1973. Elementary modern standard Arabic II. Colombus, OH: Arabic Program, Ohio State University. Badawi, S. M., and Hinds, G. M. 1986. Dictionary of Egyptian Arabic. Beirut: Librairie du Liban. Badawi, al-Said Muhammad, and Yunis, Fathi 'Ali. 1983. Al-kitab alasasi fi ta'lim al-'arabiyya li-ghayr al-natiqin biha. Tunis: AlMunazzama al-'Arabiya li-1-Tarbiya wa al-Thaqafa wa al-'Ulum (ALECSO). Bender, M. L. 1983. Color term encoding in a special lexical domain: Sudanese Arabic skin colors. Anthropological Linguistics, 25:19-28. Berlin, B., and Kay, P. 1969. Basic color terms: their universality and evolution. Berkeley, CA: University of California Press. Branstetler, K. B. 1977. A reconstruction of proto-Polynesian color terminology. Anthropological linguistics, 19:1-26. Crawford, T. D. 1982. Defining basic color term. Anthropological linguistics, 24:338-343. Cornsweet, T. N. 1970. Visual perception. New York: Academic Press. Danger, E. P. 1968. Using colors to sell. London: Gower. Evans, R. M. 1974. The perception of color. New York: Wiley. Fowler, H. W, and Fowler, F. G. (eds.). 1961. The concise Oxford dictionary of current English. Oxford: Oxford University Press. Al-Ghazzali, Abu Hamid Imam Muhammad. (d. 1111 CE). 'Ihya' 'ulum al-din.Cairo: Dar al-Sha'b. Glucksberg, S., and Danks, J. 1975. Experimental psycholinguistics: an introduction. Hillsdale, NJ: Laurence Erlbaum. Goodman, J. S. 1959. Malayan color categories. Anthropological linguistics, 5:1-12. Halibi, Abd al-Wahid ibn 'Ali (d. 962 CE). Shajar al-durr. Edited by Muhammad Abd-al-Jawad. Cairo: Dar al-Ma'arif. Hassanein, A. T., and Kamel, M. 1984. Yalla ndardish bi-l 9arabi: let's chat in Arabic Cairo: American University in Cairo Press. Heinrich, A. C. 1972. A non-European system of color classification.

A Sociolinguistic Study on the Use of Color Terminology

91

Anthropological linguistics, 14:220—227. Hussain, Taha. 1953. Al-ayyam. Cairo: Dar al-Ma'arif. Ibn Faris al-Qazsini, Ahmad (d. 1005 CE). 1970. Mutakhayar al-alfaz. Edited by Hilal Najy. Baghdad: Matba'at al-Ma'arif. Ibn Ja'far, Abu al-Faraj Qudamah. 1979. Jawahir al-alfaz. Beirut: Dar alKitab al-'Ilmy. Ibn Manzur Muhammad ibn Makran al-Ansari. 1982. Lisan al-'arab. Cairo: Dar al-Ma'arif. Ibn Qayyim al-Jawaziyah, Muhammad ibn Abi Bakr (d. 1328 CE). 1970. Zad al-ma'ad fi hada khayr al-'ibad. Cairo: Matba'at Mustafa alHalabi. Ibn Sidah, Abu al-Hasan 'Ali ibn Isma'il (d. 1066 CE). 1898. Kitab almukhassas. Cairo: Dar al-Kitab al-Islami. Ibn Ya'qub, Muhammad al-Fayruzi al-Shirazi. 1911. Al-qamus al-muhit. Cairo: al-Matba'a al-Husayniya. Lane, E. W. 1865. An Arabic-English lexicon. Edinburgh: Williams & Norgate. Libby, W. C. 1974. Color and the structural sense. Englewood Cliffs, NJ: Prentice Hall. McNeil, L. 1972. Color and color terminology. Journal of linguistics, 8:21-23. Middle East Center for Arab Studies, Shemlan, Lebanon. 1969. A selected word list of modern literary Arabic. Beirut: Khayats. Murtada al-Zubidi, Muhammad ibn Muhammad. 1888. Sharh al-qamus al-musamma taj al- 'arus minjawahir al-qamus. Cairo: Al-Matba'a alKhayriya. Nicole, A., Steckler, W. A., and Cooper, W. E. 1980. Sex differences in color naming of unisex apparel. Anthropological linguistics, 22:373-381. Al-Numayri Abu Abd Allah al-Husayn 'Ibn 'Ali. 1976. Kitab al-mulamma'. Damascus: Matbu'at Majma al-Lugha al-'Arabiyya. The Random House dictionary of the English language (unabridged edition). (J. Stein and L. Urdang, eds.) 1956. New York: Random House. Salib, M. 1981. Spoken Arabic of Cairo. Cairo: American University in Cairo Press. Shawqy, Ahmad. 1950. Al-shawqiyat. Cairo: Matba'at al-'Istiqama. Sheppard, J. J., Jr. 1968. Human color perception, a critical study of the Experimental Foundation. New York: Elsevier. Snow, D. L. 1971. Samoan color terminology, a note on the universality and evolutionary ordering of color terms. Anthropological linguistics, 13:385-390.

92

A Sociolinguistic Study on the Use of Color Terminology

Taymur, Ahmad. 1956. Al-'amthal al-'ammiya al-misriya mashruha wamurattaba 'ala al-harf al-awwal min al-mathal. Cairo: Matabi' Dar alKitab al-'Arabi. Al-Tha'alibi, Abu Mansur Abd al-Malik Muhammad. 1923. Fiqh al-lugha wa sirr al-'arabiyya. Cairo: Matbu'at al-Sa'ada. Al-Tonsi, A. 1982. Egyptian colloquial Arabic, a structural review. Cairo: American University in Cairo Press. 'Umar, Ahmad Mukhtar. 1982. Al-lugha wa al-lawn. Kuwait: Dar alBuhuth al-'Ilmiya. Al-Yaziji, Ibrahim. 1904. Kitab nuj'at al-ra'id wa shir'at al-waridfi almutaradifwa al-mutawarid. Cairo: Tab'at al-Ma'arif.

The Canons of Aristotelian Rhetoric: Their Place in Contrastive Arabic-English Studies Nancy G. Hottel-Burkhart My purpose in this paper is to describe the five canons of classical rhetoric taught by Aristotle in order to illustrate the usefulness of these canons as a heuristic for contrastive studies in rhetoric. In describing these canons, or disciplines, of Aristotelian rhetoric, I intend neither to justify nor to challenge Aristotle's model of rhetoric: 23 centuries of subscription to it by the West and a somewhat shorter period of subscription-through-adaptation of it by the Arabic East1 have justified this model, so much so that John Frederick Reynolds, one of classical rhetoric's modern proponents has characterized it as "if not the 'most complete system...for the analysis and production of discourse,' then certainly...the most tenacious" (Reynolds, 1993, p.l; italics his). As for challenges, at least a dozen variations of new rhetorics today provide them, and anthologies already exist.2 Nor is my goal in describing the canons of Aristotelian rhetoric to suggest a framework through which Arabic or English rhetorical practices ought always to be compared. Certainly the rhetorical tradition of each language and culture is sufficient unto itself and, despite historical connections to larger traditions, neither tradition need be subsumed under another. What I hope will be shown here, rather, is quite simply that the five canons of classical rhetoric as described by Aristotle in his Rhetoric treat questions that are largely ignored in comparative rhetorical studies to date but which, if taken as new directions for research, would deepen, broaden, and put more clearly into the perspective of cultural studies as a whole the notion of "rhetoric" in the term contrastive rhetoric. One's first duty in undertaking the tasks stated above is to define rhetoric. Aristotle defined it as "an ability to see the available means of persuasion" (Rh.i 1.2.1; Aristotle, 1991, p. 36). Since his Rhetoric as a whole makes it clear that one not only sees the available means but also uses them and since over time explanatory as well as persuasive discourse has come to be included in rhetoric, the revised definition attributed to Aristotle has been understood as "the ability to choose the best from the means available in order to accomplish one's discourse goals." In either case, however, although

94

The Canons of Aristotelian Rhetoric

the definition is succinct and adequate for rhetoric within the history of a single-language culture, it requires some elaboration if it is to be useful in crosscultural studies, for the first problem that one encounters in studying the rhetorics of the world is that they differ in the purposes for which they exist. Chinese rhetoric, for example, acknowledges "an organic universe" (Foss et al., 1991, p. 296) and exists for maintaining social harmony. The rhetoric of various African societies, it is claimed, values expression as action and perforce involve the participation of the entire audience in the rhetorical act (Foss et al., 1991, p. 287). Aristotelian rhetoric has been said to be unique in focusing on conflict (Al-Wali, 1999). With such apparently divergent starting points, it is easy to lose one's bearings from the outset. Thus, for the purposes of the present discussion, I propose a definition of rhetoric that is faithful to Aristotle's original description and that at the same time captures the commonalities of it and other rhetorics: rhetoric is an intellectual tradition of practices and values associated with public, interpersonal, and verbal communication—spoken or written—and it is peculiar to the broad linguistic culture in which one encounters it. This definition places rhetoric more squarely in the center of contrastive studies and renders a more revealing starting point for a cross-cultural discussion of rhetorical disciplines. Having settled on terms, one can proceed to claims: what are the practices and values peculiar to the rhetorical tradition inherited from Aristotle? And what do they suggest for research directions in contrastive rhetoric? For Aristotle, the practices comprised five canons, or disciplines, and I will take up each in turn: (1) invention, (2) arrangement, (3) stylistics, (4) memory, and (5) delivery.

Invention Known in Greek as heuresis, in Latin as inventio, translated into English as invention, the first canon of classical rhetoric is "discovery of what is to be said" in a discourse (Corbett, 1990, p. 32). This includes lines of thought, arguments that can be used, and basic propositions from which arguments can be developed. Invention treats, in short, what knowledge can be brought to bear on a subject and in what logical form. Its goal as a discipline is finding something relevant and important to say on a chosen subject. Aristotle classified approaches to persuasion into three well-known types: logos, or the appeal to reason; pathos, or the appeal to human needs and desires; and ethos, best described as a speaker's identification with and credibility in the eyes of the audience due to shared cultural values. In invention, one sought to discover—or recover, since in fact invention is linked to memory—the best ideas or propositions on which to base any of

The Canons of Aristotelian Rhetoric

95

these three kinds appeals for the rhetorical task at hand. Aristotle taught that through a system of beginning points, called the topoi, an orator could, for any given rhetorical occasion, discover the best way to develop arguments. The common topoi of his system are summarized in Table 1, from Corbett's Classical Rhetoric for the Modern Student, and the examples accompanying this list illustrate how the topoi could be used in a systematic fashion to discover propositions to use in lines of reasoning. The orator—and over time in this tradition, the writer—learned through daily practice to employ all of these beginning points, with their relationships to logic and to the structure of knowledge, in order to discover what was to be said in a discourse. The relative strengths of the various resulting lines of argument as well as the best approach or the best two or three approaches to the discourse under the circumstances were considered before a final choice was made. Note both the flexibility of this system within a single language tradition and also the implications of it for a cross-rhetorical study. As Struever (1992) has pointed out in showing the relationship of classical rhetoric to modern historical texts, Aristotle taught "we must not expect to find a single line of inquiry which will apply universally to all" intellectual situations (p. 341). From this citation alone, it would seem that a contrastive study of rhetorics lies within the very nature of the discipline of invention. I have mentioned the relationship of invention to knowledge: those in the classical tradition differed among themselves as to the exact relationship between invention and knowledge, but all agreed that invention involved discovering (remembering, in this case) what parts of knowledge could be brought to bear on a subject in question and in what logical relation to one another other. Thus, while Aristotle himself held that rhetoric had no subject matter and others in the tradition (primary among them Cicero) held that "the perfect orator had to be conversant with many subjects" (Corbett, 1990, p. 546), the system of topoi could discover arguments since in either case all knowledge was presumed to be housed in memory. In the time of Aristotle, knowledge was considered to be finite. Thus, bringing up knowledge in the form of arguments through the topoi may seem to us, 23 centuries later in more complicated intellectual times, to have been a simpler matter than today. Yet, as modern students of rhetoric have illustrated, and as new realms of knowledge have developed over the centuries since Aristotle, the framework of the topoi for discovering arguments has accommodated the knowledge of these new disciplines. History, as Struever (1990) points out, is one such discipline. The development of the social sciences, according to Cushman and Kovacic (1994), has actually provided rhetoric with a new form of argument called "the reasoned social scientific fact," which can be of sever-

The Canons of Aristotelian Rhetoric

96

al kinds and thus reside at several points in the topoi. Likewise, the field of statistics, the science of probability, has provided rhetoric with complex mathematical versions of topics such as "similarity," "difference," "degree," "cause and effect," as well as "possible and impossible." Finally, to take this discussion more directly into the professional lives of some of us, the invention exercises used in current English composition pedagogy, which seem on the face of it to bear little resemblance to the list of Aristotle's common topics, actually provide one more instance of the flexibility of invention as a discipline over time and intellectual space: for example, an invention exercise like brainstorming reflects the current intellectual culture in which knowledge is so vast and emanates from so many starting points that only free recall can draw it all forth. In this way, brainstorming conforms to the general description of invention as discovering what knowledge can be brought to bear in discussion of a subject.

Table 1. The Common Topics* TOPICS

EXAMPLES of Invention Questions for the Subject "HERO"

Definition Genus Division

What is a hero? A hero is.... What are the types of heroes?

Comparison Similarity Difference Degree

A hero is like a . The difference between a hero and a leader is How many heroes can one nation have?

Relationship Cause and effect

What factors play a role in modern hero worship?

Antecedent and consequence

At what point does a soldier become a hero?

Contraries

Would you rather live the life of the hero or the life of the common wo/man?

Contradictions

What is the contradiction of a hero? (Yielding the thesis: A hero is not a coward.)

The Canons of Aristotelian Rhetoric Circumstance Possible and impossible

Past fact and future fact Testimony

97

What are the requisite conditions for a hero to exist? (One answer: No hero exists without a nation.) Who were the heroes of Hiroshima? (The heroes of Hiroshima were the survivors.) General question: What does (authority/maxim/ law...) bring to bear on the subject of heroes?

Authority

Answers: The ink of the scholar is worthier than the blood of the martyr (a hadith).

Testimonial

General George Patton once said that heroes....

Statistics (noninferential)

The average number of persons who visit the Vietnam War Memorial every day is....

Maxims

Cowards die many times before their deaths/ The valiant never taste of death but once. (once authority; now maxim)

Law

"The Congressional Medal of Honor shall be awarded to the person who in the course of duty in service to his/her country...."

Precedents (Examples)

Antar / Ulysses / Hadrat Ali / George Washington

*The list of common topics is taken from Corbett, 1990, p. 97. Examples are mine.

For our purpose in contrastive studies, note an important assumption in the foregoing discussion: implicit in it is the notion of what counts as an argument. The culture in which the rhetoric is embedded will assign this value. Since this is so, the relative value of what counts to any given rhetoric may vary, and this is true not only of rhetorical traditions of two different language cultures, but also over time within one rhetorical tradition. As knowledge

98

The Canons of Aristotelian Rhetoric

changes and as what is accepted as disciplines in the society develop, so the relative rhetorical value of the arguments discovered by each topos may change. To take one example from within the classical framework itself, in European education from medieval to modern times, definition was a frequently used common topic for lines of development of arguments. Arguments revolving around definition, the first on the original list of common topics, could be central to a well reasoned discourse even up into the twentieth century. Yet with the development of modern science and its predilection for empirical proofs, definitions no longer constitute a very large part of most lines of argument. Scientific arguments favor instead conclusions drawn from empirical observations, and this has influenced argumentation in general. Even within one discipline, preferences for what counts as argument occur: Struever (1990) points out that in certain views of history and historical argument, narrative has been considered weak and to be nearly no argument at all because it does not compel the reader to a single conclusion as do other forms of argument, such as the reasoned social scientific fact (p. 342). To bring this discussion of invention one final step further and at last to our subject of contrastive Arabic-English rhetorics, I turn now to a wellknown article by Johnstone (1986) on cross-cultural Muslim-European discourse, at the center of which lies the unspoken assumption of what constitutes an argument. Johnstone analyzed the interview of the Ayatollah Ruhollah Khomeini by Italian journalist Oriana Fallaci, an interview that had degenerated into a shouting match over a mismatch in the expectations of the two interlocutors as to exactly what constituted a valid argument.3 Fallaci, as characterized by Johnstone, expected the ayatollah to respond to her questions with general statements supportable by reference to observable facts about the Iran of the time. The ayatollah, schooled in a different tradition of argument, offered instead answers based on the words of God and his Prophet. Quranic verses and hadith must be cited in careful fit to the context to which they are applied: this is one of the most important conditions for their being compelling arguments, and in the eyes of many in this tradition, the ayatollah's arguments were ironclad and clearly to the point of Fallaci's challenges. Yet Fallaci saw these "quotations" as hedging and was angered. In turn, the ayatollah saw his arguments dismissed as irrelevant, which is as much as to say that the words of God and his Prophet are insignificant. He was therefore enraged, and rightly so in his view of Who the source of all knowledge and argument is. This research article is the only one I know of that comes close to broaching the basic question in Arabic-English rhetorical studies that I have posed here: what counts as an argument? What is valid in Arabic argumentation, either classical or modern? What is valid in English? What are the common points; where the divergences, and with what correspondences to the respective intellectual cultures in question?

The Canons of Aristotelian Rhetoric

99

Although the Johnstone study can be framed within the Aristotelian context—"authority" being one of the topoi—it leads one to see that crosscultural studies of invention must exceed the Aristotelian context. I would suggest that what is involved in this canon encompasses vastly more of an intellectual culture than is presently captured in the description of the discipline. It extends ultimately, in the words of Rashidi and Zhan (1996), in their analysis of "Chinese arguing in English," to "what constitutes a valid intellectual activity and how the world is constructed" (p. 389). Comparative studies that would bring these issues to bear on discussions of rhetoric would profoundly enrich Arabic-English rhetorical studies.

Arrangement Taxis in Greek, dispositio in Latin, is translated into English as arrangement. According to Quintillian, "arrangement is to oratory what generalship is to war" (Corbett, 1990, p. 278). Arrangement is the manner in which the arguments discovered in invention should be organized to achieve one's rhetorical ends. First, it should be noted that the structure of ideas discovered by invention and logic is not the same as the structure of a discourse that employs those ideas, and this difference constitutes the need for arrangement as a discrete discipline, independent of invention. Arrangement is the organization of elements of a discourse in order to place its ideas in the way most likely to move a particular audience or to achieve a particular persuasive end. This implies first an identification of the elements themselves, and the elements of classical rhetoric as Aristotle identified them are: introduction, statement of fact (not the same as a thesis, but the difference is beyond the limits of our discussion here), confirmation or proof through argument, refutation of opposing arguments thorough counterarguments, and conclusion. A listing of elements is by no means a strategy, and again it was Quintillian who focused on judgments and decisions about questions like these: •"When is an introduction necessary and when can it be dispensed with? •"What evidence or documents must be made use of and where in the discourse will they be most effective? •"Should we attempt to refute our opponents' arguments as a whole or deal with them in detail?"

He also addressed more general questions on how to order proofs and refutations in the most persuasive order and when to order proofs from strongest to weakest, and vice versa (Corbett, 1990, pp. 280-281).

100

The Canons of Aristotelian Rhetoric

As with invention, one need not look as far as two different language cultures to find contrasts. Taking a historical perspective on one rhetorical tradition can provide examples, as can considering differences between genres in one language. To take an example that combines both, in a case where a genre has developed only recently, if we compare introductions in Swales' (1990) Create A Research Space (CARS) framework for research articles, we find a very different description of introductions than that of Whately (1828), who classified all introductions into one of five types: inquisitive, paradoxical, corrective, preparatory, and narrative (Corbett, 1990, pp. 283-287). Our cross-linguistic contrastive interests in arrangement should and do go back as far as Kaplan (1966) to ask what the order of elements in a discourse of a rhetorical tradition may be, but one may also ask whether the elements to be arranged in a discourse are even the same in two traditions. Published research as well as the observations of persons who are members of two rhetorical cultures confirm that the inventory of elements and subelements involved in arrangement may vary from one rhetorical tradition to another. Take, for example, the seasonal greeting in Japanese business letters as studied by Jenkins and Hinds (1987). This subelement of the introduction is one of two necessary to every Japanese business letter. It must follow the first subelement, which is the salutation, and it must make reference to the weather and stage of nature of the time. Jenkins and Hinds' example, appropriate for the month of April, is "Facing the season of glorious cherry blossoms" (p. 337). This is distinctly required in Japanese and distinctly nonexistent in English or Arabic. To take another example closer to our focus on Arabic-English rhetorical contrasts, consider in traditional Muslim-Arabic discourse, the element in ending a text in which the writer acknowledges her/his reliance on the Creator: wa min Allah al-tawfiq or wa Allahu 'a/am.4 A cross-cultural instance of the employment of this element as the ending of a text occurs in the English essay of the Moroccan student writing on a near-drowning experience when she writes as her final paragraph: "This event made me learn that we have to rely on God and only on Him and to be more wary next time" (italics mine). This instance becomes doubly interesting when we read the comment of the student's instructor, an American professor, on the italicized portion of the ending: "People in boats come in handy too," she penned, referring to the part of the young woman's story that told of her rescue by a man in a boat. "Make the connection to avoid the cliche." To the student, this ending had been a desirable conclusion in her rhetorical strategy; to her instructor, from an Anglo-American point of view of rhetorical strategy, the student's ending was hackneyed, unnecessary, and disconnected from the rest of the text.

The Canons of Aristotelian Rhetoric

101

It is possible to cite other examples of elements that are necessary in the traditional Arabic construction of discourse of certain kinds, and one example relating to the intersection of arrangement and the third discipline of rhetoric, stylistics, will be discussed in the section on this discipline below. The examples cited so far, however, may suffice to suggest that consideration of what each rhetorical tradition includes in the inventory of elements to be arranged, not simply a description of the order of presumably universal elements, is needed in contrastive studies of rhetoric in Arabic and English.

Stylistics The third canon, in Greek lexis, "word," in Latin elocutio, "speaking the words," or putting words to arranged ideas, is referred to in English with the terms stylistics and style. In the classical framework, stylistics encompasses choice of words, kinds of sentences, lengths of sentences, marked or unusual sentences, ways of varying sentences, ways to tie words of the text together, and ways of making transitions between ideas in a text (what is referred to by Halliday and Hasan, 1975, as cohesion), as well as the traditional figures of speech, e.g., schemes and tropes. Some (e.g., Vinay and Darbelnet, 1971) would include under the term stylistics the conventions of paragraphing, which can arguably be placed in "arrangement" and, depending on the language involved, in the visual considerations—i.e., "delivery" (see the next section)—of texts as well. For contrastive purposes, stylistics could be described in its largest part as the tendencies that the structure of the language imposes on the choices that the writer of that language has when creating a text. The sentence structure of English, for example, lends itself to the possibility of asyndeton, or the omission of coordinate conjunctions, between clauses in a set of clauses that are syntactically parallel and semantically related, the stereotype of which is / came, I saw, I conquered. By contrast, in Arabic the requirement for wa in such a series mitigates against such an omission.5 The agglutinative property of Arabic, on the other hand, allows for polyptoton, or the repetition of words derived from the same root (such repetition in Arabic occurring often within one clause) for certain special effects, like emphasis or humor, while inflection-poor English cannot accomplish such repetition without making the writer sound dull-witted. The few exceptions to this, like Aldous Huxley's "Few are chosen because few choose to be chosen," are memorable because of other schemes in tandem with repetition of a single root and not because of such repetition alone. Note as well in the Huxley example that the more memorable repetition of roots occurs across clause boundaries, not within them. To carry out serious comparative studies of stylistics, intimate knowl-

102

The Canons of Aristotelian Rhetoric

edge of two languages and their stylistics traditions is necessary. Certainly for Arabic alone the canon in its full detail of lexical choice, sentence type, meaning of and preferences for types, as well as figures of speech, has been described. Likewise, for English the full repertory of stylistic means is commonplace to those who study rhetoric. But the contrastive work that discusses how it is that the structures of the respective languages result in the English propensity for asyndeton and an Arabic impossibility for it and how it is that Arabic can repeat roots within one clause where English does so only rarely, to my knowledge, does not exist. I have characterized stylistics in its largest part as the tendencies that the structure of the language imposes on the lexical and syntactic choices. I should like now to note two exceptions to this generalization—one of which lies in the design of classical rhetoric itself, the other of which has grown out of a discussion at this conference. The classical exception to this characterization concerns certain kinds figures of speech called tropes, such as metaphors, similes, oxymoron, personification, and hyperbole, which hearken not so much from language structure as from broader cultural notions, from world view and how the world is organized. It is commonplace to note—although Lakoff and Johnson (1981) have done so in anything but a commonplace way—that the expressions of metaphors differ in culturally explainable ways. The relative presence or absence of the use of a certain trope in the language—the frequency with which a trope is used and the judgment assigned to its use—may also be due to a value in the rhetorical culture itself. Take, for example litotes (understatement) in English. It is considered to be a particularly refined form of humor and irony in English. In Arabic it does not merit expression, and when examples translated from English occur they are not particularly well understood. The humor in the reference by a member of the Royal Family of England to his life as work in the family firm is missed, as is the irony in referring to a $5 million house in Malibu as a hut on the beach. Puns, or repetitions of words in two different senses, on the other hand, while not valued in English and in fact mostly disparaged as low forms of humor, occur frequently in Arabic, are valued as subtle and sophisticated plays with language, and—especially as used in certain political discourses—are not always considered language play for the sake of mere humor. Observations like the foregoing can be made easily enough if by those who know enough of both the Arabic and the English systems of stylistics, but what values of the culture underlie them? What realities of the "universe of discourse" (Lefevere, 1992 a,b) of each culture are reflected in these tendencies and preferences? A contrastive Arabic-English version of Metaphors [and other figures of speech] We Live By seems to be called for in fact.

The Canons of Aristotelian Rhetoric

103

The second exception to the generalization that stylistics concerns structural tendencies and their repercussions for word choice and sentence type relates to the scope of stylistic choices. It suggests that in some languages, what we consider lexical and syntactic tendencies may in fact relate not so much to stylistics as to the textually broader canon of arrangement. The impetus for proposing this exception is a study by Aziz Khalil, the object of which was the narrative structure of high school and university students writing in Arabic and in English. Employing the framework for narratives of Martin and Rotherby (1986) to analyze Arabic and English essays of Palestinian high school and university students, Khalil found that when writing in Arabic, the younger writers in his study failed to construct narratives in the usual four expected segments (orientation, complication, resolution, coda); while they constructed the resolution and coda of their stories in accordance with the assumed universal pattern of narrative arrangement, they conflated the orientation and complication. What was noteworthy in this was not the conflation itself but the sentence arrangement within the conflated section. According to Khalil, the section consisted of repetitions of the following pattern: sentence of orientation-wa-sentence of complication; sentence of orientation -wa-sentence of complication. These young writers had not simply conflated orientation and complication: they had woven them together in a clearly discernible pattern. In effect, they were making a choice in arrangement of texts—two lines of development together in one place rather than each separately its entirety and in canonical sequence. The fact that only the students with little exposure to English texts wrote in this way and the fact that they wrote this way only in Arabic point strongly to the influence of Arabic text structure and text-structuring processes on this choice of arrangement. Furthermore, it shows clearly that the particle wa is not simply a coordinate conjunction for sentence-level choices and may be instead a crucial element in the production of a particular pattern of arrangement. This seemingly small bit of information about Arabic writers' texts provides, I believe, an extraordinary opening to our discussions of crosscultural Arabic-English rhetorical studies. The statements most often made in second-language studies about Arabic speakers writing in English are that Arabic writers use abundant coordinate, parallel sentence structure (Ostler, 1987) and that they repeat one idea at the same level of generality throughout a text (Kaplan, 1966). The above-cited detail of Khalil's study suggests the possibility that these widely referred-to Arabic abundance of coordination and parallelism are not merely governed by sentence-level decisions or even by a transfer of native-language syntactic preference. It suggests that, at least in the case of narrative, the use of wa and conjoined parallel sentences may be but two details in complex pat-

104

The Canons of Aristotelian Rhetoric

tern of arrangement, a pattern that because of its very complexity needs parallel structures and the "coordinate conjunction" to demarcate two different lines of development that are interwoven in one time-ordered train of prose.6 It may point to a use of wa and parallel structures whose text significance we have missed in the English-centered view of discourse, because we have never imagined that arrangement could work this way. Perhaps even lexically varied semantic repetitions and writing at one level of generality are related to the construction of prose in this pattern. I believe that this discovery by Khalil will allow us to begin to say something more significant about Arabic style and rhetoric than what their surface features are.

Memory and Delivery The last two canons of classical rhetoric—memory and delivery—have been mostly neglected in recent centuries by nearly all scholars of classical rhetoric, with the notable exception of those trained in classics, under the assumption that once writing became widespread, memory and delivery—both based originally on an oral model of rhetoric—were no longer relevant to rhetorical studies. That this assumption was erroneous has been recently shown by Reynolds (1993b) and his colleagues in modern English composition studies, and the recognition of these canons as indeed relevant to the production of written texts has engendered a renewal of research interest in them. Considering these canons is in fact reasonable: one must employ memory in order to construct a piece of written discourse, and as for delivery, as Connors (1993) has thoroughly illustrated, it is not only the ear that seeks to be pleased: the eye also seeks pleasure, and thus aesthetic considerations of print are centrally relevant to rhetorical studies. Yet note that almost any question we may pose about research on delivery and on memory as relates to rhetoric will apply equally to descriptive research on one single rhetoric and not only to contrastive rhetoric research, since almost any question posed is yet unbroached, not to mention unanswered. Memory. Greek mneme, Latin memoria. Memory is the place where knowledge resides, and "knowledge" means all that is known from one's culture. It includes the topoi used in invention; in fact, according to Aristotle, memory is "the key to invention" (Reynolds, 1993a, p. 6, citing Yates, 1966). It also includes the patterns of arrangements that text linguists call formal schemata. It includes familiarity with the styles of writers to whom one has been repeatedly exposed, and it includes (to look ahead to delivery) knowing the occasions for which a text must be written by hand and when rendered in print.

The Canons of Aristotelian Rhetoric

105

Memory includes the storing and recalling of single facts as well as the reciting of long texts. It consists not of a single skill nor even of several skills of the same type. Memory as treated in the classical tradition of rhetoric in fact contained all of the following faculties of memory, according to Francis Yates (1966) as cited in Reynolds (1993a): improving the memory, imprinting the memory, memorizing in order, making memorable, holding in memory, retrieving from memory, delivering from memory, and preserving in memory (p. 5). Imprinting on the memory and retrieving from memory are related to schema theory. Modern psychological studies of memory which conclude that storage of events differs from storage of facts are treating holding in memory. Gardner (1983), in recognizing the ability to recite the entire Quran as a kind of intelligence distinct from all others, is perhaps giving us an example of a phenomenon that encompasses the above memory skills in their entirety. However, most questions of concern to us in rhetoric today remain to be researched. Are schemata for arrangement—for example, the formal schema for writing a business letter—held and recalled in the same way as an event or as facts? Is there a phonological element involved in the skill of making memorable in the Arabic writers' composing processes? This question has been suggested by some of the students at my university who write in both languages and say that when they review what they have written in Arabic for editing, "we literally read it aloud...it must sound right," by which point, according to them, they are referring in part to rhythm.7 What skills are involved in memorizing and reciting of long texts like the Quran or the Psalms as recited in whole every 24 hours by monks in the Western Desert of Egypt? In either tradition, what have been the methods of teaching memory skills? Aristotle and his contemporaries devised an entire system for training the memory in all its composite skills, and if we read Yates (1966), we easily see that what is encompassed today by mnemonics covers only a fraction of the techniques and kinds of techniques in the original Aristotelian system. Was the oldest Arabic system, or are remnants of it today, anything like Aristotle's? Were the same faculties assumed to be involved? Were the techniques similar? In reference to skills used in memorizing and reciting the Quran, does mastering these have an impact on composition processes in Arabic? This canon is more than asking for research: it is begging for it. Delivery. We come finally to delivery, the aesthetics of the medium in which a message is presented. Called hypokrisis in Greek and pronunciatio in Latin, delivery is the aesthetics of the finished product—the performance of a speech, the final version of a written text. In an aurally received medium, delivery traditionally has to do with voice qualities—

106

The Canons of Aristotelian Rhetoric

pitch, volume, intensity, rate of speaking, pauses, intonation—and with patterns and variations in all of these. It has to do with articulation, with eye contact, with the distance between the speaker and his/her audience, with body stance, with gestures, with naturalness and readiness, whether a text is memorized as in classical times or read aloud as in modern ones. In what Sa'adeddin (1989) has called the "visual" medium, delivery involves the material on which writing or printing takes place—rice, cotton rag, or parchment—its size and thickness and weight, its color, and as well involves the instrument with which ink or another substance is impressed upon it. It involves the ink or other substance itself: is it black, indigo, blue, red, or green, and for what rhetorical occasions? It involves viscosity of ink: there was a time in the United States, for example, when ballpoint pens were not acceptable for school essays, and only fountain pens would do. The aesthetics of the visual mode involves choices of typeface on a typewriter and of font on a word processor; it relates to thickness of typefaces and the uniformity and spacing of letters in a given font. It involves serifs and the effect which they have on readability of a text (Connors, 1993, pp. 68-69). It involves when to write a letter or an essay by hand, when to type or print it, or when to send it by e-mail. As for contrastive studies of delivery, I know of anecdotal data only. In Morocco, any letter of application for a job, whether in Arabic or in French, is written by hand, whereas the curriculum vitae for the same application is always printed. Students in Moroccan schools must submit written work in black or blue ink; teachers correct in red. Such anecdotes are crucial at this stage, and what is important in them is not whether black ink or blue is used but what value is assigned to the choice. Despite an absence of contrastive studies of this canon of rhetoric per se, one scholar has made an important contribution to it by the turn of a phrase. This is Sa'adeddin (1989), reported in Connor (1996), who as mentioned above has used the term "visual language" to refer to the language used in written texts. Although Sa'adeddin makes use of the term to distinguish the syntactic, lexical, and semantic features of one mode of Arabic from another rather than to discuss the canon of delivery, he nevertheless has provided the canon with a term that gives the aesthetics of the written text its due. The fact that this term, though reminiscent of the title of the journal Visible Language, comes from an Arab scholar rather than from an American or an English one is noteworthy in a discussion of contrastive rhetoric and corroborates all anecdotal evidence I have found for an aesthetics in written Arabic rhetoric which seems to differ in degree if not in kind from its counterpart in English. At any rate, the values that such differences embody should interest us in contrastive rhetoric.

The Canons of Aristotelian Rhetoric

107

Conclusion What I hope I have brought to light in this paper is that it is not at the level of narrow linguistic contrasts but rather in the purview of rhetoric as a whole that profound differences between Arabic and English rhetorics are to be found. Whether in questioning what counts as an argument under the canon of invention, in considering what elements of arrangement may not be common to all discourses, in pointing to uncharted aspects of stylistics in Arabic-English studies, or in proposing a place for the forgotten canons of memory and delivery in contrastive rhetorical studies, examination of Arabic-English rhetorical studies in the Aristotelian framework with its five canons can deepen our understanding of what constructs and values may indeed be involved in a culture's rhetoric. In doing so, it not only uncovers new research fields ripe for investigation but also gives us the opportunity to attain a truly rhetorical perspective in contrastive rhetorical studies.

Notes 1 The generally accepted European view of the Arab adoption of Aristotelian rhetoric is not unchallenged. See al-Wali (1990) for the most radically stated view that Arabic rhetoric has always been its own independent (at least Eastern) tradition. 2 See, for example, the second edition of Contemporary Perspectives on Rhetoric, by Sonya Foss, Karen Foss, and Robert Trapp (1991). 3 The choice of this discourse as being relevant to Arabic rhetoric has been questioned, first on the grounds that the language of the ayatollah was Persian, not Arabic, and secondly from the claim that the reference to rhetorical arguments from Muslim discourse are not central to Arabic rhetoric. I will answer each objection in turn. First, to the objection that Persian and not Arabic is the language spoken by the ayatollah, it must be understood that the original language of the current religion of Persia has always been Arabic and that as a religious leader the ayatollah most surely knew intimately the language and those of its rhetorical traditions related to religion. Nevertheless, if one insists on reference to the secular language of the rhetorics involved (and here, it might have been added, that Italian, not English, is the language of the secular tradition of Ms. Fallaci), I restate that rhetoric, as defined at the beginning of this text, is an intellectual tradition associated with, but not limited to the linguistic culture in which it is found. Thus, the language in which a rhetoric is practiced need not be the only important language of that particular rhetorical tradition, nor need it be even the most distinctive element in that

108

The Canons of Aristotelian Rhetoric

rhetoric; nor is a rhetoric subscribed to limited to the language in which its practices and values are originally found. If the opposites of the above statements were true, English would not today have a 2,000-year-old rhetorical tradition in common with that of Europe and ancient Greece. At the same time, one of the most widely understood facts about the Arab intellectual tradition is that it and the Persian tradition have a long and multifaceted history of exchange and mutual influence and borrowing, and rhetoric is known to be among the disciplines that has benefited from this common heritage (see, among others, al-Wali, 1990). Not to recognize this common heritage is to misunderstand the multilinguistic scope that a rhetoric as an intellectual tradition can have. As to the point that the ayatollah's choice of Muslim argumentation should be removed from consideration in discussions of general Arabic rhetoric, I would answer, first, that the intellectual tradition of rhetoric, involving as it does both practices and values, is pervaded by all values of the society, including especially the religious ones. (See Garret, 1991, passim, for brief discussion of religious values and their impact on a rhetorical tradition as far removed from our focus as China.) In addition, in the rhetorical traditions of both English and Arabic, religious discourse has exemplified the best of a culture's rhetoric rather than being peripheral to it. In the English rhetorical tradition, many of the best orators and writers of their times have been ministers of the Church (e.g., Cardinal Newman, Martin Luther King) and indeed a large number of professors of rhetoric in the Scottish/American tradition in particular have been men of the cloth. (See especially pp. 563-570 of Corbett's 1990, account of 18th-century English rhetoric.) Neither the presence of religious values in the rhetoric of English nor the examples of religious leaders as English rhetoricians have ever been brought forth as reasons to question the appropriateness of persons such as Dr. King as representatives of their language's rhetoric; I find it odd that the ayatollah or his Muslim discourse and arguments should be so challenged. Thirdly, while it is true that religious discourse is not the only or even the main type of discourse historically treated in Arabic rhetorical studies, the practices of the religious tradition as culturally acquired values are nevertheless of interest—or should be—in contrastive rhetorical studies. In any case, the discourse contrast in question has been first described as Muslim-European rather than as Arabic-English in recognition of the rhetorical traditions involved as cultural and intellectual phenomena and not as narrowly defined linguistic ones. That this recognition is missed in contrastive studies is in fact the thesis from which the present chapter proceeds, and the objections raised here to my choice of example are merely cases in point.

The Canons of Aristotelian Rhetoric

109

4 The use of this example is not to suggest that all writing or rhetoric in Arabic is informed by Muslim ideals. It is, nevertheless, the case that the overwhelming majority of native Arabic writers are Muslim and are thus, depending on the educational system in which they have learned to write and the writing tasks in which they are engaged, influenced by such cultural practices. See also note 3 above. 5 I am grateful to my colleague Abdeslam Khalafi of al-Akhawayn University in Ifrane for his assistance in confirming my nonnative intuitions about Arabic stylistics. 6 Basil Hatim (1991) has argued as much for certain Arabic text types in his discussion of "through-argument." 7 I am grateful to Alaaeddine El Koussaimi and Hind Benrhanem of the spring, 1999, graduate course "Contrastive Rhetoric and Translation Theory," at al-Akahawayn University in Ifrane, for their comments on their composing processes in Arabic.

References Aristotle. 1991. On rhetoric: a theory of civic discourse. (G. A. Kennedy, trans.) New York: Oxford University Press. Connor, U. 1996. Contrastive rhetoric: cross-cultural aspects of second language writing. Cambridge: Cambridge University Press. Connors, R. J. 1993. Actio: A rhetoric of written delivery (iteration two). In J. F. Reynolds (ed.), Rhetorical memory and delivery: classical concepts for contemporary composition and communication. Hillsdale, NJ: Lawrence Erlbaum, pp. 65-77. Corbett, E. P. J. 1990. Classical rhetoric for the modern student. New York: Oxford University Press. Cushman, D. P., and Kovacic, B.1994. The rhetoric of reasoned social scientific fact. Argumentation, 8:33-47. Foss, S. K., Foss, K. A., and Trapp, R. 1991. Contemporary perspectives on rhetoric. Prospect Heights, IL: Waveland Press. Gardner, H. 1983. Frames of mind: the theory of multiple intelligences. New York: Basic Books. Garrett, M. 1991. The Asian challenge. In Foss et al. (1991), pp. 295-306 and 311-314. Halliday, M. A. K., and Hasan, R. 1975. Cohesion in English. London: Longmans. Hatim, B. 1991. The pragmatics of argumentation in Arabic: the rise and fall of a text type. Text, 11:189-199. Jenkins, S., and Hinds, J. 1987. Business letter writing: English, French, and Japanese. TESOL quarterly, 21:327-349.

110

The Canons of Aristotelian Rhetoric

Johnstone, B. 1986. Arguments with Khomeni: rhetorical situation and persuasive style in cross-cultural perspective. Text, 6:171-187. Kaplan, R. 1966. Cultural thought patterns in intercultural education. Language learning, 16:1-20. Khalil, Aziz. 1999. Analysis of superstructure in Arabic and English narrative writing produced by Palestinian EFL students. Paper at the First International Conference on Contrastive Rhetoric, held at the American University in Cairo, February 19-20, 1999. Lakoff, G., and Johnson, M. 1981. Metaphors we live by. Chicago: University of Chicago Press. Lefevere, A. 1992a. Translation, rewriting, and the manipulation of literary fame. London: Routledge. . (ed.). 1992b. Translation/history/culture: a sourcebook. London: Routledge. Ostler, S. E. 1987. English in parallel: a comparison of English and Arabic prose. In U. Connor and R. Kaplan, Writing across languages: analysis of L2 text. Reading, MA: Addison-Wesley, pp. 169-185. Rashidi, L. S., and Zhan X. 1996. Chinese arguing in English. In B. Hoffer, The 22nd LACUS forum. Chapel Hill, NC: LACUS, pp. 389-395. Reynolds, J. F. 1993a. Memory issues in composition studies. In J. F. Reynolds, Rhetorical memory and delivery, pp. 1-15. . (ed.). 1993b. Rhetorical memory and delivery: classical concepts for contemporary composition and communication. Hillsdale, NJ: Lawrence Erlbaum. Sa'adeddin, M. 1989. Text development and Arabic-English negative interference. Applied linguistics, 10:36-51. Schaub, M. 1999. Contrastive rhetoric in historical context. Paper at the First International Conference on Contrastive Rhetoric, held at the American University in Cairo, February 19-20, 1999. Struever, N. 1992. Classical rhetorical topics and contemporary historical discourse. Argumentation, 6:337-347. Swales, J. M. 1990. Genre analysis: English in academic and research settings. Cambridge: Cambridge University Press. Vinay, J.-P, and Darbelnet, J. 1971. Stylistique comparee dufranqais et de l'anglais: Methode de traduction. Paris: Didier. Al-Wali, Muhammad. 1990. Al-sura al-ashi'riafi al-khitab al-bala'iy wa al-nathri. Casablanca: Al-Markaz al-Thaqafi al-'Arabi. . 1999. Greek rhetoric and Arabic rhetoric. Lecture, School of Humanities and Social Sciences, al-Akhawayn University in Ifrane, May 12. Yates, F. 1966. The art of memory. Chicago: University of Chicago Press.

Metadiscourse in English and Arabic Argumentative Writing: A Cross-Linguistic Study of Texts Written by American and Egyptian University Students Maha El-Seidi Writers of argumentative texts use metadiscourse expressions to demonstrate their assessment of and attitude toward the information and views they present about the subject matter. As defined by Vande Kopple (1985), metadiscourse devices do not add anything to the propositional content, but represent "discourse about discourse or communication about communication" (p. 83). Writers on metadiscourse have consistently emphasized that metadiscourse acts signal the writer's involvement in the text. Brandt (1990) lists metadiscourse as one of the features of involvement interacting with features of literacy in essayist writing. Crismore et al. (1993) explicate that by using metadiscourse acts, "writers project themselves into text [to explain their] attitude toward the content and the readers. In other words, writers convey their personality, credibility, considerateness of the readers, and relationship to the subject matter" (pp. 39-4-0). Beauvais (1989) proposes that metadiscourse functions to indicate the "writer's communicative intent" in presenting the propositional content and how it is organized into a "purposeful text" (p. 17). A variety of taxonomies for metadiscourse classes has been proposed. Of these, Vande Kopple's (1985) catalog, which is primarily based on Williams (1981), is the most widely used. Two classes of metadiscourse, namely, validity markers and attitude markers, particularly function to signal the writer's credibility and attitude to the propositional content. Validity markers are used to indicate "how [writers] assess the probability or truth of the propositional content [they] express and to show how committed [they] are to that assessment" (Vande Kopple, 1985, p. 84). Validity markers include the subcategories of hedges and emphatics. Hedges are expressions that indicate the probable truth of some generic statement. By using hedges, writers "signal a tentative or cautious assessment of referential information"

112

Metadiscourse in English and Arabic Argumentative Writing

(Crismore and Vande Kopple, 1988, p. 185). Besides, hedges, according to Chafe and Danielwicz (1987, p. 109), help writers "escape blame for instances which fail to correspond to [the] generalization." Examples of hedges in English are perhaps, generally, possibly, and it seems (to me). These appear to have the following equivalents in Arabic: rubbama ("perhaps"); 'ammatan ("generally"); min al-mumkini ("it is possible"); and yabdu li ("it seems to me"), respectively. Emphatics, on the other hand, signal the writer's strong commitment to the truth of the propositional content, e. g., "undoubtedly," bi-la shakkin; "of course," bial-tab'i; and "certainly," min al-mu'akadi. An Arabic emphatic device that may not have an obvious English equivalent is the particle 'inna, (translated here, following Johnstone 1991, as indeed), which occurs initially in nominal clauses. According to Hasan (1995, p. 631), the emphatic meaning of 'inna (and some types of 'anna) consists in "emphasizing the ascription, that is, the ascription of the predicate [khabar] to the subject [mubtada'], and eliminating doubt or denial of it." Concerning the appropriate use of these two particles, Hasan points out that "it is rhetorically wrong to use them unless the predicate is subject to doubt or denial. And emphasis by them indicates that their predicate is certain and unquestionable for the speaker." In identifying the various uses of 'anna, Hasan points out that this particle functions emphatically only when it occurs with words that denote "positive cognizance and conviction" or ascertained "fear and caution" (p. 644). On the other hand, he indicates that 'anna can occur with words that denote probability, in which case it may not be used emphatically. Consequently, in the present study only clause-initial cases of 'inna are counted as emphatics. Expressions containing 'anna are assigned the emphatic function only if they contain words denoting certainty. (As will be shown by some of the examples below, 'anna may occur in hedging expressions.) As for attitude markers, they reveal the writer's mental attitude toward the informational material. They include such expressions as "regrettably," ma'a al-'asaf, "surprisingly," min al-mudhishi; and "it is wrong," min al-khata'i. Although scholars working on written language have been increasingly concerned with investigating metadiscourse, cross-linguistic examination of its use has not received adequate attention. Chief among the few studies in this respect are Crismore et al. (1993) and Mauranen (1993). Crismore et al. investigated culture and gender variation in the use of various classes of metadiscourse in the native persuasive writing of American and Finnish university students. They observed that students in both groups used all the categories and subcategories of metadiscourse listed in their catalog. Their findings, however, indicate cultural and gender variation in the frequency and types of metadiscourse. Mauranen (1993) also

Metadiscourse in English and Arabic Argumentative Writing

113

observed cultural differences in the amount and variety of metadiscourse in the English academic writing produced by Anglo-American and Finnish academics, with the former group showing the greater amount and larger variety of metadiscourse expressions. Concerning English-Arabic contrastive rhetoric studies, the use of metadiscourse, particularly validity markers and attitude markers, has hardly received adequate attention. As pointed out above, these particular categories are the ones that directly signal aspects of the writer's involvement in and interaction with the conventions of essayist prose for producing written argumentation. The present study investigates the use of validity markers and attitude markers in English and Arabic argumentative writing. The research is cross-linguistic on more than one level. It is concerned with comparing the use of these two classes of metadiscourse in native English and native Arabic argumentative essays. It also explores the variation in their use across the written argumentation by native and nonnative speakers in each language. An attempt is made to interpret the differences detected between LI and L2 essays in each language. Such differences have been usually ascribed to negative transfer of L1 discourse modes to L2 writing (Kaplan, 1966, was the seminal work in this respect; other works worthy of mention include Kaplan, 1987; Grabe and Kaplan, 1989; Clyne, 1987 and 1991). Some researchers, however, have proposed alternative explanations for these differences. Mohan and Lo (1985), Stalker and Stalker (1989), and Kaplan (1997) point out the effect of instructional and/or developmental factors in this respect. Building on this, it is proposed here that the general level of proficiency and the amount of experience with the reading and writing in L2 may account for the discourse-level problems identified in texts produced by L2 writers. The research addresses the following questions: 1. What are the similarities and differences in the use of metadiscourse in native English and native Arabic argumentative essays? 2. In what ways are the texts of native speakers different from (or similar to) those written by nonnative speakers concerning the use of metadiscourse? 3. Which of the differences detected are due to L1 transfer and which call for alternative interpretations?

Description of the Study The data base for this study consists of 160 argumentative essays, 80 were written in English and 80 in Arabic. The English and Arabic groups consist of two sets each. The first set (40 essays) were written by NSs of each language. The second set (40 essays) were produced by NNSs of the language. Care was taken to select participants with comparable educational background. The native English data were elicited from NSs of American

Metadiscourse in English and Arabic Argumentative Writing

114

English who were senior and graduate university students. The L2 English essays were written by Egyptian fourth-year English majors. Fourth-year Egyptian university students produced the Arabic L1 texts. As for the Arabic L2 essays, they were written by upper-intermediate and advanced American students (mostly graduates) of Arabic as a Foreign Language (AFL). Most of the participants wrote the required essays as classroom tasks. Only a few of the graduate AFL students were paid to complete the tasks. Instructions for completing the writing tasks were provided in writing to each subject. All the participants were asked to mention their major and gender. Additionally, the AFL students were asked to provide the number of years of studying Arabic. The essays were written on four topics. Since the two groups of participants belonged to different cultures, it was necessary to avoid culturespecific topics, i.e., topics that refer to issues that may be interesting to or perceived by one cultural group rather than another. The following topics were thought to be interesting to participants from the two cultural groups: •Arab-American relations •The major benefits as well as the challenges involved in learning a foreign language •Equality between men and women •A significant change in the writer's country during the past 10 years1 (For a more detailed description of the subjects and procedures, see ElSeidi, 1996.)

Findings Native English vs. native Arabic texts The findings related to the frequency of validity markers (hedges and emphatics) and attitude markers in the two sets of essays written by native speakers are shown in Table 1. Table 1. Categories and frequencies of metadiscourse in native English and native Arabic texts. Native English Hedges Emphatics Attitude markers TOTAL

No. 66 18 13 97

%* 11.7 3.0 2.2 16.9

Native Arabic No. 71 102 18 191

%* 8.7 12.5 2.2 23.4

Metadiscourse in English and Arabic Argumentative Writing

115

These findings indicate that the overall frequency of metadiscourse in the Arabic L1 essays is higher than it is in their English counterparts. Emphatics score higher in the Arabic L1 essays, accounting for the overall larger amount of metadiscourse in this group. On the other hand, hedges are used more frequently in the English L1 essays than in the Arabic L1 texts. Finally, attitude markers are used with an equal low frequency in these two sets. In order to avoid repetition, the findings concerning the forms, functions, and contexts of the (sub)-classes of validity markers and attitude markers as demonstrated in the two sets written by native speakers are presented in the following two sections, devoted to native English (NE) vs. nonnative English (NNE) and native Arabic (NA) vs. nonnative Arabic (NNA) texts, respectively. Native English vs. nonnative English texts Table 2 provides the findings for the frequencies of validity markers and attitude markers in English L1 and L2 essays. Table 2. Categories and frequencies of metadiscourse in native and nonnative English texts. Native English

Nonnative English

No.

%*

No.

%*

Hedges

66

11.7

60

7.8

Emphatics Attitude markers

18 13

3.0 2.2

41 11

5.3 1.4

TOTAL

97

16.9

112

14.5

* Percentage is to the total number of clauses. The table shows that the overall frequency of metadiscourse acts in the NE essays is higher than that in the NNE ones. NE writers use more hedges, while NNE writers use more emphatics. However, the difference in the frequency of hedges is not as significant as that of emphatics. Attitude markers score almost similar low frequency in the two sets. The following subsections demonstrate the forms, subcategories and contexts of the individual classes of metadiscourse. (The classification of hedges, emphatics and attitude markers is based mainly on the inventories of Vande Kopple, 1985, and Crismore et al., 1993). Hedges. The following formal devices are used to code hedges in the English data: •Verbs of cognition with a first-person singular subject, e.g., I believe, I think, I feel

116

Metadiscourse in English and Arabic Argumentative Writing •Modal auxiliaries, e.g., may, might •Adverbs of epistemic modality, e.g., probably, perhaps, possibly •Adverbs of frequency, e.g., sometimes, generally •Expressions such as in my opinion, it seems (to me) •Predications such as we can see, we can notice (only in the NNE data)

The underlying function of hedges, as pointed out above, is to indicate a tentative or cautious evaluation of the truth of the informational content. This underlying function of hedges seems to be related to the types of contexts in which they operate in the argumentative essays. In the present study, hedging devices are identified almost exclusively in clauses representing "key rhetorical points" (Barton, 1995), i.e., thesis statements, main arguments, and closing arguments. This is not surprising since it is these parts of the argumentative essay that convey the writer's interpretation and evaluation of the argued state of affairs. On the other hand, writers will not normally hedge the support points since they present them as observed facts to support their interpretations. Within these underlying contexts, various hedging devices are observed to perform subtly different functions, all of which again are drawn from the underlying function. The first of these subfunctions is to indicate the writer's direct involvement in the argumentation; in other words, hedges help the writer project him/herself into the text. This function is performed by verbs of cognition with an 7 subject and in my opinion expressions. Prefacing the argument with an / believe or in my view, the writer claims responsibility for the truth of its informational content, while delimiting its universality, or making it clear that it is not meant as a universal generalization. This is the most commonly used hedging technique in the two sets, accounting for 39% of the hedges in the NE texts and 53% in the NNE ones. In the essays of the two sets, the / believe and similar predications are detected mostly in thesis statements, the first main argument for the thesis, and closing arguments. NE writers use various verbs of cognition such as feel, see and think to form these hedges. On the other hand, NNE writers, seem to rely solely on / think in this respect. The following thesis statements from a NE and NNE texts, respectively, are hedged by this device (in all the examples emphasis is added to highlight the constructs being exemplified): (1) Though it is a challenge, / believe that learning a foreign language can be a very worthwhile and beneficial experience. (2) 7 think that equality between men and women has been achieved long ago but full equality will take long time to be achieved.2 The use of the first-person singular in these contexts seems to serve to limit the universality of the generalization while indicating the writer's direct responsibility for its truth. Writers may alternatively choose to alleviate the force of the argument, but avoid direct claim to its truth value.

Metadiscourse in English and Arabic Argumentative Writing

117

The editorial we seems to be one of the devices of this self-effacement. Predications like we can say are used only by NNE writers for this purpose. Out of the 60 hedges in these essays, 11 cases fall under this subcategory, occurring mostly in closing arguments, e.g., (3) Finally, we can say that the sound and real democracy which began years ago began to gain results. The second subfunction served by hedges is to shed necessary doubt on the truth-value of the argument, providing room for disagreement (Salager-Meyer, 1994). The modal auxiliaries may, might, adverbs of epistemic modality, such as probably and perhaps, and the clauses it seems and it appears serve as signs of the writers' awareness of the tentativeness of their interpretations. This type of hedge is demonstrated much more frequently in the essays of the NE writers (representing 30% of the total number of hedges vs. 10% in the NNE essays). These "toning down" hedging devices are detected mostly in clauses expressing main arguments and, less frequently, closing arguments. The following examples illustrate these devices as used by NE and NNE writers, respectively: (4) Perhaps all of these relations will change due to the Arab-Israeli peace efforts, rise of fundamentalism and further disintegration of the former Soviet Union. (5) ...but may be it [the social role of women] is greater than that of man. An interesting subgroup of this last function of hedging devices, present only in NE essays, includes adverbials like most probably and more than likely. These seem to convey a stronger endorsement of the argument on the part of the writer than would be indicated by their unqualified counterparts, though not strong enough to be expressed by an unhedged assertion. The following instance illustrates: (6) Whether people direct their frustration and opposition is uncertain, but people would most probably draw connections with their situations and US-Arab relations. With the third class of hedging devices writers communicate necessary limitations on the range of applicability of the informational content of their arguments. Hedges serving this purpose mostly take the form of adverbs of frequency such as usually, sometimes, often. Also adverbs and phrases that denote that the argument content is not always true but primarily, largely, or for the most part true belong to the same class. The difference in frequency between the two groups for this subcategory is insignificant. (Devices of this sort account for 11.6% of the hedges in the NE group and 10.6% in the NNE set.) Similar to the other hedging devices, this class occurs almost exclusively in clauses denoting key rhetorical functions. The following two the-

118

Metadiscourse in English and Arabic Argumentative Writing

sis statements illustrate this device in NE and NNE essays, respectively: (7) For the most part these bonds are a relatively recent phenomenon resulting from the rise of American hegemony in the wake of the second world war. (8) The Arab-American relations are based, generally, on the mutual interests of the two sides. Emphatics. Emphatics are functionally the counterparts of hedges since they indicate the writer's strong commitment to the truth-value of the propositional content of the argument. On the other hand, the underlying context of emphatics is the same as that of hedges, i.e., key rhetorical points of the argumentation. The two sets of essays show rather different types of the formal devices acting as emphatics. In the NE data emphatics mostly take the form of adverbs such as indeed, really, clearly, and of course (when it means naturally or certainly). In the NNE, although these forms can be identified, emphatics are mainly coded by predications like nobody can deny that, it is taken for granted that, and we know that. Despite the difference in the preferred forms of emphatics in the NNE essays, they occur in similar contexts to those in the NE essays. The following two examples demonstrate how NE and NNE writers, respectively, show strong commitment to the truth of the claimed state of affairs: (9) Thus, mutual feelings of superiority—cultural and religious from the Arab perspective versus political and technological from the American perspective—clearly shape the relations between Arabs and Americans. (10) No one can deny that learning a foreign language is very important in our life. Attitude markers. Attitude markers convey a different aspect of the writers' involvement with the written message, i.e., their mental attitude toward the propositional content. The two groups of essays differ with respect to the common formal devices of this metadiscourse act. NE writers mostly use adverbs such as hopefully, regrettably, and unfortunately. NNE writers prefer longer constructs such as it is a wonderful thing, it is a curious thing. The following examples from the NE and NNE texts, respectively, illustrate this subjective attitude toward the content of the arguments: (11) The words "Arab" and "terrorist" will regrettably be linked in the minds of Americans for some time to come as we witnessed last week in the bombing in Oklahoma City. (12) ...and it's a curious thing [that the Egypt-U.S.A. relation is a perfect one] if one tries to perceive perfectly because of the fundamental difference between the two countries.

119

Metadiscourse in English and Arabic Argumentative Writing

Native Arabic vs. nonnative Arabic texts Table 3 shows the frequency of the three classes of metadiscourse in the Arabic data. Table 3. Categories and frequencies of metadiscourse in native and nonnative Arabic texts. Native Arabic Hedges Emphatics Attitude markers TOTAL

No. 71 102 18 191

%* 8.7 12.5 2.2 23.4

Nonnative Arabic No. 69 48 20 137

%* 9.4 6.5 2.7 18.6

* Percentage is to the total number of clauses. These findings indicate that the overall frequency of these classes of metadiscourse acts is higher in the NA group than that in the NNA set. Also, although the class of emphatics is considerably more frequent in the former group than the latter, accounting for the difference in the overall frequency, the difference in the frequency of hedges and attitude markers is insignificant. In the following subsections the subcategories, functions and contexts of these devices are demonstrated. Hedges. The formal devices used to code hedges in the Arabic essays can be categorized into the same classes identified in the English data (with the exception of the category of modal auxiliaries): •Verbs of cognition with a first-person singular subject, e.g., 'a'taqidu ("I think"); 'uminu ("I believe"), and 'azunnu ("I guess," "I think") •Particles of epistemic modality such as rubbama ("perhaps"); qad and la 'alla ("may" or "might"; qad has this meaning only with present tense verbs) •Adverbs of frequency like 'ahyanan ("sometimes"); 'adatan ("usually"); and 'ammatan ("generally") •Expressions like min wijhati nazari ("from my point of view"); yabdu (li) ("it seems (to me))"; min 'al-muhtamali ("it is likely"); and min almumkini ("it is possible") •Predications such as nara ("we see") and nastati'u 'an naqula ("we can say") As was the case in the English essays, hedges are mostly attached to clauses expressing leading rhetorical functions. In a further similarity to the

120

Metadiscourse in English and Arabic Argumentative Writing

English essays, the hedging techniques can be classified into three subcategories that serve subtly different functions. First, using devices like 'a'taqidu ("I believe") or fi ra'yi ("in my opinion"), the writer declares direct responsibility for the content of the argument while indicating that the statement is not to be taken as universally true. This hedging function is the most frequently represented one in the essays of the two sets (30% of hedges in the NNA group and about 27% of these in the NA group belong to that category). Example (13) demonstrates a thesis statement hedged with 'a'taqidu ("I believe") in a NA essay, while example (14), taken from a NNA essay, is a main argument featured by 'azunnu ("I guess"): (13) 'a'taqidu 'anna 'al-musawata bayna 'al-rajuli wa 'al-mar'ati shayunyas'ubu 'al-wusulu 'ilayhi. bal 'a'taqidu 'annahumin 'al-mustahil 'al-wusulu 'ilyahi dunnama yahduth zulman li- 'al-rajuli 'aw li 'almar'ati. ("I believe that the equality between man and woman is something difficult to achieve—rather / believe that it is impossible to achieve it without injustice befalling the man or the woman.") (14) lakin 'azunnu 'anna kulla lughatin ta'kisu thaqafataha wa 'ab'ada mukhtalifatan min 'al-tafkiri 'al-'insaniyi. ("But / guess that every language reflects its culture and different dimensions of human thinking.") The second subfunction of hedges, indicating the tentative rather than absolute truth of the propositional content, is coded in the Arabic essays by means of particles like la 'alla and rubbama ("perhaps") and expressions like yabdu ("it seems"), and min 'al-muhtamal ("it is likely"). These devices account for about 29% of the hedges in the NNA essays and about 21% of these in the NA group. The following examples illustrate this toning down technique of arguments in Arabic L1 and L2 essays, respectively: (15) hatta 'al-'ashkhas 'al-qa'iminna 'anfusuhum 'ala ta'limi 'allughati 'al- 'ajnabiyati yumkinu 'an yumathilu 'aqabatan fi saili tahqiq 'al-ta'allumi bi-l-nisbati li-l-muta'allimina min haythu 'uslubi 'alia 'amuli 'aw tariqati 'al-tadris ('al-ta'allum). ("Even the people responsible for teaching foreign languages themselves may represent an obstacle in the way of achieving learning on the part of learners concerning the manner of treatment or the teaching [learning] method.") (16) wa 'akhiran wa rubbama 'aqyamu fa'idatin li-ta'limi lughatin jadidatin hiya tahsinu shakhsiyati 'al-talibi. ("And finally and perhaps the most valuable benefit of learning a new language is improving the personality of the student.") The third subfunction of hedges, communicating necessary limitations on the scope of applicability of the propositional content of the arguments, is expressed in the Arabic essays by adverbials like 'ahyanan ("sometimes"), 'adatan ("usually"), now'an ma ("somewhat"), and ghaliban ("mostly"). These devices are used more often by the NNA writers,

Metadiscourse in English and Arabic Argumentative Writing

121

accounting for about 23% of the hedges in their essays vs. about 9% in the NA essays. The use of this device by NNA writers is illustrated in the following example: (17) wa hakadha qad yastakhlisu 'al-shakhu min hadhihi 'al-haqa'iqi wa 'al-'alaqati 'al-tarixiyati bi-shaklin 'amin 'anna 'al-tarafayni muttafiqani 'ala 'anna 'amrika mithla 'al-khali 'al-ghanyi, 'akbaru 'altarafayni wa 'aqwahuma. ("And thus one may conclude from these historical facts and relations generally that the two sides agree that America, like the rich uncle, is the greater and more powerful side.") Emphatics. The following formal devices are used to code emphatics in the Arabic essays: • expressions like bi la shakkin ("undoubtedly"), bi-wuduhin ("obviously") and bi- 'al-tab 'i ("of course") • clause-initial 'inna. Emphatics are used to a far greater extent by Arabic L1 writers (Table 3). This considerable difference in frequency seems to be due to the more frequent use of the emphatic particle 'inna in the NA essays. This device also appears in the Arabic L2 essays, but with a lower frequency (nearly 7% of the clauses in NA essays are prefaced by 'inna vs. nearly 3% in the NNA texts). Again, the underlying context of emphatics in the Arabic essays is key rhetorical points, i. e., thesis statement, main arguments and closing arguments. The following example, the thesis statement in a NA essay, is prefaced by the emphatic 'inna ("indeed"): (18) 'inna 'alaqata misra bi-'al-wilayati 'al-muttahidati 'al-'amrikiyati hiya wahidatun min 'ahammi 'al-'alaqati ' al-dawliyati fi 'al-sharqi 'al-'awsati, 'alaqatu shuraka'in fi 'al-salam ("Indeed the relation of Egypt and the United States of America is one of the most important international relations in the Middle East, a relation of peace partners"). An example of other devices that show strong commitment to the argued views appears in the following from a NNA text: (19) ma min shakkin fi 'annata'allumi 'al-lughati ' al-'ajnabiyati yutihu bid'afawa'ida wa manafi'a li-l-talib ("There is no doubt that learning foreign languages offers some benefits and uses to the student"). Attitude markers. Writers of the Arabic L1 and L2 essays mainly use similar formal constructs to indicate their mental attitude toward the propositional content of the written message. Expressions like the following are common in the two sets: 'al-shay'u 'al-'akbaru 'ahamiyati ("the more important thing"); 'akbaru dalilin 'ala dhalika ("the most significant evidence for that"); and less frequently, min 'al-khata 'i ("it is wrong") and ma'a 'al- 'asaf ("regrettably"). In example 20 below, taken from a NA essay,

122

Metadiscourse in English and Arabic Argumentative Writing

the writer voices her rejection of the pressure faced by female university graduates in her society by prefacing the argument with an attitude marker: (20) wa 'al-shay'u 'alladhi yufajjiru ghayzi huwa 'anna 'al-mar'ata taqifu 'amamaha mushkilatu 'al-zawaji ba'da 'al-takharruji haythu yabdu 'u 'al- 'addu 'al-tanazuli ("And the matter that explodes my anger is that the woman is hampered by the problem of marriage after graduating, when the countdown begins"). The use of attitude markers by NNA writers is illustrated in the following example by use of a closing argument that features by ma 'a 'al'asaf ("regrettably"): (21) wa lakina tusaytiri al-'umur al-siyasiyati wa al-'iqtisadiyati 'ala al-'alaqati bayna al-'alami al-'arabiyi wa 'amrika ma'a al-'asaf. ("And yet the political and economic issues dominate the relations between the Arab world and America regrettably"). The results presented above indicate the following significant observations. First, the NA essays contain the largest amount of metadiscourse expressions, whereas the NNE set shows the least amount. NNA texts demonstrate more frequent use of metadiscourse than the NE ones. The most significant difference concerning the individual classes relates to emphatics, which score the highest in the NA group. Concerning hedges, NE essays contain more hedged arguments than any of the other groups. The frequency of attitude markers is almost similarly low in the four groups, but is the lowest in the NNE group.

Discussion It seems plausible to suggest that in all languages writers use metadiscourse expressions to convey their assessment of and mental attitude toward the communicated informational content. The data of the present study show that in both their L1 and L2 writing, English and Arabic, NSs used the same categories and largely the same subcategories of the metadiscourse investigated here in mostly the same contexts. Certain differences, however, in the frequency and preferred forms of the three classes of metadiscourse are detected by the comparison of the two native sets as well as by the L1-L2 comparison in each language. The comparison between the two native sets reveals the following observations. First, metadiscourse scores higher in the NA texts than the NE ones. This overall higher frequency appears to result from the fact that writers of this group used a larger amount of emphatics. This in turn is due to the frequent use of the clause-initial 'inna by these writers. This device does not seem to have an obvious English equivalent. Concerning the comparison between English L1 and English L2 essays, the findings relating to the fre-

Metadiscourse in English and Arabic Argumentative Writing

123

quency of hedges and emphatics (the most frequent metadiscourse categories in the two sets of texts) suggest two important observations. In both sets hedges occur more frequently than emphatics. This indicates that both groups of writers tend to sound modest with respect to their generalizations and attentive to the possibility of disagreement. However, the significant higher frequency of emphatics in NNE texts and, to a lesser extent, the slightly higher frequency of hedges in the NE texts reflect that this tendency to tentativeness may be stronger on the part of English L1 writers. The interesting finding related to the Arabic L1-L2 comparison is the higher frequency of emphatics in the former group, which is again ascribed to the frequent use of 'inna. This device also seems to be the preferred one for Arabic L1 writers, whereas Arabic L2 favored other devices, which are, more or less, the equivalents of the English ones, e. g., bi-wuduhin ("obviously") and bi- 'al-tab 'i ("of course"). Holistic assessment of the NNA essays reveals that 'inna is mostly used by the most proficient writers in this group. This observation naturally needs to be consolidated by further research focusing on the correlation between L2 proficiency level and the use of metadiscourse. These results suggest some significant implications. First, English NSs used more hedges than emphatics in both their L1 and L2 essays. It seems that their concern to mitigate their arguments is transferred to their L2 writing. Second, it can also be assumed that the tendency on the part of Arabic NSs to use emphatics in their L1 writing persists in their L2 essays. Nevertheless, another result in this respect would present counterevidence to the transfer hypothesis. That is, the English NSs used a larger amount of emphatics in the Arabic L2 essays than that identified in the English L1 essays. It is suggested here that L2 writers try to abide by the norms of the target language. It is further suggested that, different as they may be from those of the native language, rhetorical conventions of the target language can be learned.

Conclusion It has been proposed that metadiscourse is one of the features of involvement in written texts. It may be further suggested, based on the data investigated in the present study, that metadiscourse as an involvement signal is a universal rhetorical device (Crismore et al., 1993). It has been shown that both English and Arabic NSs, in their L1 and L2 writing, use metadiscourse expressions to "project" themselves into texts, indicating the degree of commitment to the written message and their mental attitude toward it. This universality of metadiscourse is further evidenced by the comparison of the contexts of the various (sub)-categories of metadiscourse in the four sets. Validity markers and attitude markers have been shown to be attached to key rhetorical points of argumentation in the four sets of

124

Metadiscourse in English and Arabic Argumentative Writing

essays. Furthermore, the three functional subcategories of hedges have been demonstrated in the four sets of essays. Thus, whereas the frequency and the preferred forms of metadiscourse expressions vary, both between the NE and NA sets and across the L1-L2 texts of each language, these expressions largely appear in the same contexts in the four sets. The study suggests a number of applications in the field of teaching L2 writing. The concept of metadiscourse may prove useful in L2 writing classes. Students need to be acquainted with the concept and various classes of metadiscourse. They should also learn the appropriate contexts of every class of metadiscourse. They need to learn the various expressions of each class which are available in the target language. It is necessary to train the students in the purposeful use of metadiscourse as a rhetorical device. Guided reading of authentic texts which demonstrate the effective use of metadiscourse may prove useful in this respect. There are a number of suggestions that might be taken to further the research offered here. First, the approach adopted here can be extended to other kinds of written argumentation. It would be useful to examine the use of metadiscourse in texts produced by professional writers. Second, it would be interesting to examine the frequency and contexts of metadiscourse in other genres of discourse, e.g., informative, procedural, or descriptive discourse. Finally, it would be significant to conduct an experimental study to investigate the effect of formal instruction on the students' ability to use metadiscourse effectively.

Acknowledgments This paper is based on a section of my Ph.D. dissertation. I wish to express my gratitude to my thesis supervisors, Dr. Saad Gamal and Dr. James Stalker for their guidance, constructive criticism, and encouragement. I also wish to extend my thanks to members of my dissertation committee, Dr. Hilmi Abul Fituh and Dr. Paul Stevens, for their valuable comments and detailed evaluation of the study. I am grateful to Dr. Zeinab Ibrahim for her helpful comments on an earlier draft of the paper. My thanks also go to the American and Egyptian student-writers who provided me with the data.

Notes 1 This topic was used in Scarcella (1984). 2 Some L2 and even at times L1 examples cited here show a few grammar or mechanical errors. No attempt has been made to correct these errors, nor are they marked with the distracting sic.

Metadiscourse in English and Arabic Argumentative Writing

125

References Barton, E. 1995. Contrastive and non-contrastive connectives: metadiscourse functions in argumentation. Written communication, 12:219-239. Beauvais, P. 1989. A speech act theory of metadiscourse. Written communication 6:11-30. Brandt, D. 1990. Literacy as involvement: the acts of writers, readers, and texts. Carbondale, IL: Southern Illinois University Press. Chafe, W., and Danielewicz, J. 1987. Properties of spoken and written language. In R. Horowitz and J. Samuels (eds.), Comprehending oral and written language. New York: Academic Press, pp. 83-113. Clyne, M. 1987. Cultural differences in the organization of academic texts. Journal of pragmatics, 11:211-47. . 1991. The sociocultural dimension: the dilemma of the German-speaking scholar. In H. Schroder (ed.), Subject-oriented texts: language for special purposes and text theory. Berlin: de Gruyter, pp. 49-68. Crismore, A. and Vande Kopple, W. 1988. Reader's learning from prose: the effects of hedges. Written communication, 5:184-202. , Markkanen, R., and Steffensen, M. S. 1993. Metadiscourse in persuasive writing: a study of texts written by American and Finnish university students. Written communication, 10:39-71. El-Seidi, M. 1996. Rhetorical structure in English and Arabic expository prose: a cross-linguistic study. Ph.D. diss., Cairo University. Grabe, W., and Kaplan, R. 1989. Writing in a second language: contrastive rhetoric. In D. Johnson and D. Roen (eds.), Richness in writing: empowering ESL students. New York: Longman, pp. 263-283. Hasan, Abbaas. 1995. Al-nahw al-wafi. Cairo: Dar al-ma'arif. Johnstone, B. 1991. Repetition in Arabic discourse: paradigms, syntagms, and the ecology of language. Amsterdam: Benjamins. Kaplan, R. 1966. Cultural thought patterns in intercultural education. Language learning, 16:1-20. . 1987. Cultural thought patterns revisited. In U. Connor and R. Kaplan (eds.), Writing across languages: analysis of L2 text. Reading, MA: Addison-Wesley, pp. 9-21. . (1997). Contrastive rhetoric. In T. Miller (ed.), Functional approaches to written text: classroom applications. Washington DC: English Language Programs, United States Information Agency, pp. 18-32. Mauranen, A. 1993. Contrastive ESP rhetoric: metatext in Finnish-English economic texts. English for specific purposes, 12:3-22. Mohan, B., and Lo, W. A. 1985. Academic writing and Chinese students: transfer and developmental factors. TESOL quarterly, 19(3):515-33. Salager-Meyer, F. 1994. Hedges and textual communicative function in

126

Metadiscourse in English and Arabic Argumentative Writing

medical English written discourse. English for specific purposes, 13:149-170. Scarcella, R. C. 1984. How writers orient their readers in expository essays: a comparative study of native and non-native English writers. TESOL quarterly, 18:671-688. Stalker, J. W., and Stalker, J. 1988. A comparison of pragmatic accommodation of nonnative and native speakers in English. World Englishes, 7:119-28. Vande Kopple, W. 1985. Some exploratory discourse on metadiscourse. College composition and communication, 36:82-93. Williams, J. (1981). Style: ten lessons in clarity and grace. Glenview, IL: Scott, Foresman.

The Impact of Arabic on ESL Expository Writing Cynthia May Sheikholeslami and Nabila el-Taher Makhlouf As teachers of English as a foreign language (EFL) to students who are native speakers of Arabic, we are very conscious that a student who has mastered English syntax and idiomatic usage still produces writing that sounds foreign. ELTs (English language teachers) who are familiar with the EFL writing of Arab students immediately detect that one major problem is due to the impact of a rhetorical style transferred from their native language writing. We endeavor to pinpoint the source more precisely.

Arabic Rhetoric: Oral vs. Written Genres Turning to previous literature on the contrastive rhetoric of Arabic and English, we analyzed Kaplan's pioneering 1966 study, "Cultural thought patterns in intercultural education," familiarly known as the "doodles article." Kaplan argued that rhetorical expression in English, reflecting the logic of the thought patterns of English speakers, is linear, but that the logical pattern of Arabic and Hebrew, both Semitic languages, is zigzag, due to the excessive use of repetition, coordination, and parallelism, instead of deletion and subordination. However, Kaplan's analysis reflects oral rather than written discourse, as he has subsequently (1987) recognized. In 1987 Shirley Ostler analyzed the written English of Arabic speakers, concluding that it reflects classical Arabic (CA) sentence structure. Her model of CA sentence structure is based on the Quran. After comparing the written English of Arabic speakers with that of English speakers, she concluded that Arabic speakers make frequent use of parallel constructions and balanced and rhythmic coordination between related components. However, the Quran is an oral text that has been written down, and thus has characteristics of oral texts. The use of repetition and ornate style, as well as the use of internal rhyme between two or three words or groups of words, do not reflect the sentence structures found in expository prose written in CA (see sample text A in the Appendix). Ostler states,

128

The Impact of Arabic on ESL Expository Writing

furthermore, that Arab scholars, fearing diglossic dangers to Arabic, the language of the Quran, prescriptively devised an elaborate science of CA grammar and lexicography, and 11 centuries later, the same grammar is taught in Arabic schools. Although Kaplan (1987) claimed the same grammar and vocabulary are shared by the oral and written modes of any given language, there are differences between the grammar and vocabulary of Quranic, CA, and written modern standard Arabic (MSA), and contemporary spoken Arabic. It is true that the grammar and vocabulary of CA are still taught, but not the rhetoric of an oral text for written exposition, despite frequent use of quotations from the Quran in written and spoken Arabic. In fact, in Egyptian schools today, there is virtually no rhetorical instruction in writing classes, which are in themselves inadequate. Ulla Connor, while discussing the rhetoric of Arabic in her textbook Contrastive Rhetoric (1996), cites Johnstone's 1986 comparison of a nonlinear "Middle Eastern" argument with a linear "Western" tone. The "Middle Eastern" argument is exemplified by an interview conducted through an interpreter in 1979 by the Italian journalist Oriana Fallaci with the Ayatollah Ruholla Khomeini, originally written and published in Italian. The interviewer's style was aggressive, so the Iranian president avoided direct answers to the interviewer's questions as a defensive strategy. Again, the genre of the interview was an oral one, and, moreover, it was conducted through an interpreter. Furthermore, Khomeini was replying to questions in Persian, an Indo-European language, and not even in Arabic! In Connor's discussion, the Arabic-speaking countries and Iran are lumped together as Middle Eastern with the implication that they share a common rhetoric of argumentation, which contrasts with "Western" written argumentative discourse. As Brown and Yule pointed out in Discourse Analysis (1983), spoken and written discourse in English have different characteristics (pp. 14-19). There have been numerous studies on the rhetorical organization of both oral and written discourse in English. However, to the best of our knowledge, no such comparison is available for the rhetorical organization of Arabic oral and written discourse, although there are, of course, studies of the differences between the grammar and lexicon of CA and the various colloquials. As Connor (1996) noted, the work of Swales (1990) has demonstrated that in a given language, aside from the differences in the spoken/written discourse, one must distinguish the different genres of speaking and writing. Previous work in the contrastive rhetoric of English and Arabic has compared oral genres of Arabic with written genres of English. Furthermore, the comparisons have been on the intrasentential and intersentential levels, and have not dealt with the discourse level of whole

The Impact of Arabic on ESL Expository Writing

129

texts. In the written discourse of Arabic, the rhetoric of expository prose, even in the Middle Ages, was directly influenced by Aristotelian rhetoric, and has characteristics similar to the rhetoric of modern English expository prose (see sample texts A [CA] and B [MSA] ). A text by the medieval Arab historian, mathematician, and geographer Abu al-Fida Isma'il 'Ali (1273-1331 CE) discussing the use of lines of longitude and latitude to measure the globe's circumference is an example of linear logic, an important feature in both Arabic and English expository prose. After presenting a theory stated by Ptolemy (who lived and worked in Alexandria), Abu al-Fida describes the experiment conducted to verify the validity of the theory, and draws the conclusion. No repetition, deviation, ornateness, or incoherence disturb the Aristotelian logic of the treatise (see sample text A). In a contemporary Arabic newspaper article, "Fractures in the edifice of behavior" (Al-Ahram, January 25, 1999), the writer, Dr. Muhammad Hassan Rasmy of Cairo University, complains that the once-venerated forms of behavior of Egyptians, forms that were an integral part of their culture, have fallen to pieces. The introduction has a slightly ornate style, with examples of anaphora and rhetorical questions as the author wonders about the causes of the phenomenon, for reflecting his agitation. Then, in a strictly organized form, he enumerates the "fractures," developing each one as much as space allows. The conclusion sums up the result of the collapse in two and one-half lines (see sample text B). Clearly, then, the "nonnativeness" of the rhetoric of the EFL writing of Arab students is not due to transfer from the rhetoric of Arabic expository prose. Furthermore, many textbooks that the students read in Arabic are either translations of English textbooks or might be claimed to reflect the conventions of English expository prose since they present western knowledge and ideas, and have been written by Arabs educated in the West. Neither possible influence is reflected in the inappropriate rhetoric that appears in the ESL expository writing of Arab students. In thinking about the genre of written Arabic that could be expected to have the greatest influence on the English writing of our students, we realized that, as is generally the case in modern literate societies, they had learned to write both Arabic and English in school. Moreover, the genre of English essay that we teach them is the school or examination five-paragraph essay (see, for example, O'Donnell and Pavia, 1986, p. 13). We hypothesized that our students, recognizing that the English genre that they are being asked to write is the school essay, might be expected to transfer the rhetoric of the same genre in Arabic into their EFL/ESL writing. Our study therefore compared the rhetoric of (1) the model school essay in Arabic, (2) the model English school essay in Arabic schools in

130

The Impact of Arabic on ESL Expository Writing

Egypt, and (3) the five-paragraph model school essay commonly taught in EFL and ESL classrooms worldwide. We had Egyptian students write essays in both Arabic and English to enable us to see which features were common to both and thus might have been transferred from the students' first to their second language writing. To return for a moment to the intersentential level of analysis, the study by al-Warraki and Hassanein on The Connectors in Modern Standard Arabic (1994) shows that one of the characteristic features of Arabic prose is the use of wa as a text-organizing feature that is not comparable to the English coordinating conjunction and (see sample text C, where this type of wa is underlined). This has also been pointed out by Sa'adeddin (1987). The sentence-initial wa in Arabic merely serves to indicate that the writer is still developing the same main idea, since there are no fixed punctuation or indentation rules in written Arabic. If we analyze the example of ESL writing by an Arabic-speaking student included in Kaplan's 1966 article (p. 9), we can see that half of the examples of and in the student text are sentence-initial, raising the possibility that the student is merely transferring into English a text feature of Arabic, rather than employing a coordinating conjunction (a function wa can also perform in Arabic). It may also be possible that in this sample— which is not even a complete text—the student was influenced by an oral genre, story-telling, as the text has a narrative organization. Coordination is a characteristic feature of oral narration (see Brown and Yule, 1983, pp. 14-19). Since neither the topic the student was asked to write about nor a composing process protocol is available for this sample, there is now no way to determine the source of a possible Arabic influence on this student's English writing. Equally possible as a source is the influence of the emphasis in recent decades on audiolingual and communicative approaches to the teaching of English in Arabic schools, methods that focus on oral rather than written genres of English. In other words, the text may have been influenced by oral rather than written rhetoric from either Arabic or English.

Impact of the Model School Essay One type of Arabic writing that could be a source of negative impact on the English writing of Arabic-speaking students is the school essay. A model essay in an examination-preparation book (Adb al-Wahhab and Awad Allah, 1998) that is very popular among Egyptian students is a typical example (see sample text D). The topic, expressed in a very bombastic style, is the project to expand the inhabited land near the Nile. In the essay, the writer speaks highly of the project and its benefits in very gen-

The Impact of Arabic on ESL Expository Writing

131

eral terms without ever mentioning its name. As he sings its praises, he mentions the expected returns—minerals, tourism, poultry and cattle breeding, agricultural crops, and industry. These are repeated in other terms, but the benefits are never developed in any detail. Quotations from the Quran and Arabic poetry are used to commend the work of people who plan such rewarding projects. There is no overall organization or coherence, so that, even if a student who has memorized it leaves out a part by mistake, it will not affect the flow of words in individual sentences. This is the type of model that students may have in mind when they write in Arabic. This kind of writing caters to large classes (50-60 students) where teaching writing and following up a student's progress is almost an impossible task. The EFL writing of Arab students frequently displays features of this type of Arabic writing, which is not, however, representative of expository prose in Arabic. The same topic was given to Arabic-speaking graduate students at the high intermediate level in the English Language Institute at the American University in Cairo before an intensive program in writing instruction in English began. The students wrote at home and had an opportunity to revise their texts, after reading and discussing a newspaper article on the topic in class. An analysis of one sample (see sample text E) reveals, however, that further experience with written Arabic exposition during undergraduate studies in an Egyptian university has obliterated the influence of the school essay. The logic is linear, cause and effect follow smoothly, and little deviation or unnecessary statements obstruct the flow. The writer briefly mentions the problems of overpopulation, especially in Cairo, necessitating the building of housing complexes near and away from Cairo, although those built do not have the necessary facilities and services. The student moves on to the two major agricultural projects in the desert, naming them and mentioning specific details about them, thus developing these points. The writer concludes by stating the overall benefits and recommending that services and facilities be good enough to attract residents. Model essays in another examination-preparation book that is very popular among Egyptian students learning to write in English are another possible source of influence. A typical example (see sample text F) begins with reference to an external authority and the reader is led to expect a problem-solution-evaluation structure. However, the use of the in quotation marks in the second sentence indicates that the writer does not believe overpopulation is even a problem. Lexical cohesion is largely limited to repetition of vocabulary, whereas terms such as, family planning and birth control, while in the same semantic field as overpopulation, contradict the writer's stance that it is not a problem. The second paragraph sets up a

132

The Impact of Arabic on ESL Expository Writing

contrast to the United Nations' claim and also moves into expression of personal opinion and generalizations. The two sentences about food production apparently return us to the problem-solution mode, although this is not made clear as the writer never relates this idea to the topic. Reference to the lack of fresh water is tacked to the end as a solution, again with no relationship to the problem of overpopulation indicated. In short, this essay lacks logical organization and cohesion, and gives the impression of a group of mostly unrelated sentences dropped onto the page. The same topic was given to Arabic-speaking graduate students at the high intermediate level in the English Language Institute at the American University in Cairo before an intensive program in writing instruction in English commenced. The students wrote 30-minute timed essays in class. An analysis of one sample (see sample text G) clearly showed the influence of the model essay on the same topic. The writer begins by denying that overpopulation is a problem, and then goes on to describe solutions to the problem, finally claiming that overpopulation is itself a solution. A lack of logical organization and cohesion similar to that of the model essay is apparent. These examples support our claim that the model school essay is a likely source of transfer in the EFL writing of Arab students. However, students more experienced with Arabic expository prose are able to write English expository prose in an acceptable rhetorical style. At the conclusion of his 1966 article, Kaplan recommended two types of exercises to teach students how to follow a linear rhetorical structure in their written discourse. ELTs will recognize these as pervasive exercise types in writing courses. Both exercises are restricted to the paragraph level, which Kaplan defines as a logical though "artificial thought unit employed in the written language to suggest a cohesion which commonly may not exist in oral language" (Kaplan, 1966, p. 16). The first exercise is to scramble the order of sentences from a paragraph and ask the students to rearrange them in what appears to them to be a normal order, and then compare it with the original order, which should be explained and justified by the instructor. The results will, Kaplan claims, "demonstrate the diversity of views or cultures represented in the classroom" (p. 16). The implication is that the students will thereby learn to recognize the contrasting rhetorics of their native language and English. Our experience, however, indicates that this type of exercise is dependent on the background knowledge of the topic that the student has (i.e., the student's schema) and is not a useful way to demonstrate the rhetoric of English expository prose. Kaplan's second recommended type of exercise is for the instructor to provide a topic sentence with a partly completed outline of the remainder of the paragraph, and ask the students to fill in the remaining examples and

The Impact of Arabic on ESL Expository Writing

133

illustrations to support the point. If the topic is difficult, the instructor might provide the examples in scrambled order. These exercises, Kaplan states, will fulfill the limited aim of the EFL or ESL class of providing students with an acceptable form in which they can write English. Teachers will recognize that students are likely to produce rather formulaic writing as a result of the influence of this type of exercise. The students do not develop a nativelike grasp of the rhetoric of expository prose in English. It should be pointed out that the school essay is an artificial genre and that the actual rhetorics of the various genres of written English correlate only very slightly with this form. The students who learn the school essay genre have not learned to control the rhetorics of written discourse that are accessible to the native writer of English, and thus their writing will continue to seem non-native outside the confines of the school environment. We might echo Liebman's 1992 plea "that, if ESL writing teachers want their students to succeed at a variety of academic writing tasks, they must become aware...of these different forms" (pp. 158f.) Liebman goes on to urge that ESL writing instructors also be more aware of the differences in instructional backgrounds of their students to enable them to vary their approaches to instruction in written English. Both the Japanese and Arab students surveyed by Liebman "had few experiences writing for audiences other than the teacher...[and] very little experience writing outside school...[as is the case with] their British and American counterparts.... This lack of variety extend[ed] to the functions of their writing...[as] they had written mainly informative prose for...'the teacher-as-examiner'" (p. 150). Liebman's results also support our claim that school writing is an important source of influence on EFL students' writing.

Strategies for Improvement The response to this problem in the teaching of writing to native speakers of English in the United States in recent years has been the Writing Across the Curriculum movement, where students learn to write in the various academic genres when they are enrolled in courses in these disciplines. In his 1987 article "Cultural thought patterns revisited," Kaplan himself stated, "it is the responsibility of the second-language teacher to increase the size of the inventory" available to the nonnative writer. American ESL writing instruction has recently paid more attention to strategies of invention, focusing on the writing process as well as the finished product. Egyptian students in our classes at the American University in Cairo frequently express the need for more help with invention, the lack of which may be a contributing factor to the poor content development that is such a noticeable feature of their EFL writing. Their revision of

134

The Impact of Arabic on ESL Expository Writing

their writing, however, rarely if ever extends beyond sentence-level grammar and vocabulary to discourse-level organization of the written text. Thus, they tend not even to be aware of the text's rhetorical organization, since it is not a focus of revision. Whereas authentic ESL/EFL reading materials present a variety of both academic and nonacademic genres of written English to students, the students are not expected to attempt to write in most of these genres, other than postcards, letters, and occasionally simple stories. The focus of instruction continues to be the classic five-paragraph school essay and the development of the ability to arrange ideas in a linear fashion, an ability students are not thought to be aware of from their exposure to written Arabic. Such an approach betrays the continuing influence of Kaplan's 1966 article rather than his more recent analysis, as well as the fact that researchers in the contrastive rhetoric of Arabic and English have not had a sufficient knowledge of Arabic and written discourse in Arabic to draw valid conclusions. More research in the written and oral discourses of Arabic, particularly employing composition protocols, is needed to come to a better understanding of the impact of Arabic on ESL writing. The researchers involved in these studies should be bilingual or work in cooperation with native speakers of each language. Nevertheless, it is clear that at least one source of transfer from Arabic into English writing among our students is the genre of the school essay.

Appendix Sample Text A: Sample of CA expository prose (English translation and Arabic original) Abu al-Fida Isma'il 'Ali (1273-1331 CE), from Taqwim al-buldan ("Regional geography") Reproduced in Ahmad Fahmy Abu al-Khair, 'Ulum al-'arab al-riyadiya ("Arab mathematical sciences"), 1930. The earth is spherical and geocentric. Its surface, which is convex, is parallel to the sky, which is concave. The longitude and latitude lines measured from the meridian are parallel to those of the sky and are divided into 360°, the meridian being an imaginary line passing through the North and South poles. These imaginary lines (longitude and latitude), as they start from the North Pole reaching the South Pole, are straight, undisturbed by elevations and depressions, with a space between each of one degree, equivalent to 66 2/3 miles, the sphere measuring 360°. This was the theory of ancient scientists such as Ptolemy who wrote al-Majest. A number of learned men in the time of the Caliph al-Ma'amun studied the works to verify the validity of the theory and were summoned to the wilderness of Sanjar in

The Impact of Arabic on ESL Expository Writing

135

Iraq to do so. One group went northward, and the other southward. After measuring the pole's elevation (astronomically), the two groups met again; the northward group had covered a distance of 65 l/3 miles, while the other covered 65 only. They agreed to take the lowest figure, which is 65 miles, as equivalent to one degree.

136

The Impact of Arabic on ESL Expository Writing

Sample Text B: Sample of MSA expository prose (English synopsis and Arabic original) Muhammad Hassan Rasmy, Thuqub fi jidar al-suluk ("Fractures in the edifice of behavior"), Al-Ahram, January 25, 1999, p. 10 The writer complains that the once venerated forms of behavior of Egyptians, forms that were a solid part of their culture, have fallen to pieces. The introduction, in a somewhat ornate style, laments the situation. The writer moves on to the supporting points of his argument: the first "fracture" is (1) love, then (2) generosity, followed by (3) commitment, (4) respect, (5) truth, (6) good taste, and finally (7) forgiveness. The writer concludes that man without refinement of feelings and proper behavior is only a living creature who needs thousands of years to become human.

The Impact of Arabic on ESL Expository Writing

137

138

The Impact of Arabic on ESL Expository Writing

The Impact of Arabic on ESL Expository Writing

139

Sample Text C: The use of wa as a text-organizing feature of Arabic (in the English translation, connectors are underlined; the text-organizing wa, boldfaced in the Arabic text, is not translated) Nariman Naili al-Warraki and Ahmed Taher Hassanein, The connectors in modern standard Arabic. Cairo: American University in Cairo Press, 1994. Naguib Mahfouz was born in al-Jammaliya, one of the poorer (people's) quarters of Cairo, which is situated near al-Husayn Mosque. His father had been a civil servant and then later worked in business. Mahfouz was an excellent reader for he read books by al-Manfaluti and translated some of his works. Then he read from the works of Taha Hussein, al-Akkad, Salama Musa, Ibrahim al-Mazini, Tawfiq al-Hakim and others, and did not fail to read the classics, whether Arabic or Western. He began his literary activity at an early age by writing the essay and by translation. Then he turned to writing the short story and the novel. At present, Mahfouz is considered the most famous Arab novelist for he wrote a great number of novels and short stories, which gained fame whether in Egypt or in other Arab countries.

140

The Impact of Arabic on ESL Expository Writing

Sample Text D: Model essay from an examination-preparation book popular in Egypt today A. Abd al-Wahhab and M. A. Awad Allah. Al-mu 'allim fi-l-muraja' al'amrna wa-l-niha'iyya li-l-marhala al-thanawiya al-'amma. Cairo: AlMu'tamada. 1998, pp. 331-332.

The Impact of Arabic on ESL Expository Writing

141

142

The Impact of Arabic on ESL Expository Writing

Sample Text E: Sample of Arabic-speaking graduate student at high intermediate level (English synopsis and Arabic original) Overpopulation in the old valley has become a problem in Egypt and a hindrance in its road to progress and a better life. This is particularly evident in Cairo, especially since during the day its population reaches ten million. For this reason, housing compounds have been planned around Cairo. Similarly, to get out of the crowdedness of the narrow valley, new towns far away from Cairo have been built. However, they still need many facilities to attract people. As the increase in population also needs agricultural land, the government has planned two projects. The first is the al-Salaam Canal, which will flow from the Nile to Sinai. The second is the Toshka project, which will be irrigated from Lake Nasser, thus making a new valley in southern Egypt. All these projects will provide better housing and employment opportunities. However, planning roads, facilities, and services must be well studied to vie with Cairo, which has the best in the country.

The Impact of Arabic on ESL Expository Writing

143

144

Impact of Arabic on ESL Expository Writing

Sample Text F: Typical example of a model essay from an examination-preparation book A. Saad, GEM English language special course, third year secondary (sc & arts), Cairo: Maged Printing Press, 1992. Do you think overpopulation is a problem? The United Nations Organization says that overpopulation is a serious problem. It holds conferences to deal with "the" problem. International and local funds are raised to carry out family planning and birth control programmes. One of the overt reasons given by the Super Powers for space travel race is to "search for a solution" to the overpopulation problem. However, I believe that overpopulation is not a problem. "Population" or "manpower" is one of the factors of producton and therefore, it should not be considered a problem in any sense. With the highly advanced technology and with the introduction of computers in almost every field of activity, food could be produced in abundance and even from non-organic sources. Floating farms, desert plantations, fish production and animal raising are rightfully easy activities. Sweet water is no longer a problem. Atomic and nuclear energy are being used for a wide spectrum of activities and with a great deal of safety. In short, it is true that "Many hands make light work!"

The Impact of Arabic on ESL Expository Writing

145

Sample Text G: 30-Minute essay written in an EFL class by an Egyptian student Do you think overpopulation is a problem? No doubt that there's an argument between people about Overpopulation Is it a problem or not. Both sides have a point of view. And I think it isn't a problem if there are good planning. Turning our market into a productive one from its state, as a consumer market, is the way to get good results. This will become a reality by the government encourging to the investors to come and invest in our giant projects. Huge projects like developing the desert and making big industries areas, can use overpopulation postively. Not only by making a large number of people works in the project of developing but also by employ them in the factorys and by finding them a place to live in. In conclusion, I think that good planning can give a good results, by dealing with the overpopulition as a solution not as a problem.

146

The Impact of Arabic on ESL Expository Writing

References Abd al-Wahhab, A. A., and Awad Allah, M. A. 1998. Al-mu'allim fi-lmuraja' al-'amma wa-l-niha'iyya li-l-marhala al-thanawiya al'amma. Cairo: Al-Mu'tamada. Abu al-Khair, Ahmad Fahmy. 1930. 'Ulum al-'arab al-riyadiya. Cairo: Al-I'timad Press. Connor, U. 1996. Contrastive rhetoric: cross-cultural aspects of secondlanguage writing. Cambridge: Cambridge University Press. , and Kaplan, R. B. (eds.). 1987. Writing across languages: analysis of L2 text. Reading, MA: Addison-Wesley. Inani, M. 1997. Fann al-tarjama. Cairo: Egyptian International Publishing Company-Longman. Kaplan, R. B. 1966. Cultural thought patterns in intercultural education. Language learning, 16:1-20. . 1987. Cultural thought patterns revisited. In Connor and Kaplan (1987), pp. 9-21. Liebman, J. D. 1992. Toward a new contrastive rhetoric: differences between Arabic and Japanese rhetorical instruction. Journal of second language writing, 1:141-165. O'Donnell, T. D., and Pavia, J. L. 1986. Independent writing. Boston: Little, Brown. Ostler, S. E. 1987. English in parallels: a comparison of English and Arabic prose. In Connor and Kaplan (1987), pp. 169-185. Rasmy, Muhammad Hassan. 1999. Thuqub fi jidar al-suluk. Al-Ahram, (Jan. 25), p. 10. Sa'adeddin, M. A. 1987. Three problem areas in teaching translating to native Arabic literates. Anthropological linguistics, 29:181-193. Saad, A. 1992. GEM English language special course, third year secondary (sc & arts). Cairo: Maged Printing Press. Swales, J. 1990. Genre analysis: English in academic research settings. Cambridge: Cambridge University Press. Al-Warraki, Nariman Naili, and Hassanein, Ahmed Taher. 1994. The connectors in modern standard Arabic. Cairo: American University in Cairo Press.

Teaching 'Form' in English Verse to Arabic Poetry Readers Loubna Abdel-Tawab Youssef Teaching freshman writing in English to Egyptian students, I observed that they have difficulty in understanding the idea of form. Helping them to overcome this difficulty is vital, since in the Freshman Writing Program (FWP) at the American University in Cairo (AUC), they must acquire the skill of structuring an essay and an academic paper with a plausible argument. They must also learn that unity, cohesion, and organization contribute to the overall effect of the paragraph and the essay in general. Several factors in the Egyptian students' education might account for their inability to grasp the concept of form. However, the focus here is on the effects of the way in which they studied Arabic poetry for at least six years at school. What is imprinted on their minds and what affects the way they read a text in verse or in prose, in Arabic or in English, is the concept of the line (a distich or a monostich) as a poetic unit, one that is independent of the surrounding verses and is therefore the fundamental unit for literary analysis. The purpose of this paper will be to shed light on what pupils learn about Arabic poetry and the consequences of this for the students and to show how reading an English sonnet can help them in understanding the idea of form in order to apply this understanding to structuring their own essays. Admittedly, these students will not be writing in verse, but reading an English sonnet, which is "compact, shapely, highly finished, and able to contain, in concentrated form, almost all that is human" (Spiller, 1992, p. 1), raises their awareness of many of the challenges of writing

The Impact of Egyptian School Anthologies Brought up on the tradition of Arabic poetry, Egyptian students in the FWP are aware that poetry has been the predominant mode of literary expression in Arabic. Their school textbooks have led them to believe that all Arabic poems are composed of lines, each of which is divided into independent or interdependent hemistichs. They remember certain notions

148

Teaching 'Form' in English Verse to Arabic Poetry Readers

about Arabic poetry from their school days: the Arabic "voices:" in poetry tend to be more public than private; the qasida (p1., qasa'id; "ode"), a long poem composed of hemistichs and with both meter and rhyme, starts with a description of the 'atlal (a deserted encampment) and proceeds to a traditional section about the journeys of the poet in the desert before even the real purpose of the poem is mentioned (praise of the' tribe, denigration of an enemy, or the fostering of tribal pride and solidarity); and pre- and post-Islamic Arabic poetry belong to an oral tradition and thus there can be differing versions of a poem depending on the transmission from the poet to the reciter (rawiri) who both have the ability to improvise. Enani (1986) succinctly sums up what the school textbooks focus on in the following statement: The traditional qasida usually opened with a few lines on the beauty of the beloved, on the pangs of unrequited love, sometimes on separation, then proceeded to the central subject usually of public interest, the accepted general genres being: the panegyric, the satire, the praise poem that celebrates the glories of the individual poet or his tribe, the virtues of asceticism, praise of Prophet Muhammed, love, or wine (p. 17). Rereading some of the extracts and poems that Egyptian pupils study made me realize that although in the 20th century "the Arabic poem ceased to be an open-ended collection of lines of equal length, sharing the same rhyme, with varying themes, and became the record of an emotional experience" (Enani, 1986, p. 17), the extracts and even the selection of extracts from the modern period that Egyptian pupils study belong to the classical tradition or are written by poets who revive this tradition. Thus, in addition to the pre-Islamic and early Islamic extracts cited below by al-Khansa' , alNabigha, al-' A'sha, 'Antara, as well as others like these, they read late 19thand early 20th-century poetry by Mahmud Sami al-Barudi (1838-1904 CE), Khalil Mutran (1872-1949), Ahmed Shauqi (1868-1932), Hafiz Ibrahim Muhammad (1872?-1932), Ibrahim Nagui (1898-1953), Abu al-Qasim alShabbi (1909-34), all of whom were educated in the West, but whose poetry does not sound modern. The fact is, when Egyptian youngsters remember classical Arabic songs, echoing in their minds and hearts is the voice of Om Kalthoum singing lines by Hafez Ibrahim,

[1] "Men have stood watching me erect the house of glory all by myself

Teaching 'Form' in English Verse to Arabic Poetry Readers

149

[2] And, at times of challenge, the builders of the Pyramids spoke on my behalf *

by Ahmed Ramy,

"Egypt, the land in my mind and on my mouth, I love her with all my heart and soul."

and others like these. At this early stage of their education, pupils are not introduced to the controversy between critics who belong to the conservative critical tradition that expressed the view that the structure of the qasida is fragmented and others who have recently challenged this attitude in their analyses in order to show that the different sections of a qasida reflect the poet's purpose in dealing with a variety of subjects that interest his audience. Heinrichs (1973), for example, discusses in detail the idea that scholars of Arabic poetry and poetic theory must "distinguish three classes of deficiencies: the failure of the theorists to understand (a) complex structures, (b) developments, and (c) the true extent of the segments of reality that are treated in poetry" (p. 35). In Beyond the Line, Van Gelder (1982) insisted that "Classical Arabic poems have been described as lacking 'unity' ever since Western critical standards were applied to them" (p. 14). This enlightening remark is followed by Van Gelder 's reference to the recent reassessment of the corpus of classical Arabic poetry by the application of new ideas in analysis and theory, thus demonstrating the richness of the tradition available to us. He says: In recent times a number of studies on the structure of classical Arabic poems have shed more light on this concept of 'molecularity' and have given ample attention to the ways in which poems hang together. Techniques of description and analysis differ; some look for symmetries at the levels of phonological, morphological and syntactic patterning, other construe taxonomies of opposing thematic classes, while digging for archetypal and mythical bedrock. The status of the results varies according to whether they are based on the analysis of one, five, or a large number of poems, but all authors stress coherence and unity rather than disjunction and disunity, on the assumption, tacit, or overt, that they have set things right by revealing the hidden and hitherto neglected beauties of Arabic poetry, (p. 15)

150

Teaching 'Form' in English Verse to Arabic Poetry Readers

Reading a sonnet with my students, I was struck by an immediate response: they could not understand that the poem has an end. So far as they are concerned, a line of verse is the unit, the only unit imaginable. It is grammatically complete and the idea is conveyed and can stand by itself. Teaching them to read a sonnet, i.e., a poem that advances more like a logical argument from beginning to end or from premise to conclusion, both confirmed my observations that for Arabic-speaking students "form" is a problem and helped me find solutions. Because what they have studied in this case pertains to their own culture and identity, it is engraved in their minds and being exposed to a different concept from another culture is not easy at best. It may be helpful to quote a few lines by poets from the different phases of Arabic poetry from Egyptian school textbooks. (In no way are these meant to lead to a comparison between classical and modern Arabic poetry with its different schools, an evaluation of either, or a survey or description of the different schools of Arabic poetry.) These examples show how there is "form" for the line, but not for the poem as a whole, and that the couplet is the only form that Egyptian pupils are familiar with. To start with, here are extracts that focus on different types of the theme of praise by poets from the pre-Islamic and early Islamic era. In a moving elegiac tribute to her brother Sakhr, al-Khansa' (fl. end 6th c., CE) touches on the merits of this noble individual who died in tribal combat when she writes:

[1] "Oh, my eyes, flow with tears and never dry, won't you grieve for the generous Sakhr? [2] Won't you grieve for the young master? [3] Of enormous height and estimable rank? Chief of his tribe, though still a youth. [4] If men stretch their arms toward glory, Glory would stretch its very arm toward him."

Al-Nabigha al-Zubyani (Abu Imama ibn Mu'awiya; d. 604 CE) praises Amr ibn al-Harith el-Ghasani, who helped him while he was in exile in Ghasan, saying:

Teaching 'Form' in English Verse to Arabic Poetry Readers

151

[1] "I owe Amr many a favor rendered undiluted by his father. [2] When the true men of Ghasan attack, I trust he will always be victorious."

Al-A'sha (Maymun ibn Qays; d. ca. 629 C.E.) glorifies his beloved in the following lines:

[1] "Bid Horaira farewell, the caravan is parting! Man, would it be easy to part? [2] A maiden with fair brow, long hair, and bright teeth, treading slowly like a delicate foot treading on mud."

Finally, 'Antara ibn Shaddad (fl. 6th c., CE), the illustrious warrior who was known for his love for his cousin 'Abla, expresses pride in his own prowess, saying:

[1] "My enemies were deceived by my silence And thought I had forgotten the glory of my people. [2] How can I be oblivious to the fine men Who have raised me up in their plenty? [3] If ever they were threatened by an enemy They'd call me and I'd readily answer. [4] The edge of my sword reaps lives, And my lance is a deadly blow in the chest. [5] My heart was made tougher than steel; Though steel can be molded, my heart cannot."

Although poets, scholars, and critics in the 20th century have attacked "the mistaken but commonly held belief, popularized by medieval Arabic literary critics, that the classical Arabic poem consisted of single independent lines loosely strung together" (Badawi, 1993, p. 35), Egyptian

152

Teaching 'Form' in English Verse to Arabic Poetry Readers

school textbooks do not refer to the idea of structure or unity. In the case of the examples of classical poetry cited above, it is assumed by modern textbook authors that whether these lines are originally from the beginning, the middle, or the end of the poem, there is no significance to the ordering of the lines. Pupils are encouraged to believe that the lines can be rearranged without any consequences for the meaning or the poetic significance of the work. The idea of development is relevant, Egyptian students are taught, only in so far as the description and imagery are related to the theme. Unfortunately, the late 19th- and the 20th-century Arabic poetry that pupils study is regarded as similar in many ways to classical poetry, although the syntactic and semantic links between the lines are obvious to any mature reader. The selections pupils study show that modern Arabic poets adhere to monorhyme and monometer (Moreh, 1998) and the line is still the poetic unit. These lines are from a poem that secondary school pupils study about the agony of love and the suffering from sickness by Mutran who "systematically and deliberately sought to achieve unity of structure" (Badawi, 1985, p. 108):

[1] "In vain have I consoled myself with hopes, In an exile that would, they said, cure me. [2] If this body can be cured by its fresh air, Can ever fire be put off by air so fresh? [3] In vain do I traverse those lands, And my exile is a double malady, not a recovery. [4] I live solitary with my longing, solitary with my misery, solitary with my suffering." True, the compilers of the school anthologies do in fact point out that Mutran belongs to the Romantic school of modern Arabic poetry (critics like Badawi, Van Gelder, and others explain the danger of applying Western critical terms to Arabic poetry, and state that this is done for convenience), but they discuss the lines of the poem independently. So pupils are introduced to schools of modern Arabic poetry and learn that new themes (e.g., patriotism, nationalism, freedom, independence, revolution, socialism, commitment, national causes) are dealt with; however, the diction is classical and the shape of the poem is the same: the lines rhyme and

Teaching 'Form' in English Verse to Arabic Poetry Readers

153

are divided into hemistichs. Here are a few examples. Al-Barudi, who is "the true precursor" of the Arabic poetic revival by whom "Arabic poetry was brought once more to bear upon the serious business of life" (Badawi, 1993, pp. 25-26), indirectly attacks Khedive Ismail for being greedy and unjust:

[1] "Man wishes to posses the whole earth, Not knowing what God has disposed. [2] Wealth could, sometimes, lead its owner to perdition, And greed often turns against itself."

While in exile in Spain poet laureate Ahmed Shauqi wrote about homesickness, saying:

[1] "My homeland—even if immortality lured me with its call, my soul would still drag me back to it. [2] And my heart would be moved, even when in the springs of heaven, by its thirst for the suburbs of Ain Shams."

Ibrahim Nagui, who belongs to the Apollo school of Arabic poetry, writes about a personal experience when he sits on a rock and recalls a feeling:

[1] "Oh, cliff of reunion, I ask you When will time bring together what it has sundered? [2] You are a cliff that has brought two souls together, Seeking shelter in its unsurpassed beauty."

These samples show that the compilers of the school textbooks know that Arabic poetry has been going through an era of transformation and experimentation since the beginning of the 20th century. Nevertheless, they chose extracts that demonstrate that although the themes dealt with in modern poetry might be different, the shape of the line is the same.

154

Teaching 'Form' in English Verse to Arabic Poetry Readers

Form in the English Sonnet To explain to students with this background and training the literary term form, which is inclusive, variable, variously interpreted, and therefore ambiguous, proved more difficult than expected. For homework, the students of the two classes I teach (English 112 and 113 of the FWP) were assigned to read a sonnet by Shakespeare ("When in Disgrace with Fortune") and another by Shelley ("Ozymandias"), and look up the terms form, sonnet, and all the key words in each sonnet in a collegiate dictionary. After a quick discussion about the different meanings of the word form, I distributed a photocopy of the entry on form and another on sonnet from Webster's Collegiate Dictionary.2 The first exercise was to have them exclude the meanings that are not relevant, while the second required them to formulate a definition that they can use in discussing the sonnet by combining the relevant points. Of course, the fact that item 5 of the dictionary definition of form (see note 3) directly and appropriately refers to "the sonnet form in poetry" proved helpful. They realized that at one level, form can refer to structure; at another, to metrical patterns; at a third, to the characteristics that a text may share with others; and at a fourth, to the orderly arrangement of ideas and words. In fact, all of these levels can and should be noted in a thorough analysis of the form of the two sonnets in question. In the experience with my two classes, four steps of analysis are taken to help students grasp these four levels of form: (1) form as structure (including metrical patterning); (2) form as development; (3) form as cohesion; and (4) form viewed from the end.

Form as structure For my Arabic-speaking students, the superficial structure of the sonnet was problematic: the basic difficulty was that the end of the line did not bring an end to the sentence or the idea: the Shakespearean sonnet is one long sentence, whereas Shelley's "Ozymandias" has four sentences, one of which is nine lines long and another is three words. Initial questions that relate to structure and metrical patterns can be asked: why a "quatorzain"3? why 14 lines? why an octave and a sestet? Heninger (1994) has an interesting theory that the division into an octave and sestet "when reduced to its lowest ratio, this proportion becomes the relationship between four and three" (p. 73). "Four," he adds, "is the mundane number...the number of the four elements that comprise the macrocosm, ... of the four seasons that comprise the annual unit of time.... Four signifies this world" while "in the Platonic-Christian tradition...three signifies the deity" (p. 76) and "the sum of these integers, seven, represents the entire range of human experience from lowest to highest" (p. 77).

Teaching 'Form' in English Verse to Arabic Poetry Readers

155

Both "When in Disgrace with Fortune" and "Ozymandias" have the octave-sestet pattern in which the poets convey different experiences that deal with a state of despair that they overcome in different ways leading to an image of the sublime. For example, Shakespeare's lover, the "I" in the poem "When in Disgrace with Fortune," is the central character who is transformed to a state of bliss when he remembers his beloved: When in disgrace with fortune and men's eyes, I all alone beweep my outcast state, And trouble deaf heaven with my bootless cries, And look upon myself, and curse my fate, Wishing me like to one more rich in hope, Featured like him, like him with friends possest, Desiring this man's art and that man's scope, With what I most enjoy contented least; Yet in these thoughts myself almost despising— Haply I think on thee: and then my state, Like to the lark at break of day arising From sullen earth, sings hymns at Heaven's gate; For thy sweet love rememb'red such wealth brings That then I scorn to change my state with Kings.

Shelley's "I," however, is one of four characters that speak in the sonnet: the narrator, the traveler, Ozymandias, and the sculptor each has a voice. Known for his "autobiographical impulse" (Clark, 1989, p. 7), Shelley uses direct speech here to attract the attention of the reader and to stress that this is a personal experience of great value: I met a traveler from an antique land Who said: Two vast and trunkless legs of stone Stand in the desert. Near them, on the sand, Half sunk, a shattered visage lies, whose frown, And wrinkled lip, and sneer of cold command, Tell that its sculptor well those passions read Which yet survive, stamped on these lifeless things, The hand that mocked them and the heart that fed; And on the pedestal these words appear: 'My name is Ozymandias, king of kings: Look on my works, ye Mighty, and despair!' Nothing beside remains. Round the decay Of that colossal wreck, boundless and bare The lone and level sands stretch far away.

156

Teaching 'Form' in English Verse to Arabic Poetry Readers

More questions come to mind: How does this information about the two sonnets affect the study of the structure of the sonnet and how can "form" be interpreted as structure? Clearly, the idea of structure is metaphoric, because a poem consists of words and utterances—phonological units rather than bricks and mortar. The idea of building requires the student to imagine a kind of edifice with a foundation, pillars, support walls, and so on. This itself may be difficult for a youngster to associate with a poem. So it may be easier to begin by establishing the idea that there is a starting point for every work of art that may be regarded as the basis. This may be an image, a feeling, a theme, or an attitude. The student thus can be made conscious of a central idea representing the foundation of the building. Identifying this foundation in itself can be conducted as a sort of game for the students. For Shakespeare, the starting point is a deceptive conditional clause with the relative pronoun of time, "when." This clause allows the suspension and resumption of syntax in which the poet uses grammar to deal with the plausible but unreal list of imaginary "miserable conditions" (conveyed in the words "beweep," "trouble," "look," "curse," "wishing," "featured," "desiring," "enjoy," "despising"). The accumulation of these words may be regarded as a specific poetic ruse, building up to a climax that introduces the one real operative verb in the sonnet: "I think on thee." The form takes a line marked out by grammar until a climax comes almost as a surprise, although it is necessitated by the sense. Shakespeare's two-part division invites the reader to an admirably unified octave that leads to a volte-face which is followed by a "turn" of thought and shift in feeling in the sestet. At the outset, Shelley uses the narrative mode which at the superficial level seems simple and straightforward, but the exotic mood initiated by the encounter with "a traveler from an antique land," who tells his own enigmatic story, creates an atmosphere of mystery that pervades the sonnet. The two first speakers fade in the background of the picture that Shelley portrays: the king and his sculptor take over, though we are never sure who makes the final remark. The octave and the sestet are bound together by means of three factors: the extended description of the remains of the statue of the king, the conjunction "And" at the beginning of the sestet and the interlocking rhyme scheme. The four main verbs in the four sentences that constitute the sonnet are static verbs that imply lack of movement, but the poem starts in the past tense ("met") and proceeds and ends in the present with only three main verbs: "lies," "remains," and "stretch." Unlike Shakespeare, who has three quatrains and a couplet that follow the regular rhyme scheme abab—cdcd— efef—gg, Shelley chooses to bind his quatrains together by using a rhyme scheme that suits his purpose.

Teaching 'Form' in English Verse to Arabic Poetry Readers

157

Form as development Another means of reaching the idea of form is through development. The students may be taught that the poem advances from simple to complex ideas or from the particular to the general or from one idea to the next logically and gradually or through binary opposition, such as antitheses or contrasts. The students will therefore be encouraged to make use of the analytical mechanisms used in other disciplines (entailment, embedding, etc.) in following up the progress of the poem and in tracing out a certain line of development. The theme Shakespeare deals with is developed into an idea, an argument, and a resolution. First, his idea in the first quatrain concerns hypothetical conditions of shame, ill repute, alienation, loneliness, hopelessness, failure, poverty, and more. In the second quatrain, when he compares himself to others who are more fortunate, he becomes envious, which leads not to self-pity but to more misery, contempt, and lack of selfrespect. Finally, the resolution is foreshadowed in the third quatrain with the use of the verb "think." The implication is that the mind and the heart can change this world from a living hell to a heavenly place. "Ozymandias" demonstrates how exquisitely the theme of the transience of life vis-a-vis the survival and permanence of art and nature can be dealt with by engaging the reader in interpreting the sonnet. The particulars that are disclosed one by one, namely the legs, the visage, the hand and the heart and the pedestal, convey a sense of fragmentation that shows the destructive effect of time. The extended details with which Shelley describes the statue show that he has a zoom lens that initially brings us close to see the expression on the king's face and then even closer to read the words on the pedestal. After hearing, seeing, and reading the words of the king, we take a few steps back for a full view of the "colossal wreck" in the "boundless" desert. History tells the truth and the truth is that men live and die however great. This is revealed through a series of juxtapositions: life versus death, life versus art, art versus death, dynamic versus static, present versus past.

Form as cohesion Another means of driving home the idea of form is through relating the details to one another through what is commonly called cohesion and which ultimately is conducive to coherence. When groups of images are drawn from nature and center around a specific scene or landscape, they generally have internal ideational links that may help the student grasp the cohesion of the general framework. Exercises in relating such images or clusters of images to one another will heighten the student's concept of

158

Teaching 'Form' in English Verse to Arabic Poetry Readers

form. If more links are found or established within these clusters (color, sound, etc.), the students will realize that the texture itself has form and that the use of significant details is governed by an inner logic closely tied with the poet's vision or poetic intuition. In the first quatrain of the Shakespearean sonnet, the poet uses words that allude to Adam after the Fall: "disgrace," "outcast," "deaf heaven," "bootless cries," "curse my fate." These conjure up an image of hell. The poet, who is now associated with Adam and represents man, lists one simile after another to depict himself in various conditions in which he has feelings of discontent excited by the superiority or prosperity of others. At the end of the second quatrain, in which he is indirectly claiming that envy prevails on earth, he states that he suffers from lack of satisfaction in giving an impression that nothing can alleviate his suffering. After a slow-moving octave, the sestet, which brings about relief and joy, is set in heaven. The sharp contrast between the sense of doom portrayed in the octave and the dynamic movement in the sestet is meant to show how one's mental state can have an exalting effect. In a vibrant image that fills the sestet with an uplifting feeling, the poet compares himself to a lark, which stands for joy, freedom, singing, harmony, poetry, and divine inspiration. To establish a link between the octave and the sestet, Shakespeare resorts to internal music that is achieved through the repetition of certain words ("state," "heaven," "like"), ideas ("fortune," "rich," "wealth"; "disgrace," "curse"; "alone," "outcast"), and syntactic structures (adjectivenoun combinations: "outcast state," "deaf heaven," "bootless cries," "sullen earth," "sweet love"; present participle: "wishing," "desiring," "despising," "arising"; and past participle: "featured," "possest," "contented"). The image of "sullen earth" alludes to the story of creation and directly refers to a chaotic state of existence, the one portrayed in the octave. The transformation from the "outcast state" to a state in which he "sings hymns at heaven's gate" is a reference to the sublime. Shelley's sonnet tells a highly ironic story of Ozymandias, a once powerful king who seeks immortality, but is remembered on account of his statue, which is a heap of broken stone in a pathetic state in the desert. This statue of Ozymandias shows the skill of the sculptor who captures the essence of life rather than the grandeur or the glory of the king. The closeup on the "visage" and the "pedestal" indicates that looking closely one understands the meaning of life. Ironically, the visage, which is the "power mask" that exhibits a haughty, tyrannical expression, is half-buried and although the king is not alive, his "words appear" on the pedestal. This is an epiphanic moment in which the king, who stands for history, uses direct speech. Yet the fact remains that he is being remembered because of the sculpture not because of his "works."

Teaching 'Form' in English Verse to Arabic Poetry Readers

159

Form viewed from the end Nothing can clinch the idea of form in a short poem more than the analysis of the end. Just as with the foundation, a clear ending gives shape to the whole. When the end is not haphazard or a random line thrown in for good measure, but is an inevitable consequence of all that has proceeded from stanza to stanza, the students will feel that there is an end. And when they feel an end, they will sense that there must have been a beginning and a middle. Aristotelian though this concept may be, there is no harm in utilizing it for the approximation of the idea of form at this initial stage. Variations of the Aristotelian pattern may be given at later stages, but I believe they should be avoided at this early stage. The couplet is epigrammatic and inclusive in this Shakespearean sonnet. The experience happens in one minute: initially, he is unhappy and wants what others have, but in the end he is content and regrets and scorns his previous desires. "Sweet love" enables him to break the shackles of depression that chained him to "sullen earth." The idea is that in the absence of "sweet love" he is in an "outcast state": he "trouble[s] deaf heaven" with his "bootless cries." The end thus takes us back to the beginning, to the image of Adam. The reader experiences a sense of relief that the image of the "bootless cries" to "deaf heaven" is now transformed to that of "the lark at the break of day arising/ From sullen earth, [that] sings hymns at Heaven's gate." With the allusion to Adam and the story of creation, it is now possible to infer that the "sweet love" could be a reference to the love of God. Unlike Shakespeare, Shelley's final statement is disturbing: art and nature remain, not man. True, the statue is "colossal," but it is also a "wreck" in a state of "decay" because of time, which is what is awesome. The desert too is awesome: it is the master, with the statue representing only a spot. The "sands" here are described with terms that generally refer to the sea: "sunk," "wreck," and "breathless." This metaphor depicts a fearful picture of the sand that is "bare," i.e., lifeless, especially since this is where human beings are buried. The idea that is comforting, however, is that the creator of this poem, the poet, is the one who really survives.

Conclusion Discussing these four steps in focusing on form proved to be a rewarding experience for me and for the students. Observing them struggle to identify the main subject and verb in the Shakespearean sonnet, and in so doing become aware that all 14 lines constitute one syntactic structure (that is not a run-on sentence!), and in Shelley's sonnet to compare the

160

Teaching 'Form' in English Verse to Arabic Poetry Readers

nine-line-long sentence with the one that is only three words, made me realize that this is an exercise that they will not easily forget. Moreover, apart from experiencing the vitality of a different culture's poetic tradition, Arabic-speaking students reached another goal: comprehending the two general principles that reading and appreciation of English poetry can lead to the improvement of one's own writing skills and that there is a bond between form and content.

Notes 1 Randa Abou-Bakr, lecturer in the Department of English Language and Literature, Cairo University, translated all the Arabic verses in this essay. To her, I am truly grateful. 2 "Form; 1. The shape or structure of anything; also, a mold for giving a desired shape to anything. 2. A body, esp. of a human being. 3. A set way of doing something. 4. A prescribed order of words or action: hence: a Conduct regulated by convention or custom; also, empty ceremony; as, merely a matter of form. b A social convention or manner of behaving. c manner of doing something; as, his form in diving is bad. 5. Orderly arrangement; also, a special manner of arrangement: also, the sonnet form in poetry. 6. Kind; species; variety; as, the different forms of carbon. 7. Physical and mental condition, as of an athlete; as, he was at the top of his form. 8. A long seat or bench, esp. in a schoolroom; hence, a class of students in a school; as he was in the fourth form. 9. A printed document with blank spaces for insertion of required information; as, an income tax form. 10. Grammar. Any of the changed spelling or pronunciations a word may take in declension or conjunction or compounding; as, a passive form. 11. Philosophy. a The essential nature of a thing as distinguished from the matter in which this nature is embodied; that in a thing which it has in common with every other thing called by the same name; thus, the form of a (or any) diamond is pure crystallized carbon; opposed to matter. b The pre-existing idea of which all actual things are copies. 12 In printing, type arranged and fastened in a chase. "Syn. Form, figure, shape, confirmation, configuration, outline, contour, profile. In general, form is the aspect under which a thing appear; figure is oftener form as defined by outline; Shape is more colloquial than form; it often suggests form as given or acquired. Confirmation and configuration denote form as dependent on disposition of parts. Outline suggests the bounding line of a figure; contour connotes rather body or mass as (esp. gracefully) outlined. Profile is esp. outline in side view. See CEREMONY." "Sonnet; A rhymed poem of fourteen lines (usually iambic pentame-

Teaching 'Form' in English Verse to Arabic Poetry Readers

161

ters) expressing a single idea or sentiment. The type of sonnet, the Italian sonnet, called also Petrarchan, or regular, sonnet, is quatrains (often rhyming abba, abba), and sestet, composed of two tercets (commonly rhyming edc, edc or ede, ede}. Milton and Wordsworth wrote sonnets of this type, but their division between parts is marked only by a change in the rhyme scheme. Another type, English sonnet, called also Elizabethan, or Shakespearean, sonnet, has four stanzas, or three quatrains and a couplet (commonly rhyming abab, cdcd, efef, gg)." 3 This is the Italian term that was used until Gascoigne coined the term sonnet to refer not to short poems in general but to poems "whiche are fourtene lynes, every line conteyning tenne syllables" (1.55).

References

Badawi, M. M. 1985. Convention and reward in modern Arabic poetry. Modern Arabic literature and the West. London: Ithaca Press. . 1993. A short history of modern Arabic literature. Oxford: Clarendon Press. Clark, T. 1989. Embodying revolution: the figure of the poet in Shelley. Oxford: Clarendon Press. Enani, M. M. 1986. An anthology of the new Arabic poetry in Egypt. Cairo: General Egyptian Book Organization. Gascoigne, G. 1575. Certayne notes of instruction concerning the making of verse or ryme in English. Reprinted in Elizabethan critical essays (1904), edited by G. G. Smith. London: Oxford University Press. Heinrichs, W. 1973. Literary theory: the problem of its efficiency. In Arabic poetry: theory and development, edited by G. E. von Grunebaum. Wiesbaden: Harrassowitz. Heninger, S. K., Jr. 1994. The origin of the sonnet: form as optimism. In The subtext of form in the English Renaissance. University Park, PA: Pennsylvania State University Press. Moreh, S. 1998. Studies in modern Arabic prose and poetry. Leiden: Brill. Spiller, Michael R.G. 1992. The study of the development of the sonnet. London: Routledge. Van Gelder, G. J. H. 1982. Beyond the line: classical Arabic literary critics on the coherence and unity of the poem. Leiden: Brill.

Dialectal Analysis of Freshman Writing Students' Attitudes toward American and British Dialects Christopher W. Horger The 19th and 20th centuries have seen the spread and rapid global adoption of English as a lingua franca throughout the world. Much to the chagrin of many British speakers of English, the preeminence of American (U.S.) English since World War II seems to threaten the once sacrosanct position of British English, particularly within business, communications, and education. Due in large part to U.S. world dominance in mass culture, primarily in publishing, film, and music, many people seem to want to talk like an American. But not everyone is rushing to adopt American standards of speech. Modiano (1996) claims that most language schools in Europe still cling to British linguistic forms in their curricula, including British phonology, syntax, and lexicon. This retention of the British standard is often met with great resistance from students, who see it as limiting in terms of the number of people they will ultimately be able to communicate with. Indeed, there seems something absurd about forcing on students a dialect (Received Pronunciation) that is spoken by a small minority within England. As Modiano points out, 70% of all native English speakers speak with an American dialect. To become fluent in English as a foreign language and still be frequently unintelligible to the majority of native English speakers does not sound like a very good bargain. The reason for such strict maintenance of this dialect in European language schools is tradition and the notion of prestige. Tradition maintains that British English is "better," "cleaner," "more intelligent," and "more pure," as opposed to American English, which is "lazy," "vulgar," and "unsophisticated." Even Prince Charles is on record as attacking the American dialect as "corrupt" (Modiano, 1996). But these traditional attitudes toward dialects die hard. Indeed, even some Americans see their own speech as inferior to British English. And while many L2 English learners seem desirous of adopting an American dialect, they too cling to

Dialectal Analysis of Freshman Writing Students' Attitudes

163

some of these traditional notions about dialectal superiority and inferiority when it comes to British vs. American speech. The following study grew out of a desire to measure some of the attitudes the American University in Cairo (AUC) students hold toward the two different dialects. Students at AUC represent particularly good subjects for study, since most are products of the various private language schools in Cairo. And although a considerable number clearly speak with a British accent, the majority uses the American dialect. What is more noteworthy, many of them adopt the American dialect in the course of their brief years studying at AUC. Whatever remnants of their British linguistic training are still intact on entering AUC seem to be almost wiped out by the time most are seniors. The questions guiding this study are as follows: • How linguistically aware are AUC students of the differences between British and American dialects? • Is there a positive correlation between proficiency in English and the ability to differentiate between British and American dialects? • What are some of the affective reactions to the two different dialects? • Which dialect do students prefer as a medium of instruction and to speak in general, and what are some of the reasons for this preference?

Review of Literature Within sociolinguistics, the history of attitudes toward language is a relatively recent one, slightly more than 30 years old. Beginning in the 1960s and coming into full bloom in the 1970s and 1980s, the study of attitudes toward various dialects and whole languages has taken on many forms in many regions. Most of the studies were done within the English speaking world, particularly in North America, although some notable exceptions exist. The three major focal points of attitudinal studies in North America were Black English, Latino/American, and Anglo/French Canadian, where different dialects and languages come into frequent contact. Measurements of attitudes often included such things as perceived levels of intelligence, understandability, education, and class. More recently, studies have expanded to include Japanese, Arabic, and Israeli attitudes toward English. Some of the better-known landmark studies in language attitudes are summarized below. One of the earliest studies was conducted by Tucker and Lambert (1969) who set out to measure attitudes toward standard American English and Black English. They used as their subjects three groups of college students (northern white, southern white, and southern black) to evaluate six dialects of American English: television network, educated white south-

164

Dialectal Analysis of Freshman Writing Students' Attitudes

ern, educated black southern, Mississippi peer, Howard University, and New York alumni. Respondents on the whole rated the network speaker most favorably, but when it came to ethnic and regional dialects, the northern white and southern black groups all rated the educated southern black dialect more favorably, whereas the southern white group rated itself higher. With the exception of the favorable attitudes toward network speech, southern whites and southern blacks exhibited an "us and them" attitude toward each other, finding their own dialect more pleasant. D'Anglejan and Tucker (1973) conducted a similar study in Quebec. Using both Anglo and French speakers as judges, the researchers measured attitudes between European French and two dialects of French Canadian. They hypothesized that the prestige dialect (European French) would rate higher in status but not in things that mark solidarity: friendliness, trustworthiness, and the like. The general consensus favored the European French in terms of intelligence, and socioeconomic status, but surprisingly also found them to rank higher than their Canadian counterparts in levels of friendliness and trustworthiness. Similarly, Carranza and Ryan (1975) measured attitudes among Anglos and Mexican-Americans toward their own dialects and found that both groups favored the commonly accepted high-prestige form (AngloEnglish) over the low-prestige one (Latino) in terms of intelligence, ambition, education, etc. They also found that Mexican-Americans valued the Anglo dialect in terms of friendliness, sociability, etc. More recently, attitudinal studies have diversified to include measurements of ethnic groups to whole languages. Kraemer and Birenbaum (1993) conducted a study in an Israeli context to discover correlations between ethnicity and gender and the three languages of Israel: Hebrew, Arabic, and English. They hypothesized that ethnicity would determine preferences toward Hebrew and Arabic, but that preferences toward English, since it is considered a neutral language in Israel, would find a correlation only in gender, and that females would favor it over males. They found that ethnicity determined not only preferences toward Hebrew and Arabic, but English as well; the Hebrew-speaking students rated English more favorably than the Arab students. The proffered interpretations for this finding were both economic and political. More recently Levin et al. (1994) conducted a study to measure attitudes toward levels of formality, asking respondents to rate a speaker who read a passage with a proliferation of Latinate English words against a reader reading with a primarily Germanic lexicon. As expected, the results showed that the highly Latinate version was measured positively in terms of intelligence, ambition, and social stature, as opposed to the Germanic version, which generally scored low on these constructs. Conversely, however, the speaker who used the

Dialectal Analysis of Freshman Writing Students' Attitudes

165

more Germanic lexicon was valued for his sincerity, friendliness, and trustworthiness, while the Latinate one was not. Giles et al. (1995) took a different approach in studying attitudes toward Anglo- and Latino-accented speakers by adding the level of persuasiveness to the semantic differential scale. Using a matched guise technique, the researchers recorded a bidialectal speaker to read two brief argumentative passages in each dialect, one an argument in favor of the English-only Movement (EoM), the other against it. The researchers employed questionnaires that asked subjects to measure the recorded voice according to the level of dynamism, superiority, and attractiveness. The results confirmed the hypothesis. Subjects who had originally been in favor of the EoM were largely convinced to alter their view after hearing the Anglo version against the EoM, whereas the Latino-accented version did not carry as much persuasive weight. Curiously, those who were against the EoM were not strongly affected by the anti-EoM argument when it was articulated by the accented recording, and indeed showed a slight movement toward altering their view after hearing the Anglo speaker. The implications of this study are quite clear and frightening. Mainstream dialects carry more intrinsic authority than do minority dialects. Often it is not so much the meaning in the message, but rather the dialectal quality of the message that persuades people. The final two studies in this literature review have a direct bearing on the study reported here, since they attempt to measure L2 English speakers' cognitive and affective attitudes toward various English dialects. The first, by Alford and Strother (1993), was an attempt to measure native English speakers' affective reactions toward regional American dialects against those of normative speakers. They wanted to know whether L2 speakers of English could detect regional dialects as most American L1 speakers can. Furthermore, they hoped to discover whether L2 speakers shared some of the same attitudes and perceptions toward these dialectal differences as Lls did. Six readers, three males and three females, represented the three dialects: northern (New York), southern (South Carolina), and midwestern (Illinois). The respondents listened to each recording, and then evaluated personality characteristics on a seven-point Likert scale. The personality constructs measured were levels of intelligence, family training, education, ambition, confidence, professionalism, trustworthiness, sincerity, friendliness, patience, gentleness, and extroversion. The findings clearly indicate that L2 English speakers are capable of detecting differences in dialects and ascribing to them the same characteristics as L1 speakers. The question that arises from these results is, how do L2s acquire these attitudes, especially since many had been in the States for less than a year? Are they media influenced? Or do these perceptions come from immersion in the regional dialect? More studies need to be

166

Dialectal Analysis of Freshman Writing Students' Attitudes

conducted to shed light on such linguistic attitudes. The final study reviewed most directly ties in with the present study and was the guiding inspiration for it. Al-Kahtani (1995) conducted an attitudinal study among 14 Saudi Arabian students studying in the United States to measure perceptions toward three dialects: standard American English, Black English, and Indian English. He hypothesized that correlations exist between the independent variables of L2 English speakers' level of proficiency, motivation for learning, and age and the dependent variables of character traits perceived in the various dialects. The general results indicated a positive preference for Standard American dialect. Black English scored a bit lower, and Indian English consistently scored the lowest. Furthermore, the combined means suggest a marked preference for standard American English as a medium of instruction and in overall linguistic quality. In terms of the hypothesized correlations, Al-Kahtani found a positive correlation between proficiency and age, and desirability of standard American English. The implications of this study, according to Al-Kahtani, are that ESL students do establish dialectal preferences and attitudes similar to native mainstream speakers, often in a short period of time. Clearly, social biases and prejudices permeate language at all levels of proficiency. And as will be demonstrated, students at the American University in Cairo (AUC) are no exception to this phenomenon.

Procedure All of the above studies employed some form of a semantic differential scale to measure both cognitive and affective reactions toward the various dialects and languages. Hence, this study incorporates this research design with minor modifications. Since one of the purposes was to discover any traditional attitudes held toward British English (BE) and American English (AE), I drew on some of the better known stereotypes of BE and AE in putting together the semantic differential scale. For example, a common stereotype of British speakers is that they are more sophisticated and cultured, yet often arrogant. Americans, on the other hand, are often thought to be less cultured and sophisticated, but more friendly. The questionnaire distributed contained three parts (see the Appendix). The first part contained eight personality constructs to which subjects were asked to rate in response to the speaker's reading. Six traits were affective (in the sense that students were asked to respond intuitively): level of intelligence, sophistication, friendliness, class, arrogance, and culture; two, level of education and articulateness, tried to measure a more tangible trait. Each personality trait could be ranked on a five-point scale (e.g., from very intelligent to very unintelligent). The second part of the questionnaire consisted

Dialectal Analysis of Freshman Writing Students' Attitudes

167

of personal data designed to elicit information about the subject's cognitive reactions to the two dialects. The third part was a lexical questionnaire designed to elicit either an American or British form. Students were given 10 sentences with a key word omitted. Below each sentence were four multiple-choice answers that contained at least one commonly used standard American lexical item and one British lexical item, e.g., apartment/flat. Respondents were asked to choose the form they would most likely use. A portable cassette recorder was used to record four speakers, two British and two American. The two British speakers consisted of a standard British dialect (in this case, B.B.C. London-based British) and the other was a northern English (Manchester) dialect. The two American speakers consisted of a standard American (TV network) dialect and a more highly marked Wisconsin dialect. Regrettably, the Manchester and Wisconsin speakers were dropped from the study as a result of time and space constraints. (Out of curiosity, however, I did play these two dialects after the questionnaires were complete and asked both groups to guess at the dialectal origin. Most could not place the Manchester dialect within England and none guessed the Midwest for the Wisconsin dialect.) The speakers all read the same short passage from a recent Cairo Times article about al-Horreya Cafe/bar. The average time duration for each speaker was roughly a minute and a half. The speakers were encouraged to read as naturally as possible in their own dialect. The study was conducted at two different times: May of 1998 and February of 1999. The first time did not include part three, the lexical questionnaire; thus, the sample is small.

Methods The subjects for this study came from four of my own freshman writing classes at AUC, two sections of English 112 and two of English 113. They were a sample of convenience and should in no way be considered representative of the larger AUC student body. Nevertheless, they do provide a somewhat heterogeneous sample, since they run the gamut of varying levels of proficiency. Another virtue of this sample is that they already comprise two separate levels of English proficiency. English 112 is the entry level writing course required of all AUC students. As such, many often exhibit poor English language skills, more so in writing than in speaking. The total number of English 112 students in this study was 29. The English 113 students are usually much better speakers and writers than English 112 students, by virtue of the fact that they have had to demonstrate a certain level of English proficiency to pass from 112 into 113. The total number of English 113 students was also 29.

168

Dialectal Analysis of Freshman Writing Students' Attitudes

Students were prepared for the questionnaire/testing a week in advance to ensure a good turnout. I explained to them the nature of the study, but to insure objectivity, I kept my own feelings quiet. On the day of the study, I carefully explained the questionnaire and asked them to respond as honestly as possible. They were not told who the speakers were or where they came from. I played the two recorded speakers and asked them to respond to the first part (the semantic differential scale) before having them fill in the second part on personal attitudes toward the two dialects. After compiling all the data, I first determined ability to discern between dialects. Table 1 shows these results. Because there were surprisingly quite a few who could not discern between British and American dialects, and who had no sense of linguistic differences, their responses were pulled from the rest of the study. It seemed fruitless to include someone's reactions to the semantic differential scales, someone who gave very low ratings to both speakers on most items, and then claimed those very dialects would make excellent media for instruction. Thus, only the 16 subjects who were able to accurately differentiate between British and American dialects were included in the rest of the data analysis. All items were given a mean score based on a five-point scale.

Findings The first question to answer was how linguistically aware students were of the differences. Could they distinguish between a British and American dialect? Table 1 shows the overall number of participants from both classes and the combined number and percentage of those who could distinguish between dialects. As we can see, 45 out of a total 58 students (77.58%) were able to discern between dialects. If we were to answer the first question according to the results in this table, the answer would be "somewhat aware." Table 1. Profile of study population. Study Participants

No.

English 112

29

19

65.51

English 113

29

26

89.65

Total

58

45

77.58

Dialect ability*

%

*Dialect ability = the ability of students to distinguish between the British and American dialects. More revealing, however, are the differences between the two levels of English classes. Clearly, the English 113 classes were better able to dis-

Dialectal Analysis of Freshman Writing Students' Attitudes

169

tinguish between dialects. In the English 112 class only 19 out of 29 students could place the speaker's country of origin (65.51%), whereas 26 out of 29 students (89.65%) could in the 113 class. This would seem to support the hypothesis that language proficiency correlates positively with linguistic and dialectal awareness. Following al-Khatany (1995), affective responses to personality traits were identified. Some modifications to al-Khatany's original series were made to account for perceived stereotypes of British and American dialects. Students were asked to characterize the speakers as intelligent, educated, articulate, sophisticated, friendly, upper class, arrogant, and cultured, and to rate these traits on a five-point scale (see Table 2). Table 2. Judgment of personality traits of readers by entire study population (N = 45). Personality Traits* (meanst+) Dialect American

INT 3.77

EDU 4.44

ART 4.04

SOPH 2.84

FR 3.73

UC 3.68

ARR 1.88

CUL 3.80

Briti4.06 4.64 3.73 4.11 2.26 4.11 3.71 3.93

* INT = intelligent; EDU = educated; ART= articulate; SOPH = sophisticated; FR = friendly; UC = upper class; ARR = arrogant; CUL = cultured + Based on a five-point scale.

In terms of some of the affective responses to personality traits, the findings are perhaps not surprising. I averaged all the personality data from Part I in the questionnaire into three major groups: the total population, the English 112 population, and the English 113 population (tables 3 and 4, respectively). In general students ranked the individual traits of the British dialect more favorably, with the notable exceptions of friendliness and articulateness. Invariably, the American dialect gets the higher score in these two categories, that is, American English sounds more friendly and is easier to understand. It is worth noting, however, that high scores for the personality traits should not always be equated with a positive attribute. For example, arrogance is seldom a highly valued social construct. In effect then, the low scores for the American dialect under the category of arrogance speak more in its favor. To sum up the findings of Part I, students generally find British English to be more intelligent, educated, sophisticated, upper class and cultured, whereas American English is more articulate, friendlier, and less arrogant.

170

Dialectal Analysis of Freshman Writing Students' Attitudes

Table 3. Judgment of personality traits of readers by English 112 students (n = 19). Personality Trait* (meant) Dialect American British

INT 3.89 4.05

EDU 4.30 4.76

ART 4.00 3.36

SOPH FR 2.89 3.63 3.89 1.94

UC 3.47 4.10

ARR 1.84 3.89

CUL 3.89 4.89

* INT = intelligent; EDU = educated; ART= articulate; SOPH = sophisticated; FR = friendly; UC = upper class; ARR = arrogant; CUL = cultured + Based on a five-point scale. Curiously, it is under the three categories of sophistication, friendliness, and arrogance where the most dramatic and statistically significant differences in mean scores appear. The other five personality traits show marginal differences in scores, but sophistication, friendliness, and arrogance consistently show more than a whole percentage point in difference (the exception being English 113 students' scoring of sophistication, Table 4). This would seem to suggest that these three traditional stereotypes have the greatest influence on student's perceptions of dialectal differences. Another explanation might be that both speakers were actually similar in their levels of education, intelligence, articulateness, and so on, and students were not able to perceive significant differences. Table 4. Judgment of personality traits of readers by English 113 students (n = 26). Personality Trait* (meant) Dialect American British

INT 3.69 4.07

EDU 4.30 4.53

ART 4.07 4.00

SOPH FR 2.80 3.80 3.61 2.50

UC 3.47 4.11

ARR 1.84 3.57

CUL 3.73 3.96

* INT = intelligent; EDU = educated; ART= articulate; SOPH = sophisticated; FR = friendly; UC = upper class; ARR = arrogant; CUL = cultured + Based on a five-point scale. Questions 9 and 10 in Part I of the questionnaire provided the data for tables 5, 6, and 7. These questions asked students to rank both dialects in terms of appropriateness for educational instruction and overall linguistic quality. These findings are surprising and a bit ironic. Table 5 shows the mean scores for the entire population of participants. In terms of overall

Dialectal Analysis of Freshman Writing Students' Attitudes

171

linguistic quality, BE scored 4.06% out of a possible 5.00%, whereas AE scored 4.00%—a difference that is not significant. What is significant is the dialectal preference for educational instruction. After showing slight preference for the British dialect in overall linguistic quality, students generally seemed to agree that the American dialect was more suitable for classroom instruction. Table 5. Attitude of entire population (N = 45) toward educational and linguistic value of each dialect. Trait (meant)

Dialect American British

Appropriateness for education

Linguistic quality

3.51 2.77

4.00 4.06

+ Based on a five-point scale. If we take a close look at tables 6 and 7, which represent the preferences of English 112 and 113 students respectively, we find some striking differences. Table 6 shows English 112 students ranking AE higher in linguistic quality as well as educational appropriateness. But their desire for AE as a medium of instruction over BE is dramatically significant: almost two points greater than BE on a five-point scale (roughly 40%). Table 7, on the other hand, shows a considerable reversal of this trend among 113 students. They ranked BE marginally higher in both overall quality and appropriateness for instruction. Table 6. Attitude of English 112 students (n = 19) toward educational and linguistic value of each dialect. Trait (meant)

Dialect American British

Appropriateness for education

Linguistic quality

4.0 2.05

4.0 3.63

+ Based on a five-point scale.

172

Dialectal Analysis of Freshman Writing Students' Attitudes

Table 7. Attitude of English 113 students (n = 26) toward educational and linguistic value of each dialect. Trait (meant) Appropriateness for education Dialect American British

3.15 3.30

Linguistic quality 4.00 4.38

+ Based on a five-point scale. How do we account for these discrepancies? At least two interpretations for these findings can be offered. The first is that English 112 students are not yet confident in English and therefore have a low ambiguity tolerance. Since most students perceive AE as clearer and more understandable, it stands to reason that less proficient students would cling to that which is more straightforward and comprehensible. The second explanation has to do with the sample size of this study. Only 26 English 113 students were included in this study, half from the spring 1998 class and half from the spring 1999 class. When this study was first conducted in 1998, there was a unanimous preference for AE as a medium for instruction. After including the 1999 sample, however, the pendulum swung in the opposite direction. Notably, 11 of the 15 students in the 1999 class have been to British schools, an unusually high concentration for one class. Evidence of this group's British preferences is reinforced by findings from the lexical questionnaire, which shows a high frequency of British forms, e.g., interval and surgery were preferred over intermission and office. Still, the primary inclination among most students is toward American English for instruction and as a dialect to speak. One wonders why students would favor a dialect that they generally considered inferior. To answer this question I returned to the open-answer segment from Part II of the questionnaire. The majority of respondents claimed that American English is easier to understand, that it will better prepare them for a career in the professional world, and that almost everyone else is speaking it. A few even said it just sounds "cool," and that exposure to American film and music has a huge influence on their choice of dialects. The final findings in this study came from Part III, the lexical questionnaire. As was stated earlier, this was only administered to the spring, 1999, groups; thus only 30 participants were included: 15 from the

Dialectal Analysis of Freshman Writing Students' Attitudes

173

English 112 class and 15 from the English 113 class. A quick glance at Table 8 shows a general preference for American lexical items, but it is revealing to see how frequently British items were chosen. The fact that words such as interval, surgery, rubber, pavement, and zebra crossings were chosen at all suggests a stronger British influence than was anticipated. Clearly, British English is alive and thriving at AUC. Table 8. Lexical choices of English 112 (n = 15) and English 113 (n = 15) study subgroups. American (112/113)* Intermission 4/6 Office 7/6 French Fries 12/11 Eraser 10/9 Sidewalk 6/9 Trunk 11/11 Crosswalks 12/8 Apartment 14/11 Elevator 9/8 Great 13/10

Other

British Interval Surgery Chips Rubber Pavement Boot Zebra crossings Flat Lift Smashing

6/8 5/8 2/2 4/6 9/6 1/2 1/5 1/4 4/5 2/4

5/1 3/1 1/2 1/0 0/0 3/2 2/2 0/0 2/2 0/1

* Number of students from English 112/English 113 choosing the particular lexical item.

Conclusions and Recommendations The findings of this study need to be taken cautiously. The sample was too small to generalize to the larger AUC population. Still, one can not ignore the clear correlation between proficiency in English and the ability to recognize dialects. In terms of how AUC students pick up some of the traditional notions and stereotypes about British and American English, it might just be the result of mass media, language school training, or even Egypt's British colonial past. This fondness for a glorious British past could certainly influence the way people regard the two dialects, and traditions in Cairo (linguistic or otherwise) seldom fade quickly. Besides, American influence in Egypt is relatively recent and still on the rise. Attitudes toward languages ultimately evolve like languages themselves. A similar study conducted 20 years from now might find students claiming British English to be hopelessly archaic, provincial, and rustic, while

174

Dialectal Analysis of Freshman Writing Students' Attitudes

American English may get high grades for cosmopolitanism, articulateness, and intelligence. Future dialectal studies certainly promise to shed light on the complex issue of language attitudes.

Appendix: Questionnaire on Dialects Part I. Instructions: Listen carefully to the different speakers. Feel free to silently read along with them from the handout. After hearing each reader, place an X in the appropriate blank. For example, if you think the speaker is very intelligent, you should mark it as follows: : : : : Unintelligent Intelligent: X If you thought the speaker was very unintelligent, mark it as: Intelligent: : : : : X : Unintelligent If you had no opinion or could not tell, mark it as: Intelligent: : : X : : : Unintelligent If you thought he was somewhat intelligent, but not very intelligent, mark it as: Intelligent: : X : : : : Unintelligent If you thought he was somewhat unintelligent, but not very unintelligent, mark it as: Intelligent: : : : X : : Unintelligent Speaker #1: Do you think the speaker is

?

1. Intelligent: : : : : : Unintelligent 2. Educated : Uneducated 3. Articulate : Inarticulate : Unsophisticated 4. Sophisticated 5. Friendly : Unfriendly : Lower Class 6. Upper class 7. Arrogant : Not arrogant 8. Cultured : Uncultured 9. Do you think this5 form of English should be used as a medium of instruction? : : : : : Disagree Agree ? 10. Do you think thatat the speaker used : : Poor English Good English 11. Do you know what kind of English the speaker was using? yes no 12. If yes, specify

Dialectal Analysis of Freshman Writing Students' Attitudes

175

Speaker #2: Do you think the speaker is 1. Intelligent Unintelligent 2. Educated Uneducated 3. Articulate Inarticulate 4. Sophisticated Unsophisticated 5. Friendly Unfriendly 6. Upper class Lower Class 7. Arrogant Modest Uncultured 8. Cultured 9. Do you think this form of English should be used as a medium of instruction? Agree : : : : : : Disagree 10. Do you think that the speaker used ? Good English : : : : :_ : Poor English 11. Do you know what kind of English the speaker was using? Yes No 12. If yes, specify

Part II. Instructions: Please read through the following questions and respond as honestly as possible. l.Name:. 2. Age:_ 3. Years studying English:_ 4. Major: 5. What is your goal in learning English?

6. Do you use English for purposes other than school? Yes No 7. If so, what other purposes do you use English for? (examples: with your friends, family, when shopping, work or business, etc.)

176

Dialectal Analysis of Freshman Writing Students' Attitudes

8. Which language do you use most frequently? Please rank in order of frequency with 1 being the most frequent and 3 being the least. Arabic English Other 9. How aware are you of the differences between British and American English? Very aware Somewhat aware Not very aware 10. Which dialect do you think is more prestigious? American British Neither 11. State some of your reasons:

12. Which dialect do you think you come closest to using? American British Other 13. How did you acquire this dialect?

14. How would you describe the differences between these dialects?

15. Which dialect of English would you prefer to use? American British Other_ 16. State some reasons for this preference:

Part III. Instructions: Circle the letter to the answer which you think best fits the blank. 1. I went to see the movie Titanic last week and during the I ran into an old friend. a. intermission b. intervention c. interval d. introspection 2. The operation was minor and the doctor was able to take care of it in his a. cabinet

b. store

3. I ordered some a. pomme frittes b. chips

c. surgery

d. office

with my cheeseburger and shake. c. french fries d. fingers

4. "May I borrow your pencil? The a. eraser b. leaf

on mine is used up." c. rubber d. button

Dialectal Analysis of Freshman Writing Students' Attitudes 5. The a. field

177

is for pedestrians, and the street is for cars. b. sidewalk c. pavement d. tarmac

6. Most people keep a spare tire in the a. glove compartment b. trunk c. boot

of their car. d. shoe

7. In Cairo are often painted onto streets at intersections, but many drivers barely slow down for them. a. atlases b. crosswalks c. hotels d. zebra crossings 8. My cousin lives in a beautiful a. apartment b. compartment

in Maadi. c. pad ' d. flat

9. She had to walk up ten flights of stairs because the wasn't working. a. raiser b. ascensior c. lift

d. elevator

10. I think Braveheart was a a. perfidious b. smashing

d. egregious

film. c. great

in her building

References Alford, R. L., and Strother, J. B. 1993. Attitudes of nonnative speakers toward selected regional accents of English. TESOL quarterly, 24:479-95. Carranza, M. A., and Ryan, E. B. 1975. Evaluative reactions of bilingual Anglo and Mexican American adolescents towards speakers of English and Spanish. International journal of the sociology of language, 6:8-104. Giles, H., Williams, A., Mackie, D., et al. 1995. Reactions to Anglo- and Hispanic-American-accented speakers: affect, identity, persuasion, and the English-only controversy. Language and communication, 15:107-120. Al-Kahtany, A. 1995. Dialectic ethnographic "cleansing": ESL students' attitudes towards three varieties of English. Language and communication, 15:165-180. Kraemer, R., and Birenbaum, M. 1993. Language attitudes and social group memberships. International journal of intercultural relations, 17:437-149. Levin, H., Giles, H., and Garrett, P. 1994. The effects of lexical formality and accent on trait attributions. Language and communication, 14:265-274.

178

Dialectal Analysis of Freshman Writing Students' Attitudes

Modiano, M. 1996. The Americanization of Euro-English. World Englishes, 15:207-215. Shuy, R.W., and R. Fasold (eds.). 1978. Language attitudes: current trends and prospects. Washington, DC: Georgetown University Press. Tucker, G. R., and Lambert, W. E. 1969. White and Negro listeners reactions to various American-English dialects. Social forces, 47:463-468.

The Acquisition of the English Copula by Native Speakers of Lebanese Arabic: A Developmental Perspective Abdel-Hakeem Kasem One of the main objectives of theories of second-language (L2) acquisition is to account for the manner and order in which a second language is acquired and develop an explanation that is applicable across language-specific boundaries regardless of the learner's first language (L1). Most current theories of second-language acquisition assume that the learner's native language plays a role in the acquisition; however, what role the native language plays is something less certain and controversial. Some researchers are of the view that speakers of L1 are initially transferred to the interlanguage grammar, but given the appropriate input will ultimately be adjusted to the correct L2 setting (e.g., White's "transfer hypothesis" of 1987). On the other hand, other researchers believe that L1 serves as a "surrogate" Universal Grammar (UG) for the learner and that only those aspects of UG that are manifested in the native language will be acquired by the learner (e.g., BleyVroman's "fundamental difference hypothesis" of 1989). The main objectives of the present study are: (1) to investigate and establish the developmental sequences or patterns for the acquisition of English copula structure by native speakers of Lebanese Arabic in a classroom environment, and (2) to establish what role LI plays in the acquisition of L2. According to Wode (1976) and Meisel et al. (1981), the term developmental sequences implies that language learners go through a number of steps before achieving a targetlike proficiency in any L2 structure. These steps, more importantly, are not random but rather systematic. The changes, or variations, in a learner's language are the result of a number of operations, such as modification and generalization, that learners apply to linguistic structures as they gradually move closer to mastery of the target language. The present study draws on the assumption that L1 and L2 are acquired in the same way (L1=L2) and that L2 learners' errors are similar to those made by L1 learners. This assumption is substantiated by longitudinal data of Lebanese young adult learners acquiring the English copula in a classroom environment.

180

The English Copula by Native Speakers of Lebanese Arabic

Developmental studies focus on the learning process and the learner's strategies rather than simply on the order of certain morphemes. The view taken by researchers working within the developmental approach is that while still learning the grammar of the target language, second-language learners use forms that do not belong to either the second language or the native language. These forms are known as transformational forms (Dulay et al., 1982). A number of other researchers (e.g., Wode, 1976, 1978, and 1981; and Meisel et al., 1981) employ the term developmental sequences to refer to these same transitional constructions and the order in which they occur. Furthermore, researchers in developmental studies argue that the similarities between first language and second language (L1 = L2) far outweigh the differences. When differences exist, they are usually the result of the fact that adults are mentally more sophisticated and, therefore, can produce a wider variety of a particular structure than children learning the same language as their native language (Al-Buanain, 1987, p. 53). Several researchers, including Wode (1978), Hatch, (1978), Meisel et al. (1981), and Ellis (1994), found that errors observed in the transitional constructions produced by L2 learners bore no relation to their L1. These errors are intralingual in nature, resulting from a developing system, rather than interlingual resulting from the learner's L1 interference. The learner's language is treated as a linguistic system in its own right, rather than a distorted version of the target language system (cf., Larsen-Freemen and Long, 1991). The learner's language is also viewed as a dynamic system, which undergoes continual and constant change as learning unfolds and progresses toward target-language norms. The dynamic nature of the learner's language is, nonetheless, systematic, and it is possible to formulate rules and principles that account for its development from one stage to another. A dynamic linguistic system cannot be completely described in terms of categorical rules, however, but must permit the use of variable rules. Variability could result from any of a number of factors, such as transfer, age, motivation, context, and input. These variables have, in fact, an effect on the manner in which the learner's language evolves as we shall see below (see "Analysis of learners' errors," p. 185).

The Sample As this was a longitudinal study over a period of six months, it was not appropriate or feasible to deal with too many subjects. The sample for the present study involves a total of 10 newly arrived Lebanese-born students (five males and five females) randomly selected from Brunswick Language Centre in Melbourne, Australia, to provide data for this investigation. The 10-student sample was selected on the basis of their age, their

The English Copula by Native Speakers of Lebanese Arabic

181

ethnolinguistic background (i.e., L1 speakers of Arabic), their motivation for learning English (i.e., academic purposes) as well as their previous experience with the language (i.e., little or no exposure prior to beginning their intensive English course. Of the 10 subjects, eight were 13 to 15 years old. The remaining two subjects were 16 and 17 years old at the time when the study began (March, 1997). All the subjects had arrived in Australia shortly before entry into the study (<6 months). The subjects were all native speakers of Lebanese Arabic. They all had a very moderate knowledge of English at the time of data collection and they all spoke their L1 at home with their parents and other family members. The subjects were highly motivated and had very positive attitudes toward learning English. Indeed, they were very keen to improve their English-language skills as soon as possible so that they could join the Australian school system, and also be able to communicate more effectively with native speakers of English.

English-language experience Information about the amount of exposure to English the subjects had since they arrived in Australia was obtained through direct questioning of the learners, through a written questionnaire, and through my own observations during my regular visits to the language center. All learners were considered intermediate learners on the basis of their linguistic proficiency at the time of collecting the first set of data. Data obtained from the subjects revealed that the female subjects' exposure to and use of use of the target language (English) outside the classroom was extremely Limited in comparison with the male group, who had better exposure to L2. Subjects' (especially females') main exposure to English was confined to the classroom in the language center, which they attended for six hours per day, five days a week for a period of six months. During this period the subjects received formal instruction in the classroom. The subjects tended to use English at the language center only when they spoke to their non-Lebanese friends and to thek teachers, but when talking to each other during breaks and other recreational activities, they used their L1 for the most part. L1 was likewise used at home on most occasions, especially with parents and relatives, most of whom knew little or no English at that point. The fact that the female group had very little exposure to English outside the classroom may explain, as we shall see in the data analysis, why the female subjects did not perform as well as the male group during the three stages of data collection. This was quite evident in the number of errors produced and the accuracy rate achieved at the end of each stage. This observation seems to correlate with and confirm the findings of

182

The English Copula by Native Speakers of Lebanese Arabic

McLaughlin (1987), who reported on the importance and significance of contact with the target language in situations in which languages are learned in classroom settings and there is no daily contact with native speakers of the target language outside the classroom. Other researchers, such as Huntsbury (1972) and Ervin-Tripp (1974), have reached similar conclusions.

Data Collection The developmental nature of this study requires data that is produced in as natural and spontaneous a manner as possible. This type of data is ideal for establishing developmental stages and acquisition of structures. It is essential that the subjects are unaware of the structure under investigation so that their performance is as close to spontaneous as possible. Larsen-Freeman and Long (1991) argue that in the case of elicited data, it is desirable that the learners' performance occurs in a general context rather than in single sentences. This, it is hoped, should minimize learners' linguistic awareness, on the one hand, and maximize naturalness of data, on the other hand. At the time of the collection of the first set of data, the 10 subjects involved in this study had received approximately 240 hours of formal instruction in English. The main data-eliciting procedure was written compositions, both guided and unguided compositions, collected three times over a six-month period beginning in March 1997 and ending on August 27, 1997. Written data were collected on a bimonthly basis. The compositions selected for analysis reflected a sample of the learners' written production at three periods: stages 1, 2, and 3. A total of 30 compositions were collected and analyzed. The average length of each essay was 200 to 250 words. The benefit of using an essentially quasilongitudinal data set is that one is able to gain a full picture of the development process within a sixmonth period. As long as the learners are selected randomly and the main variables (such as previous exposure to the target language, motivations, age, and aptitude) are accounted for, the data reflect a continuum of developmental stages, which reflect the psycholinguistic realities of learning L2 syntax in a classroom environment.

Copula Structure in Arabic and English The following is a brief description of the structures under investigation. In both colloquial and standard Arabic, a sentence can correctly be formed without a verb. The copula "to be" is not commonly used in sentences of the following patterns, characteristic of English:

The English Copula by Native Speakers of Lebanese Arabic

183

+ to be + noun noun + to be + adjective + to be + adverb The optional use of (to be) in Arabic depends on whether the sentence is marked for time other than the present or not marked at all. If it is not marked, no form of "to be" is used. If it is marked, a past or future form of "to be" is required, as illustrated in the following examples: (Hussein, 1971, p. 83, and Abboud, 1975). 1. hadha bait-na this house our "This is our house" 2. hadha kan bait-na this be (past) house our "This was our house" 3. hadha haykun bait-na this be (future) house our "This will be our house" In Arabic the verb kana may correctly be translated as "to be" and it has some of the linguistic functions of "to be" of some other languages, yet these functions can also be performed without a form of kana, both in the sense that words, some of which are not even verbs, can be substituted and that no such words need to be used at all (Shehadi, 1969, p. 112). It is necessary to consider certain features of Arabic syntax with regard to the use of the copula. Arabic linguists usually divide the categorical attributive sentences into nominal and verbal. The first is composed of a noun (ism), which is the subject (mubtada'), and a predicate (khabar). The second one is composed of the verb (fi'l) and its subject (fai'l). In a verbal sentence no special connection is needed to relate the verb and its subject. However, it is in the nominal sentence that one finds the most notable feature of Arabic syntax. In many languages the juxtaposition of the words "the man" and "tall" would not form a complete, grammatical sentence. What is missing is a connective between the two words that brings them together as subject and predicate. In English the verb "to be" is usually the connective. But in Arabic the nominal sentence can occur without such a connective. Of course, this is not to say that the mere juxtaposition of "the man" and "tall" suffices to form a sentence. The relationship expressed in English as "the man is tall" is indicated in Arabic by putting the two parts

184

The English Copula by Native Speakers of Lebanese Arabic

in the nominative case to form a complete sentence. One would say in modern standard Arabic (MSA): 4. al-rajul-u tawil-un the man (nom.) tall (nom.) "The man is tall." Since it is not necessary to use a connective between subject and predicate in the nominal and verbal sentence, and since these two kinds of sentences constitute the class of assertoric sentences in which the connective is expected to appear, it would seem reasonable to assert that there is a complete absence of the copula in Arabic. However, it must be noted that it is possible, and in some cases required, to have a word that, strictly speaking, can be said to have a copulative function, for example, kana (cf. Shehadi, 1969, pp. 112-125).

Data Analysis and Discussion Method of data analysis The method used in the present study for identifying and describing developmental patterns is obligatory occasion analysis .This has been widely used by L2 acquisition researchers and is clearly described in Brown (1973, 1982). The basic procedure is as follows. First, samples of naturally occurring learner language are collected. Second, obligatory occasions for the use of specific target-language features are identified in the data. In the course of using the L2, learners produce utterances, which create obligatory occasions for the use of specific target-language features, although they may not always supply the features in question. Third, the percentage of accurate uses of the feature is then calculated by establishing whether the feature in question has been supplied in all the contexts in which it is required. A criterion level of accuracy can then be determined in order to provide an operational definition of whether a feature has been acquired. Usually the level is set at 80% to 90% to take account of the fact that even adult native speakers may not achieve complete accuracy (Ellis, 1994, pp. 74-75). It is important to point out here that the description and explanation of L2 learners' errors have always been both difficult and problematic (Ellis, 1994). Generally, learners' errors are usually divided into two main categories: developmental errors which can be related to the developmental stage and are typically explained in terms of a learner's proficiency or competence in the target language, and nondevelopmental errors usually accounted for from a performance viewpoint. The widely held view that

The English Copula by Native Speakers of Lebanese Arabic

185

learners' errors are essentially a result of cross-linguistic influences1 (i.e., L1 interference) is not supported in this study. The position taken in this paper treats learners' errors as key developmental signals that manifest an ever-evolving linguistic system in its own right, rather than a distorted version of the target language.

Analysis of learners' errors The following is a structural/linguistic analysis of the written data produced by the 10 learners during the three stages of data collection. All data were examined for the occurrence of copula constructions and analyzed from linguistic (structuralist) and communicative theoretical perspectives. The linguistic analysis of the data concerning the development of the copula structure is summarized in the tables below, along with some illustrative examples of errors found in the learners' compositions during the three stages of data collection: Data collected at the end of stage 1 strongly indicate early development of the copula structure. Learners' errors in the use of copula in this stage were classified into two major categories: (1) copula omission and (2) incorrect use of copula or auxiliary. Copula omission errors in stage 1 occurred in three environments: • After "there" constructions (error type 1). Examples: 5. There four people in my family. 6. In Brunswick there two language centres. 7. There a big tree in my house. • Np - [cop] - adj (error type 2). 8. My sister unhappy in Australia. 9. ...and she always sad and alone. • NP - [cop] - NP (error type 3). 10. My brother name Fadi. 11. Her husband name Anwar 12. My brother a doctor. The incorrect use of the copula or auxiliary errors (i.e., lack of agreement) was found primarily during stage 2 (error type 4). Let us consider some illustrative examples: 13.1 have four teachers their names is.... 14. There is two Vietnamese students in my class. 15. The boys names is.... 16. My brothers is going to school. 17. There are one Lebanese student in my class. 18. There are a big tree near my class.

The English Copula by Native Speakers of Lebanese Arabic

186

The following tables give types and percentage of copula errors per student. Table 1. Stage 1: Copula omission errors. Number of errors by type* Subject

Total use

(no.)

Errors

(no.)

Errors rate

(%of total use)

1

2

3

1

7

3

3

25

13

52.0

2

9

5

5

34

19

55.9

3

7

6

6

32

19

59.4

4

8

8

6

37

22

59.5

5

8

6

6

36

20

55.6

6

9

7

7

29

23

79.3

7

9

6

7

33

22

66.7

8

10

7

6

33

23

69.7

9

9

8

8

36

25

69.4

10

11

6

5

34

22

64.7

Total

87

62

59

329

208

63.2

*Errors were catalogued into three types during this stage: (1) after "there" constructions; (2) NP-cop-adj; and (3) NP-cop-NP. t Total use refers to the number of times the copula was used correctly plus the number of times it was omitted. Typical errors observed in the learners' compositions may elucidate the relation between such "errors" and the overall interlanguage developmental process. The methodology of this study rests largely on two assumptions relating to a learner's development of linguistic accuracy in the target language: (1) learners have reached a particular developmental stage if they make few or no errors of a given syntactical structure; and (2) the more frequent certain errors are, the further the learners are from attaining a particular stage. If one accepts these assumptions, the succession of structures that are mastered constitute a developmental pattern, that is, a succession of phases of learning toward mastery of new structures. A brief look at the specific error types in Table 1 reveals that omission of the copula was the greatest problem to our learners. Errors of omission proved to be the most frequent (208 errors out of 329). Error rates ranged from 52% to 79%, yielding accuracy rates of 21% to 48%. A careful analy-

The English Copula by Native Speakers of Lebanese Arabic

187

sis of copula omission in examples 5-12 above may suggest that these errors result from cross-linguistic influences (L1 transfer) since these sentences mirror L1 structure. This view is not supported in this paper. Despite the fact that examples 5-12 may be viewed as a direct translation of Arabic structures, second-language acquisition research (Huang and Hatch, 1978; Itoch and Hatch, 1978) has shown that copula errors are common in the emerging second language of speakers of virtually any native language. For example, the omission of the copula in cases such as "My sister unhappy in Australia" and "...my brother name Fadi" are errors made not only by Arabic speakers, but also by speakers of Spanish, Japanese, Russian, Chinese, and other languages. For example, speakers of Spanish, Japanese, and other languages that, like English, have copula verb forms frequently omit forms such as am and is. The pervasiveness of copula-omission errors has thus been among the most significant counterarguments against the importance of transfer. Moreover, the transfer explanation for such errors seem questionable in light of the fact that omission of the copula also occurs in the speech of children learning English as their native language, for example, "That a kitchen" (Brown, 1973). This suggests that copula-omission errors are nothing more or less than indicators of developmental process found in both L1 and L2 acquisition, and accordingly this type of error is termed developmental. Table 2. Stage 2: Incorrect use of copula . Subject

Number of errors of type 4 (incorrect use of copula)*

Total use (no.)

Errors rate (%of total use)

1

5

18

27.8

2

4

16

25.0

3

6

24

25.0

4

6

23

26.1

5

8

19

42.1

6

9

17

52.9

7

7

13

53.9

8

10

19

52.6

9

12

19

63.2

10

10

17

58.8

41.6 Total 77 185 * This figure tracks only errors of type 4 during stage 2, since all participants stopped omitting the copula at this stage. t Total use refers to the number of times the copula was used correctly as well as incorrectly.

The English Copula by Native Speakers of Lebanese Arabic

188

Incorrect use of the copula was the second type of error found in the learners' interlanguage in the data collected in stage 2 (see Table 2). Errors of this type were found in 77 cases out of 185, with error rates ranging from 63% to 25% (for accuracy rates of 37% to 75%). Examples 13-18 illustrate how learners use the incorrect form of the copula. The argument one would present for type 4 errors (the incorrect use of copula), such as those in sentences 13-18, is that they result from the differences between the verb structures of the target language and the source language. Although learners at this stage seem to be aware of the need to use the copula structure, apparently they have not yet internalized the rule for it. These were clearly developmental errors attributable to intra-English confusion, and as far as the structure of these sentences is concerned, these learners essentially followed a universal pattern of sentence development: language learners, regardless of their L1, tend to use the incorrect form of copula in the early stages of development. Table 3. Stage 3: Type and percentage of copula errors per student. Total

Number of errors by type* Subject

use

(no.)

Errors

(no.)

Errors rate

(%of total use)

1

2

2

3

1

0

1

1

2

28

4

14.3

2

0

0

0

1

22

1

4.5

3

1

2

1

1

33

5

15.2

4

0

1

0

4

30

5

16.7

5

1

0

2

1

26

4

15.4

6

2

2

0

2

22

6

27.3

7

0

1

1

5

26

7

26.9

8

0

0

1

5

24

6

25.0

9

1

0

2

3

21

6

28.6

10

1

0

1

6

27

8

29.6

Total

6

7

30

259

52

20.1

9

*Errors were catalogued into four types during this stage: (1) after "there" constructions; (2) NP-cop-adj; (3) NP-cop-NP; and (4) the incorrect use of the copula or auxiliary errors. t Total use refers to the number of times the copula was used correctly as well as incorrectly.

The English Copula by Native Speakers of Lebanese Arabic

189

Finally, it is interesting to note that omission errors found in stage 1 decreased significantly at the end of stage 2 and errors of type 4 almost disappeared at the end stage 3 as shown in Table 3 above. In this stage, only 22 errors of copula omission (types 1-3) were found and 30 errors of type 4 (incorrect use of copula). As the table shows, by the end of stage 3 learners were using the copula with a much higher degree of accuracy (70%-95%; error rates of 5% to 30%), which is strong indication of significant progress in the learners' competence. Furthermore, male learners (subjects-1-5) in particular successfully acquired the copula structure. It is interesting to note that the female learners (subjects 6—10) had a lower degree of accuracy, calculated from the error rate, throughout the three stages of data collection: Females achieved accuracy rates of 21% to 35%, in comparison with 40% to 48% for males in stage 1; 37% to 47%, in comparison with 58% to 75% in stage 2; and 70% to 75%, in comparison with 83% to 95% at the end of stage 3. This is a very interesting finding given that all learners were selected at random and the main variables, such as motivation, attitudes, language competence, and age, were accounted for. A general deficiency of language input may well explain the variation in the accuracy rate between the two sexes. As stated earlier, the female learners' exposure to English outside the classroom was limited in comparison with the male group. This finding supports Young's (1991) claim that the process of second-language acquisition is characterized by a high degree of systematicity and variation (Young, 1991). Therefore, it is essential to account for variation in the evolving system of the learner's language as he/she progresses from zero competence to nativelike competence in the target language by defining the factors (sociolinguistic, psycholinguistic, and purely linguistic) that contribute to such variation so that every aspect of the morphosyntactic variation in the learner's system is accounted for in terms of specific factors (cf., Corder, 1981; Ellis, 1994; and Young, 1991). Meisel, Clashen, and Pienemann (1981) argued, on the basis of works by Haugen (1956), Schumann (1975), and Gardner and Lambert (1972), that sociopsychological factors, such as social distance, L2 input, attitude, and motivation, play a central role in second-language acquisition, and should be incorporated into a model that specifies the exact role of such factors in the acquisition and use of a second language. They propose a multidimensional model of second-language acquisition in which the kind of interlanguage system acquired by the learner depends on the learner's sociopsychological characteristics. This model is considered to be the most comprehensive, at least in terms of its ability to account for variation in a learner's interlanguage.

190

The English Copula by Native Speakers of Lebanese Arabic

Conclusion The overall linguistic and acquisitional analysis of the data collected for this investigation support its two main objectives: (1) to establish and explain interlanguage development for the acquisition of English copula by native speakers of Arabic during the early stages of learners' interlanguage and (2) to establish to what extent errors are the result of crosslinguistic influence or are intralingual in nature (i.e., the result of a general process of language development similar to that observed in L1 acquisition). Much of the second-language acquisition research on early learners' error focused on determining the extent to which L2 acquisition was the result of cross-linguistic influences (L1 transfer) or of creative construction (the construction of unique rules similar to those that children form in the course of acquiring their mother tongue). The presence of errors that mirrored LI structures was taken as evidence of cross-linguistic influences, whereas the presence of errors similar to those observed in L1 acquisition was indicative of creative construction. (cf., Dulay and Burt, 1973, 1974a, and 1974b). The overall linguistic analysis and findings of this investigation seem to be consistent with the claim that the process of L2 acquisition is similar to that found in L1 acquisition and they further support the predictions concerning the development of learner's interlanguage—errors of a developmental type occur when the learner attempts to build up hypotheses about the target language on the basis of limited experience. Second-language acquisition research of the 1980s and 1990s strongly indicated that a substantial number of adult learners' errors are not attributable to L1 interference. For many researchers, such errors are nothing more or less than indicators of developmental processes found in both first- and second-language acquisition, and therefore such errors are often termed developmental errors (see Dulay and Burt, 1973; Meisel et al., 1981; Dulay , Burt, and Krashen, 1982; McLaughlin, 1987; LarsenFreemen and Long, 1991; and Ellis, 1994). In fact, many researchers have argued that cross-linguistic influences, or L1 transfer, accounts for between 5% and 25% of grammatical errors. The important point to be made here is the fact that many errors are clearly not the result of L1 transfer. Another important finding in this study is the reported lower accuracy rate achieved by female learners throughout the three stages of the study may be attributed to learners' restricted exposure to L2 outside the classroom. This fact has highlighted the significance of accounting for other factors, such as attitude, motivation, language input, and the learning environment, that affect the development of learner's interlanguage.

The English Copula by Native Speakers of Lebanese Arabic

191

Notes 1. The term cross-linguistic influence, which is also known as language transfer, is a theory-neutral term, allowing one to subsume under one heading such phenomena as "transfer," "avoidance," and "borrowing" (Odlin, 1989, p. 1).

References Abboud, P. F, and McCarus, E. N. (eds.). 1975. Elementary modern standard Arabic. Ann Arbor, MI: University of Michigan. Bley-Vroman, R. 1989. The logical problem of second language acquisition. In S. Gass and J. Schachter (eds.), Linguistic perspectives on second language acquisition. Cambridge: Cambridge University Press, pp. 41-68. Brown, R. 1973. A first language: the early stages. Cambridge, MA: Harvard University Press. . 1982. From topic to subject: dominance in interlanguage of Hmong speakers. Ph.D. diss., University of Hawaii. Al-Buanain, H. A. F. 1987. Second language acquisition of Arabic: the development of negation and interrogation. Qatar: Dar al-Thakafa. Corder, S. 1981. Error analysis and interlanguage. Oxford: Oxford University Press. Dulay, H., and M. Burt. 1973. Should we teach children syntax? Language learning, 23: 245-258. . 1974a. Errors and strategies in child second language acquisition. TESOL quarterly, 8:129-136. . 1974b. A new perspective on creative construction process in child second language acquisition. Language learning, 24:253-278. , Burt, M., and Krashen, S. 1982. Language two. New York: Oxford University Press. Ellis, R. 1994. The study of second language acquisition. Oxford: Oxford University Press. Ervin-Tripp, S. 1974. Is second language learner like the first? TESOL quarterly, 8:111-127. Gardner, R. C., and Lambert, W. E. 1972. Attitudes and motivation in second language acquisition. Rowley, MA: Newbury House. Hatch, E. 1978. Second language acquisition. Rowley, MA: Newbury House. Haugen, E. 1956. Bilingualism in the Americas: a bibliography and research guide (publication no. 28 of the American Dialect Society). University, AL: University of Alabama Press.

192

The English Copula by Native Speakers of Lebanese Arabic

Huang, H., and E. Hatch. 1987. A Chinese child's acquisition of English. In Hatch (1978), pp. 118-131.. Huntsbury, R. 1972. Second language acquisition in childhood. Manuscript, Language acquisition laboratory, University of Connecticut. Hussein, A. 1971. Remedial English for speakers of Arabic: a psycholinguistic approach. Ph.D. diss., University of Texas.. Itoch, H., and Hatch, E. 1978. Second language acquisition: a case study. In Hatch (ed.), 1978, pp. 76-88. Larsen-Freeman, D., and Long, M. 1991. An introduction to second language classroom research. London: Longman. McLaughlin, B. 1987. Theories of second language learning. London: Edward Arnold. Meisel, J., Clashen, H, and Pienemann, M. 1981. On determining development stages in natural second language acquisition. Studies in second language acquisition 3:109-135. Odlin, T. 1989. Language transfer. Cambridge : Cambridge University Press. Shehadi, F. 1969. Arabic "to be." In J. M. Verhaar (ed.), The verb "be" and its synonyms: philosophical and grammatical studies. Part IV: Twi, modern Chinese, Arabic (Foundations in lanugage supplement no. 9). Dordrecht: Reidel, pp. 112-125. Schumann, J. (1975). Second language acquisition: the pidginization hypothesis. Ph.D. diss., Harvard University. White, L. 1989. Universal grammar and second language acquisition. Amsterdam: John Benjamins. . 1987. Markedness and second language acquisition: the question of transfer. In Studies in second language acquisition, 9:261-286. Wode, H. 1976. Developmental sequences in naturalistic L2 acquisition. Working papers on bilingualism, 11:1-13. . 1978. The L1 vs L2 acquisition of English negation. Working papers on bilingulalism, 15:37-57. . 1981. Learning a second language 1: an integrated view of language acquisition. Tubingen: Gunter Narr. Young, R. 1991. Variation in interlanguage morphology. New York: Peter Lang.

Categories of Comprehension in Argumentative Discourse: A Cross-Linguistic Study Salwa A. Kamel This paper deals with argumentative discourse in English and Arabic. The focus is on the comprehension1 of authentic English-language argumentation by native Arabic-speaking students at the Department of English, Cairo University. What was tested is not the formal grammatical aspect of the language, what Van Dijk and Kintsch (1983) call microstructure, but the global organizational aspect of the text—its superstructure (see below, "Categories of comprehension of argumentation"). This paper, therefore, falls within the broad area of contrastive rhetoric. Comprehension as a major component in acquiring language skills demands special attention from a cross-cultural perspective. There is a critical need for research on the source of problems that native Arabic speakers face in understanding discourse in a foreign language. Graduating students of English, who are expected to use English professionally and to understand discussions, advertisements, pamphlets, editorials, articles, etc., need special attention. This paper investigated the degree of proficiency acquired by a group of such students in understanding authentic argumentative texts in English. Of particular concern in this study was whether their understanding is hampered by both cultural and rhetorical differences between the sender's encoding strategies and assumptions and the receiver's decoding strategies and expectations. Because of its reliance on global organization, argumentative discourse poses extra difficulties compared with other types of discourse, such as exposition, narration, or description. Methods of signaling the topic and arranging the evidence in both languages are assumed to be different. Organization and methods of expression are dissimilar. The present study, thus, aims first at presenting categories of logical organization and rhetorical strategies characteristic of English argumentative discourse that may be inaccessible to Egyptian students of English. It then compares these categories to native Arabic ones that may be culturally embedded in the student's cognitive system. Such a situation may require the develop-

194

Categories of Comprehension in Argumentative Discourse

ment of an awareness of other categories in English, which assist students in making the correct inference, rather than leaving them to discover them by themselves. In this respect, the problem is more cultural than linguistic, and drilling the student with more and more reading of arguments would not help, as long as s/he is not made aware of the cultural differences. Maria (1990) reports on a study by Harowitz (1985) in which a group that was given reading and writing instruction with text-structure patterns was found to do "significantly better than a group who received only reading instruction" ( p. 158). Limitations of this study. With the exception of Johnstone-Koch (1983) and Hatim (1997),2 little attention has been paid to contrastive Arabic-English work in the area of argumentative discourse. However, the sample selected for this research limits its applicability: it is meant only to document the problem that graduates of English face rather than to provide solutions or answers. According to Brown and Yule (1983), research on inference and comprehension remains difficult and somewhat questionable. Production is generally more easily testable and quantifiable than comprehension: it is a matter of what one does as opposed to what one knows. Smith (1988) takes a more extreme view: "Comprehension cannot be measured at all, despite constant educational efforts to do so, because it is not a quantity of anything...." (p. 53). As a result, more research has been done on production than on comprehension. A second limitation on this study is that when data-gathering tests are given out of context, as this one was, motivation, which according to Van Dijk and Kintsch (1983, p. 107) is essential for optimal performance on behalf of the volunteer subjects, is at a minimal level, no matter how aware they may be of the importance of the test for the purpose of the study. It is assumed that this may have affected the results in some way.

Definition of Terms Comprehension. According to Frank Smith (1988), comprehension is a tool not an outcome of reading. "Comprehension of a text is a matter of having relevant questions to ask...and of being able to find answers to at least some of these questions" (Smith, 1988, p. 166). Absence of comprehension means not being able to predict or ask questions ranging from word meaning to style and point of view. The generally accepted theory of comprehension (see Smith, 1988; Renkema, 1993; Bernhardt, 1993; and Maria, 1990, among others) places great emphasis on background knowledge as a major element. Bernhardt (1993, p. 120) quotes Grassner et al. (1987, pp. 218-219):

Categories of Comprehension in Argumentative Discourse

195

We assume that the reader constructs a structure of propositional units (called nodes) during comprehension. Some of these nodes are explicitly mentioned in the text, whereas other nodes are inferences. The comprehender needs to construct bridging inferences in order to establish conceptual connectivity between an incoming explicit statement and prior passage context. Such knowledge-driven operations are culture dependent. "Seen" texts appear in black and white, whereas "unseen" texts are intended by the writer and carry "implicit sociocultural elements" (Bernhardt 1993, p. 72) assumed to be known to the reader. Unseen elements are implicitly held in the text and are relevant to processing the text coherently. Coherence is the "connection which is brought about by something outside the text, namely knowledge assumed to be shared by speaker and hearer" (Renkema, 1993, p. 35). Unlike cohesive elements, such connections among successive sentences are not apparent in text elements (Renkema 1993). Cohesion uses explicit linguistic ties, whereas coherence is an implicit pragmatic phenomenon. Inferences, therefore, are derivable implicit information, including entailment, conventional implicature, conversational implicature, and connotation (Renkema 1993). They are used in filling in gaps or discontinuities in interpretation. Brown and Yule (1983, pp. 223-224) comment that We certainly rely on the syntactic structure and lexical items used in a linguistic message to arrive at an interpretation, but it is a mistake to think that we operate only with this literal input to our understanding.... We also rely on some principle that, although there may be no formal linguistic links connecting contiguous linguistic strings, the fact of their contiguity leads us to interpret them as connected. We readily fill in any connections that are required. This is inference: getting from the literal meaning to the intended meaning, i.e., implicature. It is largely based on background knowledge. Prior knowledge is used to process new information. The theory of knowledge places more emphasis on items of prior knowledge such as schemata, scenarios, scripts, field, and topic familiarity, etc., than on lexical or grammatical categories in the process of comprehension. The lack of such prior knowledge creates a good deal of problems for second/foreign language learners. De Beaugrande and Dressier (1983, p. 36) have found that "text utilization fails if unresolvable discrepancies persist and regulative integration of an occurrence into one's store of knowledge cannot be

196

Categories of Comprehension in Argumentative Discourse

done." Bernhardt (1993, p. 94) classifies types of knowledge as highly idiosyncratic, or local, knowledge shared by two or more people; domainspecific knowledge, such as history, music and so on; and culture-specific knowledge, such as rituals and aesthetic values. Smith (1988, p. 223) holds that each one of us has his/her own theory of the world that helps us in interpreting our interactions. "The basis of comprehension, whether of language or of the world in general, must be some internal organization of knowledge (or belief) about the world." Anything that cannot be related to the theory is bewildering. To share a culture, therefore, is to share similar categories for organizing experience and hence deriving meaning from language. All these elements facilitate comprehension. Bernhardt (1993) and Van Dijk and Kintsch (1983) add one more element, namely text structure. Knowledge of text structure increases comprehension and helps in overcoming difficulties resulting from the lack of cohesive devices, such as conjunctions, or from syntactic obscurity or implicitness. Such knowledge enables the reader to make predictions, thus making comprehension possible. Meyer (1985) found that knowledge of text organization and superstructures (see "Categories of comprehension of argumentation," infra) is systematically employed by skilled readers as a strategy to facilitate comprehension. Making sense of any text is a function of introducing some degree of coherence into it. Ricento (1987) sees that coherence is not a permanent and inherent property of a text; for in the absence of a familiar cultural background, a reader is unable to provide answers or even ask questions, and therefore fails to understand. A reader may use the author's cultural background or his own. The latter, according to Van Dijk (1983), dominates the former.3 Creating coherence and making inferences, therefore, are functions of background knowledge and the knowledge of text organization, both being highly cultural issues. Argumentative text: type. Hatim's (1997) model rests on three basic sociotextual units: the text (i.e. the rhetorical purpose, such as argument), the discourse (or the writer's attitude or orientation, such as a committed Christian), and the genre (or the conventional type, e.g., editorials). He defines "texttypes" as "global frameworks utilized in the processing of rhetorical purposes in discourse" (Hatim 1997, pp. 54-55). Text-type focus, which determines text structure and "regulates the way texts are organized as cohesive and coherent wholes" (p. 55), is the "predominant rhetorical purpose served by a given text" (p. 36). Texts are multifunctional, i.e. hybrid, in that one function is predominant while the others are subsidiary (p. 41). Exposition and argumentation share techniques, such as comparing-contrasting, narrating, cause-effect, generalization, classification, definition, and analogy (in an experiment described in Connor, 1986, the argumentative text

Categories of Comprehension in Argumentative Discourse

197

that used narration for support was rated highest by all judges). But rhetoricians distinguish them on the grounds that exposition seeks to inform through the use of such techniques, whereas argumentation uses the same methods to change the way a listener/reader thinks or acts (Kane 1983, p. 457). If a writer weighs the pros and cons, this is comparison; it turns into an argument when s/he takes a definite stand. De Beaugrande and Dressier (1983,162) see that exposition "monitors" the situation, whereas argument "manages" it, i.e., focuses on the formation of future behavior. In other words, exposition requires explanation in order to make things clearer and understandable, but argument requires proof to justify a choice between two or more already well-understood issues. Exposition is nonevaluative: its purpose is to analyze concepts, to narrate or to describe, whereas argument "focuses on the evaluation of relations between concepts"—it starts off with an evaluative thesis (Hatim, 1997, p. 38) and continues from there. The global processing pattern is the plan of how events and states lead up to the attainment of a goal. Levin (1990) requires both types of texts to be clear, but an argument must be also convincing. Argumentation is more advanced than other text types, such as description, narration, and exposition, in that it can use these three types as evidence or support. According to Fareh (1988) narration is interesting because it is chronological and person-oriented, while exposition is impersonal and is oriented to subject matter, with no time focus, and with regressive directionality into an already constructed universe that is "uniform and familiar" as Hatim (1997, p. 86) puts it. But argumentative texts are difficult, challenging and unpredictable because of their zig-zag progressive diretionality into a world to be constructed. As such more weight falls on the reader. Baker (1896, p. 33) adds another difference: exposition may have a "term" as the subject of discussion, whereas argumentation is always built around a "proposition," which is "an assertion in regard to a term or terms." He reserves the term "argument" for discourse that seeks to convince by means of intellectual and logical reasoning: "it is the art of producing in the mind of someone else a belief in the ideas which the speaker or writer wishes the hearer or reader to accept" (Baker, 1896, p. 33). "Persuasion" is a term referring to a rhetorical act which makes the hearer or reader act as the speaker or writer's desires through exciting the emotions; it has been used mostly in public speaking since ancient times. But in written arguments, it is the structure, not the eloquence, that matters. According to Kane (1983), argument establishes equality between readers and writer, whereas persuasion "manipulates," and as such is relegated to a lower moral order. Such techniques are today exploited in advertising and politics, misleading people by fears, sympathies, material desires, or prejudices, rather than by reason. The very act of writing involves "stepping back from emotion, and even genuine feeling, when powerfully expressed in prose, must be focused by a

198

Categories of Comprehension in Argumentative Discourse

guiding intelligence" (Kane 1983, p. 469). Thus, reasoning and logic are highly regarded and structure is considered essential to argument. The audience in argumentation (in a purely Aristotelian tradition) demands attention. Kane (1983, p. 467) sees "the effect upon the reader" as the principal purpose, but effectiveness is not a test of value because what may convince an incompetent gathering of hearers may fail to convince a highly qualified audience: truth and rhetorical effectiveness are different things. Perelman (1969) holds that dialectic since ancient times concerns opinion, thus argumentation is not connected with truth as such, but with "adherence." Whately (1969, p. 5) similarly sees that reasoning in argumentation is applicable to two purposes: "inferring" is the ascertainment of the truth by investigation," and "proving" is the "establishment of [the truth] to the satisfaction of another..." Thus, Aristotle judged argumentation amoral (Cooper, 1932). Similarly, Perelman (1969), from a philosophical rather than a rhetorical point of view, also gives the audience priority on the scene: "it is in terms of an audience that an argument develops.... A change in audience means a change in the appearance of the argumentation" (Perelman 1969, pp. 5-7). An argument, therefore, is expected to bear the cultural and social mark of the audience it addresses. Supporting evidence in argumentation is relevant to the type of audience and their biases and concerns which may affect their receptiveness. Support is basically a means of winning over the audience by preventing them from suspecting deception. (See "The cultural nature of logic," below, for differences between Arabic and English argumentation.) Argumentative text: form. Since organization creates strength in argumentation, the structure or schema of an argument is an integral part of it. The logic by which arguments develop is of two types: either it uses specific facts (experience and observation) to support a generalized conclusion, known as induction, or it works from a generalized premise toward a specific conclusion, known as deduction. The two types can be mixed, as when the premises on which a deductive argument rests need supportive evidence. McDonald (1983, p. 51) also suggests that the major premise of a deductive syllogism is usually derived by induction. Evidence comes from experience or opinion, factual information, reliable statistics, diagrams and charts, and authority and expert testimony. Baker (1896) enumerates the components of argument: analysis (which sorts out main and subordinate ideas and ensures their truthfulness), finding and selecting evidence, and rhetorical structure. McDonald (1983) describes this structure in terms of a syllogistic schema in which the first paragraph usually contains the major premise and the last paragraph contains the conclusion. First and last positions are the strongest.

Categories of Comprehension in Argumentative Discourse

199

The "body" presents the minor premise. Premises are "propositions [facts or truths] accepted by the hearers" (Perelman, 1969, p. 104). Facts are universal agreements on things that require no justification for adherence: they are presupposed. If the premises are presumptions, they need reinforcement. Stevens (1983) suggests that statements about beliefs or judgments beyond personal preference always need supporting before they are accepted. The syllogism is valid only if the form is correct, and it is sound only if the premises are true. In a deduction, the logical rules are inference rules; they are supposed to support the conclusion conclusively because if premises are accepted [i.e., no rules that may lead to fallacies are broken], it is not possible to reject the conclusion. By contrast, in induction, evidence rules are used: these always lead to probability not proof because no comprehensive evidence can ever be presented. They support the conclusion to a lesser degree. Refutation and concession are part of this structure.4 Refutation is attacking or disproving another argument; it is often the basic purpose of arguing (Kane, 1983). It is done by "showing that the 'assumptions' upon which the claim is based are 'untrue'"(Maccoun, 1989, pp.101-102). It is needed even when one is advancing a positive conclusion, when one is pointing out weaknesses in logic or evidence. Concession is temporarily admitting premises or evidence contrary to what one hopes to establish because it cannot be refuted in a bid to neutralize them or to turn them to positive account by showing that they are not important in light of the counterargument. Categories of comprehension of argumentation. The categories inferred for testing in this study are adopted from Van Dijk and Kintsch (1983) and Maccoun (1983). Van Dijk and Kintsch sketch the way knowledge is used in discourse comprehension. This model is suitable for the purpose of the present study because it is more concerned with strategic processes on the higher-order discourse comprehension level than with lower-order perceptual processes or linguistic parsing processes. In this model, verbal input is decomposed into atomic constituent propositions that are organized into larger units on the basis of some knowledge of structure to form a coherent text. From such a base a "macrostructure" is constructed, which represents the most essential information in the text base. The comprehender's knowledge, beliefs, and goals play a role in the process of interpreting new information and integrating it into the already existent system of world knowledge that s/he has. Maturity of knowledge determines the quality of thinking, and beliefs begin to change, hence the importance of memory. Thus, interpretation is based on previous knowledge derived from experience. "Local coherence strategies are mixed with global coherence strategies" controlling each other (Van Dijk and Kintsch, 1983, p. 152).

200

Categories of Comprehension in Argumentative Discourse

In reading a text, a hierarchically structured macroproposition keeps order among semantic details. It can be kept in short-term memory and is an important cue for the actualization and processing of (complex) knowledge. Macropropositions are inferred through a bottom-up process. Their best position is the beginning or end of units. They need not be directly expressed, but can be inferred from semantic interpretation. Once acquired, they exercise top-down control: "...the relevant macroproposition for an episode remains highly available during processing; this information is constantly available for strategic comprehension" (Van Dijk and Kintsch, 1983, p.180). This is known as "macrocontrol," or top-down control, as distinguished from linear or sequential control. Hatim (1997) adds that in top-down macroprocessing, the text user is concerned with the relation of context to text and to the text-type focus, while in bottom-up microprocessing, users focus on the various elements of texture, such as the exponents of cohesion and coherence (see also Brown and Yule, 1983, p. 234). Macrostrategies are used to infer macropropositions from the sequence of local propositions. Levels of macropropositions form the macrostructure of a text: the topic of the text. A reader need not wait till the end to infer what the text is globally about. These strategies have a heuristic character. A reader is led by thematic words or first sentences and other information from the text. Superstructure is another category that Van Dijk and Kintsch propose in their theory of comprehension. It is the schema5 that organizes the macropropositions. (In the case of this study, it is argumentation.) From the first cue in the text, the reader activates a superstructure, a schema that he uses as a powerful top-down processing device. By allowing a reader to anticipate information, such a schema facilitates comprehension (see supporting experiments in Van Dijk and Kintsch, 1983, pp. 254-257). Meaningful organization also assists retention. Schematic organization aids comprehension, and comprehension is partly finding an appropriate organization for a text. Van Dijk and Kintsch require that such schematic superstructures should be not only in the text but also in the mind of the reader. Other stylistic strategies exist, which are options controlled by the situation to monitor the degree of (in)formality to infer information about the speaker and his class membership, as well as his tone of voice.6 These are rhetorical strategies that, if effective, realize the goal of comprehension and the success of the speech act (e.g., acceptance of the author's position). The conventional nature of language, therefore, facilitates understanding. "The more we can anticipate and employ the formal structure that an author uses, the more we can understand and remember what we read" (Smith, 1988, p. 41). Even at the level of perception, units are easier to perceive when they are part of a meaningful larger unit, i.e., macrounit.

Categories of Comprehension in Argumentative Discourse

201

Context, thus, facilitates perception and controls hypothesis testing. A good reader is less concerned with bottom-up processing. Meyer (1985) found that skilled readers approach texts with organizational knowledge and apply strategies that use the highest level structure; absence of signaling in well-organized texts did not have much effect on good comprehension. According to Van Dijk and Kintsch (1983, pp. 251-252), superstructures are macrosyntactic rhetorical conventions beyond grammar, such as syllogisms, for example. These are learned by the language user as organizational patterns to aid understanding, much like principles, rules, strategies, categories, or other units of language and culture. Such schemata vary according to culture, and part of comprehension is discovering these schemata. Superstructures, therefore, are culture specific. Maccoun (1983) relates this model to argumentation and finds that in order to understand argumentative discourse on a macrolevel the reader must be able to recover the author's position or logic, his evidence, the opposition's argument, and the author's refutation or skepticism about the opposition's argument. Added to this is the ability to infer the type of argument used, the position of intensity (i.e., whether the best argument is placed first or last) and the tone (see note 6). In order to attain full comprehension, one should successfully infer relations of cohesion and coherence, both local and global (see below, "Coherence"), markers and their meanings, including metadiscourse markers (see Grismore, 1983, pp. 43-47), topic shift and its role in sequential relations7 infrequently marked by adverbial expressions of temporal sequence (Longacre, 1979), and connectives, such as "however," "furthermore," and "so"(Van Dijk 1977), as well as semantic gaps and implications.

Coherence. Coherence is thematic continuity. It is defined by Kaplan (1972, p. 120) as "an orderly flow of sentences marked by repetition of key ideas." Van Dijk and Kintsch (1983, p. 43) assumed that coherence exists when two propositions share a common argument or a common referent and are related, which facilitates comprehension. The relationship is not always explicit. We often need bridging inferences to construct a coherent memorizable relationship. What is missing is inferred by the reader, and thus the focus is shifted from the linguistic to the conceptual level. Bridging inferences are based on background knowledge and they usually take time to become acquainted with. Brown and Yule (1983, p. 257) call them "missing links." Each link is activated by certain schemata or scenarios. Such inferential processes produce macrostructures, i.e., help reduce a text to its communicative message, from the detailed text base, the lowest level of macrostructure (Van Dijk and Kintsch, 1983, p. 52). The text is thus reduced to its essential components in successive steps, resulting in a

202

Categories of Comprehension in Argumentative Discourse

hierarchical macrostructure. This is strictly related to the summary section of the test designed for this study. Ricento (1987, p. 43) sums up the above argument in the three levels of text identified by Meyer (1985): the sentence, or microproposition, level concerned with how ideas are organized within and across sentences; the macropropositional level, which "is concerned with the issues of logical organization and argumentation"; and the highest level, which is the top-structure, "or overall organization of the entire text." This last level covers the rhetorical relationships across complexes of paragraphs. Coherence is a property of discourse that is underpinned by specific grammatical manifestations known as cohesion. Local coherence also needs macrocontrol in the form of discourse topic, which is a discourse function of sentences. According to Van Dijk and Kintsch (1983), both local coherence strategies, which are accounted for in terms of relations between propositions in subsequent sentences, and global coherence strategies, which are characteristic of discourse as a whole including topic, theme, upshot, etc., are needed, as well as propositions derived from text understanding and those inferred from world knowledge. Pragmatic information is also relevant; according to Van Dijk and Kintsch (1983, p. 239): Some text types are not merely defined in terms of surface structure or semantic content and their schemata, but above all in pragmatic terms. Argumentative discourse is a case in point. Premises and conclusions are schematic categories which not only are linked through a semantic chain of implication, entailment, implicature, or practical inference, but also involve speech acts of asserting, assuming, and drawing conclusions.

Knowing the parameters of cultural and social conventions results in successful conversational interactions; similarly, "the cultural aspects of these very general and basic strategies of comprehension reside in the fact that the context types, text types, and schemata may vary according to culture" (Van Dijk and Kintsch, 1983, p. 238).

Differences as a Source of Difficulty A comparative study of rhetorical conventions in the area of argumentative text types reveals fewer similarities than might be expected. Concise style. Concreteness, which appeals to the senses, is a mark of good writing in English, but do not tell everything; it is boring. Good writing is economical. This is a warning offered by Levin (1990, p. 447) that

Categories of Comprehension in Argumentative Discourse

203

reveals a very English-specific characteristic, by no means a universal requirement of style. Arabic, for example, is a language that favors both semantic and syntactic repetition as a mark of good style (see Johnstone 1991). English style, therefore, abounds with gappings and ellipses, and a good deal is thus left unsaid. Hatim (1997, p. 2) describes the intuitions and expectations of a native English reader concerning "effectiveness": if something is dramatized, it is "unsubtle" and inappropriate; it is also not persuasive if it is direct and means exactly what it says. Although in both Arabic and English, argumentation aims at convincing the reader, the onus in English falls on the reader, while in Arabic it falls on the writer. In English, a valid argument cannot be rejected unless a counterargument has been strongly presented. In Arabic, an argument is anchored into the reader/listener's awareness through repetition, another instance of what Kaplan (1983, p. 150) calls a "culture-bound preference." Hatim (1997, pp. 172-173), therefore warned that the Western reader ignores the seemingly superficial features of the Arabic text [e.g. repetition, parallelism, paraphrase, solidarity, and warmth, etc.] at his peril. They are all there for a purpose and while the means of expressing the attitudinal meanings involved may differ from one language to another, the ends are universal values which are globally recognizable.

In other words, the Western reader may be too involved in his/her own expectations derived from his/her own culture to be able to judge objectively the effectiveness of other rhetorical styles which s/he may find awkward8 (see Hatim, 1997, ch. 14, for a detailed counterargument against accusations directed against the popular qualities of Arabic style). Surface formats and logical markers. Hatim (1997, p. 65) noted that Arabic marks "surface formats more explicitly than, say, English"; by contrast, Johnson (1983) asserted that explicit logical markers are not always present in English texts and most of the time one has to rely on context for interpreting a text's communicative value. These differences between English and Arabic make the comprehension especially of argumentative texts a demanding task for L2 readers. El-Shershabi (1988) described Arabic style as more explicit and cohesive, whereas English is more implicit and coherent. El-Shershabi notices that the rather imprecise usage of pronominal reference in Arabic reduces the degree of textual coherence for foreign readers,9 although a shift in the editorial discourse of Arabic "towards the distributional cohesive patterns of the Anglo-American discourse" (El-Shershabi, 1988, p. 255) is taking place.

204

Categories of Comprehension in Argumentative Discourse

Topic. Other types of difficulties are related to topic. For example, the topic sentence may not be explicit and the reader has to formulate it him/herself. Another case involves topic shift. Hatim (1997, p. 59) remarked that English suppresses signals of topic shift within and between paragraphs, and the shift is recognized intuitively by being aware of the writer's intentions. When rebuttal requires that claim and counterclaim be juxtaposed, unless the reader is aware of the superstructure of argumentation in English, one paragraph could be mistaken for two by Arabic-speaking students. Ellipsed ideas and coherence. Pragmatic gaps in English are also a source of problems in comprehension. Johnstone (1991) remarked that statements such as /huwwa rufayya9 mish tixi:n/ ("he is thin not fat") are pleonastic in English. Presupposition in English is part of the meaning, whereas Arabic tends to assert. The statement in English violates the default presupposition. Such "reverse paraphrase," saying what something is and then what it is not is typical of Arabic. Thus, it is the area of coherence that creates comprehension problems in English. Coherence and unity as such may be English concepts only; there is no such strict requirement in Arabic, Chinese, or even German, according to Kaplan (1986). Each language has its preferences and valued choices. If we accept Miller and Kintsch's (1981, p. 335) view that "readability can be viewed as an interaction between a text and the reader's prose-processing capabilities, rather than as some innate property of a text," such capabilities partly spring in culture. Halliday (1976, pp. 22-23) also asserted that register, and argumentation is one type, and is understood by a receiver in the context of a situation—in a culture. Cultural background and relevance. As Hatim (1997, p. 3) put it, reading is, in a sense, feeling one's way through a text. It is an epistemological and heuristic attempt at finding relevance, therefore background knowledge is used to fit the new information in and to help in making inferences. It is a cultural issue, of which the superstructures of texts is a part (Van Dijk and Kintsch 1983). If the writer and reader come from different backgrounds, difficulties result. Bernhardt (1993) holds that each cultural context will bring a different set of values into play; as such, texts are cultural artifacts. Presuppositions are about what everyone knows, but not a nonnative speaker (Kaplan, 1983). Presupposed information may be absent from the second-language reader's knowledge, his/her sociocultural skills may be deficient, and his comprehension will be based on linguistic data that are not adequate to recapture the implicit information carried by the text. The reader in this case uses inappropriate cognitive strategies to build an inappropriate model of the text, without being aware of the prob-

Categories of Comprehension in Argumentative Discourse

205

lem; as a result, inappropriate inferences are made (Johnson, 1983, p. 31). If speaker and hearer do not share knowledge about categories or if the speaker presupposes sociocultural events, actions, or situations that are not shared by the hearer, there is a break in comprehension. Culture is a shared system of knowledge and beliefs, a kind of a conceptual code, by which people order their perceptions and experience and make decisions, and by which they act, interpret themselves and the world, and formulate behavior (Vivelo, 1978). Superstructure. The difficulty in understanding argumentative discourse in English can result from the macrolevel alone. Sentences at the microlevel or even paragraphs can be understood separately, but this does not guarantee an understanding of the overall idea, i.e., the logical development of the text. Maccoun's (1983) experiment shows that macroissues are more problematic than local or micropoints. The overall organizational pattern resulting in coherence may escape the reader due to loss in cues. Superstructures facilitate the accessibility of cues and, when they are inaccessible, comprehension fails. Recall and summarization are good judges of such overall comprehension. In argumentative discourse in particular, where the meaning resides in the relationship of parts to the whole, i.e., the development of the argument from claim to refutation to evidence and then to conclusion, the knowledge of superstructure guides and controls comprehension at the same time. The absence of macrocontrol may result in making the wrong predictions or in preventing the student from locating the writer's bias (see "Categories of comprehension of argumentation" above and "Results and discussion," below). A student who is preoccupied with the local analysis of sentence-by-sentence (bottom-up) comprehension, unaware of top-down control in argumentative discourse, is bound to be confounded by the uncontrolled multiplicity of incoming information. Macrostrategies of comprehension, therefore, require much training to complement the inferencing of local properties by local coherence strategies. A common developmental error that Maccoun (1983) discovered in her case studies of nonnative comprehension of argumentative texts in English is to mistake the opposition's argument for the author's. The student gets an impression from reading the first paragraph, for example, and never changes it, even when the paragraph questions and does not assert (Maccoun, 1983, p. 44). This finding is highly corroborated by the present study: many students either totally lost sight of the opposition or mistook it for the writer's viewpoint (see "Results and discussion").10 Hatim also agreed that counterargument "for many Arab users of English...seems to be a blind spot" (1997, p. 50). They are unable to appreciate "the subtle rhetorical functions involved in counterargument in English" (p. 158). The concept of refutation in argu-

206

Categories of Comprehension in Argumentative Discourse

mentation seems to be a very difficult point to hammer home to Arabicspeaking students. It is a cultural issue and requires very special attention. That is why Kaplan (1983) insisted on confronting difficulties for nonEnglish speakers, not by more linguistic training, but by teaching a language in its cultural context, i.e., its communicative functions. How much is left out of a text and controlled by a wider cultural context is thus a serious problem (Kaplan, 1972, p. 121). Following the analysis of George Steiner, Kaplan (1983) called these types of difficulties "ontological." They are epistemological in nature and cannot be resolved by linguistic means. They occur when "the contact of intelligibility" between sender and receiver is broken (Kaplan, 1983, p. 141). Kaplan's example is that of discourse "focus" as opposed to "topic." It is a culturally bound syntactic strategy that marks prominence of information in a text. Fronting in Arabic, and clefting, passive voice, and subordination in English are examples. If the reader is not aware of these hierarchies of levels of information, s/he can never get to the gist of an argument. The cultural nature of logic. As stipulated by Raphael and Newman (1983, p. 314), the highest form of argument appeals to reason. Reason is what distinguishes argumentation from persuasion according to Baker (1896). Reason appeals to Aristotelian logic, which has always permeated Western argumentation. Kaplan (1966, p. 1) warned that "logic,...which is the basis of rhetoric, is evolved out of a culture, it is not universal." People interpret the world differently and so they express it differently. Kaplan saw the interpretation of the world rather in the reader. Mastering the logic of a language is part of learning it (see al-Khawli, 1931). English develops deductively or inductively, and the maturity of style depends on the degree of subordination. Arabic, on the other hand, works by coordination (Kaplan 1966, p. 8). Arab rhetoricians may have been aware of Western logic since the third century CE. They were among the earliest translators of the classics, particularly Aristotle. But although they were aware of Western logic and widely studied and practiced counterargumentation for a period of tune, they did not finally adopt the Western model. According to Hatim (1997, p. 160), "the sociocultural milieu of Arab Islamic society was quite different in negotiating texts." Al-Sharkawi (1986) confirms the fact that Islamic theology had been greatly influenced by the Greek method of reasoning; however, this trend had been greatly opposed by the Islamic mainstream of thought. The conviction was that any approach to reason or argumentation must not outgrow the limits for human thought laid by the Creator.11 Al-Khawli (1931) confirms the roots of Arab/Islamic philosophy, theology, and rhetoric in Western thought from the time of the translations of Greek texts in the third century. Poetry and logic were fused

Categories of Comprehension in Argumentative Discourse

207

together. But the development of Arabic rhetoric has been limited to lexical and sentential boundaries. The concept of paragraph as a unit of thought and logic and the autonomy of text structure does not exist in Arabic. What Johnstone-Koch (1983, p. 49) called presentation is a typically Arabic mode that does not depend on rigid organization but on stating and restating a claim, i.e., when "emphasis serves to uphold a conviction." It is convincing by repeating, "something English writers are taught to avoid." This is what Hatim (1997, p. 8) called through argument. Maccoun (1983) found that L2 learners scored better with a text that contained one point of view rather than two or three, although the vocabulary and the syntax of the single-viewpoint text were more difficult. In a patriarchal society, social order is imposed by authority that is always assumed to know more and, therefore, is fit to direct and to be used as a "persuasive" tool to change things. Compliance by group or individual in that case is a social virtue. In the Islamic tradition, the Quran is the ultimate authority, and strong adherence is a tribal trait (Daif, [1963], pp. 18-19). Thus, an argumentative method is based on authority: the speaker is in no position to be refuted. Repetition drives the message home safely, hence the "presentational" method discussed by Johnstone (1987; 1991) in connection with Arabic, whose goal is "not to convince but to instill in the reader a sense of identification with its point of view" (1991, p. 105). In that case, fact and opinion, which must be differentiated in an English argument in order not to distort the message, often converge in an Arabic argument, while coordination is preferred to subordination. Johnstone-Koch presents a sociolinguistic explanation of argumentative rhetoric in Arabic. However, a historical one is also due. Argumentation in a large part of Arabic literature was directed toward the defense of i'jaz ("inimitability") of the Holy Quran, where denial or opposition had no place. Explaining i 'jaz became the ultimate aim of rhetoric. A good number of sciences were derived from the Quran. For one school, rhetoric was a philosophical issue aimed at establishing proof for the Muslim people. A second school, headed by Al-Sakkaki and his followers, opposed the possibility of explanation or proof. In this context, the initial statement or any claim at all had to be accepted as sincere. Johnstone-Koch (1983, pp. 53-54) remarked that an argument about a universal truth is presentation, whereas an argument about a doubtful truth is proof. Since Arabic arguments were mostly religious, they were of the first type. The audience was not skeptical. Counterargument is purposeful only in the case of a skeptical audience, where the initial statement cannot be taken at face value. The lip-service paid in concessions, for example, would appear to an Arabic-speaking reader as "devious" (Hatim 1997, p.170). The "truth" should be made available through direct means (see "Results and discussion" for findings concerning the present sample in this connection).

208

Categories of Comprehension in Argumentative Discourse

Even a quick look at commercials on the Egyptian television confirms this situation. The typical commercial attempts to persuade by making the commodity actually present and not by building up a logically convincing situation. Whether this is symptomatic of the attitude toward persuasion in the culture is not the issue here: the issue is whether frequent exposure to such methods affects the expectations of the public, including students of English. Hatim (1997) argues that such a mode is not necessarily less persuasive than a logical structure. It is just another mode addressed to a different public with different expectations. However, the question that is of interest to the present research is not whether students of English in this part of the world are ready to adopt an English argumentative method in their writings or in advertising—the question is whether they can understand an English argument when they read one. Hatim (1997), however, raises an issue in connection with Johnstone's analysis of the "presentation" method in Arabic argumentation. He identifies two types of argument: through-argument and counterargument. The first is preferred in Arabic, whereas the second is preferred in English. Through-argument states a viewpoint to be extensively substantiated or argued through no reference to an adversary (thesis-substantiation-conclusion). The counterargument selects a summary of someone else's viewpoint and offers a rebuttal or an explicit concession. Such a format is culture specific and is not available to all languages. The choice between one type and the other is motivated by matters such as politeness, ideology, power, social life, and political system, or family role, and sociopolitical norms, such as "saving face," attitudes toward truth, freedom of speech, and so on (Hatim, 1997, p. 47). "It is therefore the speakers and not the languages which must be held countable" (Hatim, 1997, p. 53). Although in modern standard Arabic, Western-educated writers use counterargument, the general tendency is traditional: the writer "advocates or condemns a given stance, glossing over beliefs entertained by an adversary" (Hatim, 1997, p. 47). As such, through-arguments are masked expositions (Hatim, 1997, p. 154). Since most rhetoricians agree that in argumentation the focus is on the audience, Hatim (1997, p. 171) proposes the following rhetorical maxim to govern the relation between arguer and audience in Arabic: On a given occasion, assume that the world is divided into those who vehemently oppose your views and those who wholeheartedly endorse them, but when it comes to whom your contribution is designed to address, talk only to your supporters and ignore the opposition.

In short, there is no recognition for the opposition. Such is the

Categories of Comprehension in Argumentative Discourse

209

macrorhetorical organization of an Arabic argumentative text: support is taken for granted. If we take into consideration Grierson's (1945) definition of the Aristotelian view of rhetoric (quoted in Johnson, 1983, p. 249), which is the study of how to express oneself correctly and effectively, bearing in mind the nature of the language we use, the subject we are speaking or writing about, the kind of audience we have in view, and the purpose, the last of which is the main determinant, there should be no absolute value attached to either approach to argumentation so long as each is suited to its purpose and its audience. Perelman (1969, pp. 116-117) describes a theory of presence that may be a psychological phenomenon but also an essential factor in the rationalistic conceptions of reasoning. The thing that is present to the consciousness assumes an importance that the theory and practice of argumentation must take into consideration. Accordingly, one of the preoccupations of a speaker is to make present, by verbal magic alone, what is actually absent but which he considers important to his argument. Subordination vs. coordination of ideas. The structure of a paragraph in English is based on subordination: a number of ideas are subordinated to one central idea, which is the topic statement. This is essentially a hierarchical structure, in which some elements enjoy a higher communicative value than others. A paragraph, therefore, is marked by a topic shift, which is often signaled by topic markers (Brown and Yule, 1983). The same structure is repeated in the larger text, creating a similar relation between the larger chunks, i.e., the paragraphs. Such a nuclear structure, which highlights the topic, is not present in Arabic, where parallelism and repetition are established as a strong tendency (see the example in note 10). Such a tendency highlights the topic not by factorizing it, but by underlining it, hence the frequent use of the term tawkidu al-ma'na ("emphasizing the meaning") in describing the function of such tools in Arabic rhetoric. These tools are aided by a large system of connectives and an elaborate pronominal reference. (See "Score results" for an example of how this can result in miscomprehending an English text.) Thinking hierarchically renders the gestalt clearer through creating prominence. The "body" of information is easier to grasp. Thinking linearly through coordination, creates an informationally accumulative effect without establishing relations among the parts, and the gestalt can be missed. Students trained to read in Arabic, therefore, are confronted with problems because they are not trained in looking for the central idea, which is used in macrocontrol to facilitate comprehension (see "Categories of comprehension of argumentation").

210

Categories of Comprehension in Argumentative Discourse

Study Design This study is based on a test administered to 61 fourth-year students of English language and literature at the department of English, Cairo University, who were expected to graduate within five months of the test. It consisted of a passage of argumentative discourse and a number of questions designed to test the students' aptitude in comprehending the passage. Three judges were asked to provide answers to all the questions, which guided the researcher in assessing the students' answers. The three judges were teachers of writing, two American and one British, all native speakers of English. They were also asked to provide opinion as to the suitability of the test in terms of the study's goal and the degree of difficulty and clarity of the text and questions.

Selection of text used in study test The passage (Appendix 1) for this test was selected because it met the following requirements; it was • an authentic text written for native speakers of English in a native environment and for a practical purpose • of medium difficulty. For this purpose, the judges' opinions were decisive. In fact, the judges had no difficulties with the text, with the exception of sentence 28, which seemed obscure to the British native speaker. His comment was "sentence 28 is a bit obscure," because, as he commented, "we have to know who Adam Smith is and what he theorizes." Three top-ranking students were given an extra question that asked them to rate the test as a final examination in reading comprehension in terms of difficulty as "fair, difficult, very difficult, or unfair." Two of them considered the test and the passage to be "difficult" and one "fair" • of reasonable length with familiar vocabulary, touching on some foreign cultural idiosyncrasies to test the students' inference strategies • about a serious subject of universal interest to "involve" the students learning something new in the process of reading, not just to answer questions. For this purpose the budget expenditure topic seemed suitable in that it did not contain inaccessible knowledge which would obstruct comprehension, but has enough cultural information to test the importance of background knowledge in understanding a text. However, in answering question 1 (whether the reader enjoyed the passage and why), 47.5% found it enjoyable for reasons ranging from discussing a real, important, and up-to-date topic to the fact that it was challenging because it posed indirect, difficult, and new problems that were hard to four subjects. The passage was found difficult and/or boring by 52.5%12 for reasons such as "it is puzzling," "not straightforward," "complex and unfamiliar topic,"

Categories of Comprehension in Argumentative Discourse

211

and "too technical" (with one subject response for each), and "not interested in the topic" and "incoherent and no unity" (both of which were represented by two subject responses each) • a selection from a book on teaching reading skills to L1 English learners to ensure the passage's representative value as a comprehension text • characterized by a style of English more typical than that used in teaching passages in general, i.e., it was a text that tended to be more implicit than explicit, with few logical connectives—it was thus selected to encourage inference strategies and deducing of presuppositions for appropriate interpretation of the text. The text was then analyzed by the researcher into its units and their functions (see Appendix 2) in keeping with the adopted model (see "Categories of comprehension of argumentation," p. 199).

Test: method and practicalities The study test (Appendix 1) was designed as a proficiency test for a specific purpose: to determine how students apply the reading skills they have developed through three and a half years of instruction. Students in the English department of Cairo University receive intensive instruction in reading different types of texts, including argumentative texts, during the first year of their studies, and continue to attend combined courses in reading and writing throughout the three successive years, including the first term of the fourth year. A proficiency test to evaluate students' language abilities, according to Harrison (1986, p. 8), should ask, "having learnt this much, what can the student do with it?" Each item of the ideal proficiency test should be designed to elicit an answer that will demonstrate the student's competence in relation to the three major areas of textual organization: the micropropositional level, the macropropositional structure, and the superstructure. In planning the test, the explicit naming of categories was avoided in order not to lead students to interpretations. Questions asked while the text passage was available for review by the students (part I) were of two general types: open-ended (OEQs; section A) and multiple-choice questions (MCQs; section B). The test included attitudinal/opinion questions (nos. 1, 2, 3, 11, and 12) that were considered for discussion but received no score. The students were given one and a half hours to read and answer the questions, after which all papers including the text were collected; they were then given another half hour to write the combined recall-summary exercise and answer the biographical questionnaire. The questionnaire is meant to divide the students into two groups of Arabic- and English-medium background of at least secondary school education in a bid to compare the impact of a prolonged period of exposure to English on comprehension, and also to disqualify from the study students who may have had

212

Categories of Comprehension in Argumentative Discourse

more exposure through foreign parents or by living for a long time in an English-speaking country. The test starts with questions of a global nature, before moving on to details. It assumed that the student understands the passage if s/he recognizes the following components: • the nature/function of the passage as a whole as an argumentative text type (OEQ 8; summary) • elements of argumentative text type: major claim, opposition, evidence, the writer's bias, and their loci (OEQs 5, 7, 9, 15, 16, 17; MCQs 1,4; summary) • presence of discourse markers, such as hedges, related to the writer's tone and attitude toward the subject and audience (OEQs 3, 4, 6) • elements of local and global coherence, both implicitly and explicitly marked, and cohesion (OEQs 2, 3, 7, 10, 11, 12, 13; MCQs 7, 8 ). In addition, an understanding of the passage is assumed if the student demonstrates: • the ability to use inference strategies (OEQs 7, 13, 14, 18; MCQs 3, 5, 6), and • the ability to tell fact from opinion (MCQ 9).

Rationale for question types used It is commonplace knowledge that not all types of questions are relevant for every purpose. For example, to find out about global organization in connected discourse cloze tests are irrelevant, because they tend to test local information and grammatical structure within clause boundaries (Van Dijk and Kintsch, 1983); cloze tests are, therefore dismissed as "inadequate" and "unfair" for second-language learners and function more as a vocabulary exercise than as an assessment of reading comprehension (Bernhardt, 1993, p. 197). A number of criticisms are equally leveled against multiple-choice questions. They may be coercive in the sense that they strain the student to find the answer, knowing that the right one is there. They encourage guessing, but it is suggested that this problem can be partly overcome by increasing the number of answer choices. In the present case, the number of answer choices is four, which minimizes random guessing to one fourth. In spite of further, blanket criticism that multiple-choice questions are inadequate for determining a test taker's actual skill and comprehension, these questions are suitable for testing particular information, provided they are passage dependent and hold no grammatical clues (Bernhardt, 1993, p. 198). Conversely, we should remember that comprehension is a subjective activity, and therefore, open-ended questions, which are concerned with relationships rather than isolated facts, serve to free the reader's imagination.

Categories of Comprehension in Argumentative Discourse

213

Summary and recall are two different activities of the test; the latter includes as many details as possible, but avoids the negative effects of consulting the text, whereas the former contains only important and general ideas and, as such, tests the degree of macrocontrol exercised by the test takers. The present study combines the two, making use of the positive aspects of each. Concerning the diagnostic functions of recall techniques, experts have divided opinions. Bernhardt (1993), for example, thought that recall helps focus on important text information: "it is the preferred method of investigating foreign-language readers' comprehension"(p. 187) especially when recall is performed in the native language, because it allows gaps and misunderstandings to surface. Not all authorities concur. Smith (1988, p. 43), for example, remarked, "you do not prove that you have understood anything by repeating it." Recall is a short-term memory activity, and in Smith's opinion, only what is committed to long-term memory is what is understood. In order to avoid the pitfalls of rote memorization in direct recall, the present research uses translation, so that by converting the message into a different code there is a guarantee that what is operating is a comprehension strategy. Moreover, translation bridges the language barrier in expression. Apart from the lack of appropriate cultural knowledge, the lack of linguistic knowledge is sometimes the principal component in comprehension failures, noted Bernhardt (1993, p. 98). Writing in the reader's native language thus reduces the degree to which the message is deflected. For that reason, the subjects were asked to write their summaries in Arabic. The exercise combines the merits of both recall and summarizing. According to Maria (1990), summarizing can be done only when the text is understood. "Getting the main idea or summarizing is almost synonymous with comprehension"(p. 189). Making summaries is among the best learning and reading strategies for comprehension, according to Maria (1990, p. 160), since they are suitable for and helpful in a variety of texts. The summary is a form of condensed paraphrase that is essentially a sign of comprehension, a form of translation. For these reasons, the present test combines both summary and recall in one exercise (see Appendix 1). It does not require lengthy repetition as in recall, at the same time the requirement of a summary objectifies the structure and unity of the text.

Scoring and evaluation Open-ended questions (OEQs) were assessed subjectively. Model answers were obtained from the judges' responses, qualified by information from the passage analysis presented in Appendix 2. Questions that received no scores (OEQs 1, 2, 3, 11, 12) were attitudinal and opinion questions. The

214

Categories of Comprehension in Argumentative Discourse

summary was also assessed subjectively, since no reliable method for scoring such an exercise is available. Maria (1990) recommends this subjective approach, which allows the researcher to spot misinterpretation or restructurings in retelling. The open-ended questions that received scores were uniformly assessed with scores ranging from zero to five. Multiple-choice questions (MCQs) likewise scored between zero and four, with the exception of 8 and 9, which was scored between zero and four.

Sample population Sixty-one randomly selected fourth-year students (41 females, 20 males) who had just finished the first-term examination (for graduation in June, 1999) took the test. The students who came from Arabic secondary schools with fewer years of exposure to English, were 26; 35 students came from secondary schools that used the English language as a medium of instruction. Only three mentioned English or French as a medium of instruction in the biographical questionnaire. Ages ranged between 19 and 28 years, with 90% between 20 and 22; the extremes of age, 19 and 28, were each represented by a single student—the rest were 23 or 24 years of age.

Results and Discussion Opinion among the sample subjects as to whether the passage was enjoyable was almost equally divided. Most of those who considered it "boring" did not understand either the organization or the content or both, as could be deduced from analysis of the relevant answers. To exemplify, answers by two students were picked at random to represent each view: subject 16, who found it "boring" had a general test score of 28 of a possible 113. On the other hand, subject 46 scored 67 and found the passage interesting, because "I haven't understood it directly. It is somehow puzzling." Although both students had difficulty understanding the passage, the two attitudes were different: the one who performed less well overall was intimidated while the other, who performed fairly well, was challenged. Answers to OEQ 2 about how clear the reader found the passage also show that opinion was equally divided. Subject 12, who thought the writer "presents a hard to determine issue," found that in paragraph 7, the writer "sides" with the identifiable lives. As such, this student was unable to determine the writer's bias. Two subjects, who also said they enjoyed the passage, were aware that the writer is arguing for and against, but the latter subject thought that the writer is "leaving the choice to the reader"

Categories of Comprehension in Argumentative Discourse

215

because he is arguing for and against "each side." Another student who enjoyed the passage, "except paragraphs 8 and 9," found that the writer was "not able to get what he wants to say." One subject who answered OEQ 1 negatively found that the writer "seems as if saying the opposite of what he wants" (italics mine). In other words, several opinions are expressed explicitly but the writer's own is implied. Another student who did not enjoy reading the passage, in answer to OEQ 2 (clarity of passage), commented that the writer talks about saving two types of lives, but he is in favor of neither; in other words, the bias is not determined. Yet another student who did not enjoy the passage wrote that the writer clearly "hesitates"; this opinion was echoed by another subject who felt that the writer "goes in circles." Concerning the answer to OEQ 3 about places of special difficulty in the passage, only five subjects answered "no" and one wrote "I don't know." The answers by the remainder of the study population are summarized and listed below: • Difficult vocabulary includes "technical words," such as monetary theory, Medicaid, occupational safety, and liberal. • Other difficult issues are: the main idea (one subject), difference between statistical lives and identifiable, political names, the connection with the terrorist attacks, the conservative account of the problem, and how welfare-state policies focus on identifiable lives. One subject could not understand why, in paragraph 6, the writer says that "the resources are scarce and this is the reason why they cannot do both: helping those in need and preventing future needs." Another subject could not tell whether the writer is for or against open-heart surgery. Most prominently, Adam Smith's opinion proved enigmatic. • The difficult sentences are generally nos. 16, 20, 23, 24, 26, 29, 31, 34, and 39. • One subject found the passage "all difficult." • Four subjects found paragraph 7 difficult, while paragraphs 8 and 9 scored very high in difficulty: 21 subjects rated paragraph 8 and 10 rated paragraph 9 as difficult. These proved to be the two most problematic paragraphs of the whole text. In keeping with these judgments, 12 students suggested removing paragraphs 7, 8, and 9 in answer to OEQ 11, while in answer to OEQ 12, one subject found these paragraphs incoherent. These subjects among others probably lost track of the argument as the writer summed up his claims in preparation for the counterattack, which leads to his bias (see Appendix 2). Paragraphs 1-4 also proved a little troublesome: a number of subjects found at least three of them dispensable, whereas others found them confusing.

216

Categories of Comprehension in Argumentative Discourse

Score results Tables 1 and 2 quantify the general scores for each question for 26 students who had an Arabic secondary school education and 35 from English secondary school education respectively. OEQ's 1, 2, 3, 11, and 12, since they concern opinion and attitude, received no scores. Column 2 tabulates the total score on each question of each subpopulation, as calculated in points; questions ranged in value from four to six points (for details, see above "Scoring and evaluation.") Column 3 marks the cases in which the answer was "I don't know," was left blank, or was part of an incomplete paper (so-called nonanswers). Column 4 quantifies all the questions that were answered. Column 5 tabulates the percentage of scores obtained by all valid answers in the subgroup collectively to the total scores calculated for each question. For example, in the case of OEQ 6 of Table 1, there were only 22 valid answers. Since that question was scored from zero to four (five possible scores), the total against which the percentage is calculated is, therefore, 22 times 5, or 110. In this example, total points earned by the 26 students was 28.5; the total percentage of the valid answers in this subpopulation is thus 25.9% (28.5 divided by 110). Answers to microlevel questions (e.g., OEQs 7, 13, 14, 18; MCQs 3, 5, 6, 7) scored very low in the case of coherence elements connected with the superstructure, such as OEQ 7. Questions on coherence connected with local prepositional issues, such as OEQs 13, 14, and 18 and MCQs 3, 5, 6, and 7, scored much higher. OEQs connected with organizational issues on both propositional and superstructural levels (Qs 4, 6, 8, 10, 15; summary) scored much lower. In Table 3 the total percentage of scores drops lower as the questions increase in comprehension difficulty, i.e., as the categories became more abstract. The lowest scores for both groups is connected with the superstructure component. Macropropositional structure is higher and the micropropositional level is higher still. It is noticeable that all the percentages are, however, below 50%. It is also noteworthy that of 23 questions, the scores of 15 questions are below 50%. The only remarkable score in the test is that of OEQ 14 in both tables 1 and 2. The answer by two subjects of subgroup 2 included the title Wealth of Nations by Adam Smith, who was identified as a theorist in economy by another subject from the same subgroup. Other answers varied among "capitalist," "economist," "conservative," "philosophy," "politician," and "writer." The response to this question is relevant to background information on the topic of discourse. This particular item as I found out later from the students themselves is covered in the civilization course required for English majors at Cairo University. However, even knowing who Adam Smith was, the students revealed in their answers to OEQ 3 that his theory was not very clear to them.

Categories of Comprehension in Argumentative Discourse

217

Table 1. Score percentages per question attained by students (n = 26) from Arabic-medium secondary schools: open-ended questions. Subject

Total Score (points)

Nonanswers (no.)

Valid answers (no.)

Total Score (%)

26

12.3

OEQs 4

16

5

71

1

25

56.8

6

28.5

4

22

25.6

7

18

2

24

15.0

8

51

2

24

42.5

9

44

2

24

38.2

10

44

2

24

36.6

13

39

2

24

32.5

14

72

6

20

72.0

15

0

8

18

0

16

44

7

19

46.3

17

33

9

17

38.8

18

61

13

13

93.8

1

55

4

22

50.0

2

75

26

57.6

3

70

25

56.0

4

50

26

38.4

5

55

26

42.3

6

50

1

25

40.0

7

55

2

24

45.8

8

41

3

23

44.5

9

10

9

17

16.1

Summary

38

26

29.2

MCQs

1

OEQs = open-ended questions; MCQs = multiple-choice questions

Categories of Comprehension in Argumentative Discourse

218

Table 2. Score percentages per question attained by students (n = 35) from English-medium secondary schools. Total Score (points)

Nonanswers (no.)

Valid answers (no.)

Total Score (%)

4

5

1

34

2.9

5

52

3

32

32.5

6

23

3

32

14.3

7

3

4

31

1.9

8

41

3

32

25.6

9

60

3

32

37.5

10

31

4

31

20.0

13

37

35

21.1

14

101.5

35

58.0

15

6

5

30

4.0

16

67

9

26

51.5

17

44

6

29

30.3

18

76

15

20

76.0

1

60

1

34

35.2

2

100

1

34

58.8

3

115

35

65.7

4

75

1

34

42.8

5

20

1

34

11.7

6

50

1

34

29.4

7

70

35

40.4

8

62

1

34

45.5

9

14

1

34

10.2

Summary

21

2

33

12.3

Subject OEQs

MCQs

OEQs = open-ended questions; MCQs = multiple-choice questions

Categories of Comprehension in Argumentative Discourse

219

Table 3. Total score percentages classified according to the categories of comprehension used in this study. Micropropositional level

Macropropositional level

Subgroup 1

Subgroup 2

Subgroup 1 Subgroup 2 Subgroup 1 Subgroup 2

49.6%

38.2%

52.5%

29.1%

Superstructure

35.1%

28.7%

Subgroup 1 = 26 students with an Arabic-medium secondary school background; subgroup 2 = 35 students with an English-medium secondary school background. OEQ 4 scored very low. The concessive function of the item "true" completely escaped the test takers' comprehension. The highest score recorded for this question was one (of five), for an answer by a single subject from subgroup 1 that only hinted at the idea of an argument, with descriptors such as: "to expose both sides of the issue," "a fact and its opposite," as well as "to expose opposite opinion." The most popular answer here was "to assert a fact" or "to emphasize his idea." Equally disconcerting is the answer to OEQ 6: all forms of hedging in the article were not recognized, neither the concessions in paragraphs 1-6, and 9 nor the modal verbs and hypothetical start (sentence 1) and finish (sentences 36 and 37), which set the tone for the whole passage. The answers to OEQ 8 were expected to elicit a response containing the word "argue" or any of its derivatives or synonyms. Answers, instead, varied among "wanted to choose between...," "explain...," "show...," "state a fact...," "how to...," etc. Only two subjects from subgroup 1 hinted at "discussing two opposite views" while a third, from subgroup 2, mentioned the word "argue." Finally, another student from the subgroup 1 explained that "the writer wants to solve the conflict or the dilemma of saving souls and saving money. The conflict is between helping identifiable lives and saving statistical lives." This is the closest that anyone got to the topic of the argument, despite the misunderstanding about "saving money." Equally important is that in answering OEQ 9, responses ranged from getting the right answer to completely missing the point. Many mistook the writer's position for either the side of saving identifiable lives or being able to do both or "don't have to choose." Other curious answers include one that considered one of the views expounded in the passage to be that of Adam Smith, and another that indicated that the "writer's view is the whole text." In answering OEQ 10 (identify one word and one sentence as the most important in the passage), very few got both word and sentence right. The

220

Categories of Comprehension in Argumentative Discourse

majority got the sentence right, but missed the word. Some of the answers to what is the most important word were "regrettably," "true," "dilemma," "identifiable," "welfare state," "needs," "statistical," "goals," "policy," and "choose." The only acceptable answers were considered to be "choose" and "dilemma." The answers to OEQs 15 and 16 about the passage's superstructure scored widely: for Q 15 (the relation between paragraph 5 and paragraph 6) ranged from 0% in Table 1 to 4% in Table 2. Once more, the functions of stating the claim in paragraph 5, i.e. confirming the dilemma, and the counterargument in paragraph 6, were not clear to any of the students. It is interesting to note that the judges too were not sure about this function either. Only one answered that paragraph 6 supports the assumptions in paragraph 5, without referring to the counterargument. The other two identified the relationship as "problem and solution" and "reality and choice" (see Appendix 2). The response to OEQ 16 (the role of paragraph 7 in the development of the text) reveals a very important fact concerning the judges' opinion. It was called "argument," "explanation," and "further classification and setting up the rest of the argument" by the three judges respectively. None of these answers touches on the real function of the paragraph. Sixteen subjects, however, gave the answer "summarizes" what went before, "is a summary," or "sums up." All were considered correct (see analysis in Appendix 2). One subject, from group 2, called it a "climax," two said it "raises/reveals conflict," and one said "analysis of problem." Responses to this question are strongly confirmed by the score of MCQ 2, which asks the same question in a different form. Paragraph 7 is transitional between the first half of the argument, which attempts to confirm the dilemma by asserting that we have to choose, and preparation for the counterargument and attack against the welfare-state position in order to declare the choice we ought to make. Although study participants were aware of the propositional content of the paragraph, nowhere else in their answers, including the summary, is there a sign of fitting it into an overall schematic structure. The writer's choice, his conclusion, is the subject of OEQ 17. Five participants said the writer took the side of identifiable lives, one of whom added that in so doing, he is subtle and indirect. This is actually the opposite of what the writer concludes in the passage. Six students said the author had no definite conclusion, one said the writer concluded that we could do both, and yet another thought the writer was puzzled. The MCQs were mostly on the microlevel. Both subgroups scored fairly well on MCQ 8, which was about fact or opinion, whereas scores for both groups on Q 10, which marks the sentence relations in the paragraph, an organizational issue, again, were low. The summary was meant to further clarify and probably underline

Categories of Comprehension in Argumentative Discourse

221

some of the findings resulting from the response to questions. Several subjects missed the central issue in the argument, and no one mentioned the word "argument." One student turned the issue into a moral question of whether it is right to let people who happen to be sick for no fault of their own suffer, while others saw it as a sociopolitical issue of class distinctions. To many, the writer's conclusion "was to do both." A good number of the summaries were either too vague or inadequate, giving only part of the issue. But the most important remark to make is that the article was invariably received as a piece of exposition, either explaining a viewpoint or showing/explaining a situation. The closest to understanding the article as a semiargument was a student from subgroup 1 who missed the aspect of "choosing" altogether. (By semiargument I mean recognizing the content as explaining a dilemma but unaware of the organizational pattern.) A general remark is due here: Subjects mostly quoted sentences verbatim when answering questions nos. OEQ 8, 9, 10, 17, etc., instead of using their own words.

Findings The above results show that in an attempt to comprehend the passage, the greater part of the sample population attended to local and sentential details of prepositional meaning and organizational structure at the expense of macropropositional meaning and superstructure. The answers to OEQs 8, 9, and 10 reveal that subjects were not in possession of a macropropositional component, which is an essential mechanism for securing macrocontrol and allowing for successful predictions. The correlation between the scores of OEQ 16 and MCQ 2 is further substantiation of this fact. More students in both groups connected the phrase "make three points" to "summarize" or "sum up" in MCQ 1, but fewer could connect the paragraph with the transitional function of "summing up" what came before to prepare for the next phase. Subjects were aware of fragments of the propositional and superstructural levels, but could not make them cohere. They could not weave the pieces together because the controlling schema was absent. Since students at the Department of English, Cairo University, are taught the patterns of argumentative discourse throughout their four years of study, it can be concluded here that either they are not able to generalize the schema or that when the rigid pattern they are acquainted with is slightly broken—they feel lost. The zig-zag nature of the argumentative pattern used in the test passage is not very common in writing argumentative discourse. The writer kept shifting from claim to opposition, from argument to counterargument in almost every paragraph (see Appendix 2). These shifts must have been rather puzzling

222

Categories of Comprehension in Argumentative Discourse

to students, who are rigidly trained to expect a stereotyped pattern—once the sequence is changed, the text turns featureless. The importance of text structure in comprehension, particularly in the absence of cohesive devices, was alluded to by Bernhardt (1993) and Van Dijk and Kintsch (1983); see "Comprehension" above. The misinterpretation of the repeated use of "true" in the early part of the passage points to the fact that in the absence of a control schema, the concessive function of this item was missed. Significantly, "true" was understood primarily as an "emphatic" element, which brings to mind Van Dijk and Kintsch's assertion that the reader's cultural background usually dominates the writer's in the process of reader comprehension of the writer's work. The "directness" required by readers of Arabic for making proof by emphasizing and presenting the writer's argument, which is a linear development, hampers the indirect functions of concession and counterargument in an English text, which follows a top-down hierarchical development. The summary, therefore, demonstrates that the passage was received as a piece of exposition rather than argumentation and as a problem-solving technique at best. However, the discrepancy in the two subgroups' scores to OEQ 7 (15% and 1.9%) and MCQ 1 (55% and 60%) is puzzling, since both questions are related to the same claim in the same argument. However, the type of inference involved in MCQ 1 is different from that involved in Q 7: the former is on the macropropositional level, whereas the latter is on the schematic superstructural level, requiring the answer aptly given by one of the judges, concession-assertion. The conventional schema of argumentation in English (claim, counterclaim, refutation/concession, conclusion) with its ordering variation (see Maccoun, 1983, for a detailed discussion of these variations) may be present in the subjects' awareness only by name. Application is in the most part faulty. The lack of comprehension for those who did not even partly grasp the schema resulted in their judging the passage as boring. For those who were aware of the intricacies of the schema it was challenging and enjoyable. However, one subject who seemed to appreciate the "subtle and indirect" tone of the passage still failed to recognize the bias of the writer; to this student, the writer "prefers" identifiable lives. Comprehension strategies by study participants followed a bottom-up direction, no macrocontrol was exercised, and due to the zig-zag nature of the argument, the data proved to be puzzling and off-putting at times, let alone completely misinterpreted. Maccoun (1983, p. 95) suggests a tentative scale of developmental stages of acquisition order of the comprehension of argumentative discourse; at each stage the learner acquires:

Categories of Comprehension in Argumentative Discourse

223

• views discourse as neutral (stage 1) • mistakes opposition's argument for author's argument (stage 2) • acquires local bias (stage 3) • correctly identifies author's global argument; acquires direct refutation (stage 4) • acquires concession (stage 5) • acquires local and global bias, direct refutation and concession (stage 6) • acquires sarcastically marked disagreement (stage 7). If placed on such a scale, at least 51.7% of the sample population is in stage 1. These students those who considered the text an exposition, commenting that the writer "tries to make people care about...," "states a new policy...," shows two types of....," "talks about...," or draws the reader's attention to " We find 13.7% in Stage 2; 12% in Stage 3, and 10 refrained from answering OEQ 8. The remainder of the stages are not represented in this sample. No one was aware of the categories of refutation or concession combined with the global bias. According to this classification, after three and a half years of training in reading comprehension, most students are still in the early stages of understanding argumentation. They have learned the labels, because words like "argument," "counterargument," and "opposition" appear sporadically in their answers. But in answering OEQ 8 about the writer's intention, the answers included comments like "showing the reader," "exposition," "wanting to choose," etc. Only one subject used the term "argue," but it was to "argue for saving identifiable lives" rather than statistical lives. Similarly, in answering OEQ 17 about the writer's conclusion, six said "I don't know," eight said "no conclusion," a few said "no solution" or "we could do both," and some thought the writer sided with identifiable lives or that the writer was "puzzled." It is clear, therefore, that the subjects did not show any ability in locating components of the argumentative passage or in identifying their functions. The theoretical aspect of their knowledge could not be put into practical use. They are not in control of a discovery procedure whereby, when faced with something new, they can probe into its nature to discover its identity. Labels may be required for comprehension, but what is more important is to be able to attach those labels to their respective functional loci in order to recognize the type of discourse and its superstructure, without which no true comprehension is possible. The students' control of coherence strategies also proved quite inadequate. They tended to concentrate on details of an intrasentential nature and on word meanings more than on global coherence strategies. In answering OEQ 6 (use of modals), most students connected modals with

224

Categories of Comprehension in Argumentative Discourse

their habitual meanings, e.g., many students remarked to the effect that "the writer is not sure if he is right or wrong," "expresses uncertainty," "evokes doubt," or is a "sceptic," but none connected them with their more global function as hedges. Students may have understood the semantic content of the word "true," but they totally missed its pragmatic force, its concessive function. Used so many times in a sequence, in their answer to OEQ 4, the repetition was mistaken for "emphasis," obviously a case of cultural transfer. The transfer from Arabic argumentation by repetition of one's claim and ignoring the opposition may have been what prompted the study subjects to interpret "true" in this context as the author's way "to assert a fact" (eight students), " to state a fact" (four students), "to assure" or "to make sure" or "to emphasize facts" (one student each). Equally significant has been the lack of global coherence: most students were unable to make the correct inference as to the direction the writer's argument was taking, and thus they mostly missed the writer's bias, i.e., the purpose of the argument. It may be already clear from the facts of Table 3 that the impact of a prolonged period of exposure to English and consequently more proficiency in the language is not a determining factor in comprehending global designs or superstructures. Since argumentation is a text type that greatly depends on its form, it follows that proficiency in the language alone does not aid comprehension. Table 3 shows that on all three levels, students from an English-medium background scored substantially lower than those from an Arabic-medium background. The macropropositional level shows a wider gap than the micropropositional level, and the gap closes a little on the level of superstructure. The consistency is surprising in view of the fact that the sample of the English group is larger than that of the Arabic group. It can be concluded that the comprehension of texts such as argumentation depends on training rather than language proficiency. Students with less proficiency in English may be working harder on the technical sides of their courses to make up for the language deficiency.

Acknowledgments I would like to thank Professor Robert Kaplan of the University of Southern California for his encouragement during our discussion of the initial plan for this research at a very early stage. My thanks also go to Khaled Tewfik and Manar Shalaby of the Department of English, Cairo University, for assisting in carrying out the test and in processing the data.

Categories of Comprehension in Argumentative Discourse

225

Notes 1 El-Seidi (1996) compares the rhetorical structure of English and Arabic expository and argumentative discourse in the writings of native speakers and finds more similarities than differences in applying argumentative strategies between the users of both languages. The present study differs in two ways: it assumes that argumentation and exposition are two different text types and it investigates comprehension rather than production. 2 El-Shershabi (1988), for example, discusses aspects of cohesion and coherence in Arabic in general, irrespective of any constraints or special specifications that may be imposed by argumentative discourse. 3 Such cultural strategies have a wide scope; they involve knowledge about different geographical areas, social structure, institutions and events, different communicative events, different languages and discourse types, speech acts, superstructures, different local and global coherence conditions, styles and rhetorics, different symbolic or ritual values and functions, different beliefs, opinions, attitude ideologies, norms and values, conceptual ordering of the world and society, and finally, different objects of reference (Van Dijk and Kintsch, 1983, p. 81). 4 Aristotle demands that the arguer "first present [his] own argument, and then meet the opposing arguments by direct refutation or by pulling them to pieces...because the mind of the hearer...refuses an argument if the opposing speaker has made a good impression" (Cooper, 1932, p. 235). 5 They are "higher complexes (and even conventional or habitual) knowledge structures"(Van Dijk 1981, p. 141), which prop up ideas in the organization and interpretation of experience (Brown and Yule (1983, p. 247). They help us predict aspects of our interpretation of discourse. Therefore different cultural backgrounds result in different schemata. (This is a different use of the term schema from that of the same term in "Categories of comprehension of argumentation" above.) 6 Tone, or writer's attitude toward subject or reader, i.e. his/her voice, can be angry, sarcastic, conciliatory, objective, whimsical, bitter, facetious, mocking, joyful, advisory, indifferent, sympathetic, ironic, judgmental, distanced, flippant, etc. It is revealed either directly (i.e., stated) or indirectly by parody, exaggeration, understatement, paradox, and so on (Levin, 1990, p. 391). 7 Topic shift marks boundaries between paragraphs. However, Hatim (1997) demonstrates that this is not always the case. Mistaking an

226

Categories of Comprehension in Argumentative Discourse

intraparagraph topic shift for the continuation of the same topic results in faulty chunking of the text, miscomprehending, and mistranslating. 8 Sa'deddin (1989, p. 44) reports that the English speaker regards the text producer in Arabic as "trespassing" and "breathing down the neck of the audience." 9 In the following extract from Mohammad Sadiq al-Rafi's I'jaz alqur'an ("The inimitability of the Quran," p. 210), all pronominal references are in boldface. It is a case of complexity caused by excessive anaphora within one sentence, which may render understanding difficult: fa-laysat fiha [ i.e. anwa'i al-balagha] 'isti'aratun wa-la majazun wa-la kinayatun wa-la shay 'un min mithli hadha yasihhu fi al-jawazi aw fima yasa'hu l-'imkanu 'an yasluha ghayruhu fi mawdi'ihi 'idha tabaddaltahu minhu fadlan 'an 'an yafi bihi, wa-fadIan 'an 'an yurbi 'alayhi, wa-law adarta al-lughata kullaha 'a/a hadha al-mawdi'. 10 Through personal contact, I learned from a Syrian linguist that students in Syria find it difficult to understand what refutation is all about in an English argument. 11 Ibrahim Salama (1952) tells the story of a debate between Al-Sirafi who rejected Aristotelian logic and Qudama who upheld it, in which the former's argument rested on moral, authoritarian opinion based on eloquence and style with no attempt at proof. He was commended for his argument and is reported to have won the debate against Qudama's austere defense of down-to-earth logic. 12 It is interesting to note that judge 2 (British) also found the passage difficult on the first reading, particularly the difference between "identifiable" and "statistical" lives.

Appendix 1: Test used for data collection READ EVERYTHING BEFORE WRITING ANYTHING. Carefully read the following passage several times and then answer the questions below. Please attempt all the questions, but if there is any question you feel you could not possibly answer, just write "I don't know" in front of the question number. Otherwise try your best because a good deal depends on your response. There is a big difference between the answers "I don't know" and "No there isn't" in response to Questions 5 and 12, for example. There is no "right" or "wrong" answer, so please write only what you know. Please put your name and class on each paper you hand in. Sentences are numbered in superscripts at the beginning of each sentence, while paragraphs are numbered in bold digits at the beginning of each paragraph.

Categories of Comprehension in Argumentative Discourse

227

Part I: The following abbreviations are used: S = "sentence," S's = "sentences," P = "paragraph," and P's = "paragraphs." Statistical Lives by Alan Wertheimer, New York Times, 1980. 1 Suppose the following were true: 1 2At least some money spent on open-heart surgery could be used to prevent heart disease. 3True, patients in need of such surgery might die, but many more lives would be saved. 2 4Some money spent on treating tooth decay among low-income children might be used on fluoridation and dental hygiene. 5True, some decay would go untreated, but fewer children would ever need such treatment. 3 6We could prohibit ransom payments to kidnappers. 7True, kidnapped children might die, but by lowering the incentive to kidnap, fewer children would be taken. 4 8We could drastically reduce unemployment compensation. 9True, the unemployed would suffer, but by converting the money saved to private investment and by lowering the incentive to stay jobless, there would be substantially less unemployment. 5 l0These cases exhibit a similar structure. 11All involve choosing between a policy designed to help specific persons and one that seeks to prevent the need for such help. 12These choices are especially difficult because we know who needs help. l3The patient requiring open-heart surgery, the kidnapped child, the unemployed autoworker—they have names and faces, they are "identifiable" lives. 14 On the other hand, we do not know whose lives will be saved or who will benefit from the prevention of heart disease, tooth decay, kidnappings, or creation of new jobs. 15Some people will, and we may be able to estimate their numbers with precision. l6These are real lives, but they are only "statistical" lives. 6 l7We might say we do not have to choose between helping those in need and preventing future needs. 18After all, we could do both. 19But resources are scarce, and even when resources are not at issue (as in the kidnapping case), we often must choose between competing persons and goals. 20We cannot do everything we might like to the extent we might like. 21We must often choose between helping identifiable lives and saving statistical lives. 7 22I wish to make three points about these dilemmas. 23First, we do seem to favor the interests of identifiable lives (saving the kidnapped child) and it may not be irrational to do so. 24Second, we nevertheless do see the need to attend to the interests of statistical lives, even if this injures identifiable lives. 25Thus it is now common to hear people advocating directing more medical resources to primary prevention of disease and fewer to treatment. 26The policy of refusing to negotiate with terrorists may risk the lives of some hostages, but we do see the point. 27 Third, welfare-state policies focus on identifiable lives, whereas conservative economists prefer to focus on statistical lives.

228

Categories of Comprehension in Argumentative Discourse

8 28Monetary theory and other technical issues aside, the new Adam Smith tells us that however well intentioned, welfare-state policies have not (always) worked— on the policies' own terms. 29Minimum wage laws, unemployment compensation, consumer protection, occupational safety, Medicaid, Social Security—by interfering with market efficiency, by discouraging individual initiative, by impeding private capital formation, by incurring large-scale expenditures on governmental bureaucracies—all these policies (and others) have been self-defeating.30They argue that liberal economics, filled with concern for the genuine needs of identifiable lives, has swelled the future ranks of statistical lives in need. 31Welfare-state humanitarianism is shortsighted, they say, and if thus less humanitarian than we may believe. 9 32we need not dwell on the accuracy of this account. 33Conservative accounts may be wrong about the facts. 34We certainly need not assume that market choices and private-capital formation always serve the interests of all social groups, that regulation always does more harm than good. 35But suppose conservative economists are (sometimes) right about the facts. 36suppose that attempts to serve the needs of identifiable lives do end up harming statistical lives. 37should we turn our back on the needs that we see in order to prevent those that we cannot see? 38Regrettably, the answer may sometimes be yes. (From Levin, 1990, pp. 250-252) A. Answer the following questions on a separate sheet: 1. Did you enjoy reading the passage? Why? 2. Does the writer make himself clear in what he wants to say? How? 3. Did you find some points particularly difficult to understand? Give as many examples as you can. 4. Why do you think the writer uses the word "true" so many times? (See S's 1, 2, 3, 4, and 5). 5. Is there a sentence that can be said to aptly summarize the first four P's? Which one? 6. Why does the writer use words like might and would in S's 1-9 so repeatedly, and the word suppose in S's 1, 35, 36? 7. What do you think is the relationship between S 2 and S 3? 8. What do you think the writer is trying to say and/or do in this passage? 9. How many points of view do you think are presented? Name them. Which is the writer's? 10. Can you point out one word in this passage that you may call the most important word? Can you do the same with a sentence? 11. Are there any sentences/paragraphs in this text that can be safely removed without affecting its development? 12. Is there a sequence of sentences that do not follow each other clearly and meaningfully, i.e. a sequence, which seems to puzzle or confuse you? 13. Join S's 8 and 9 into one sentence, using the suitable connective.

Categories of Comprehension in Argumentative Discourse

229

14. If you do not know who Adam Smith is, can you make a good guess? 15. Describe the relationship between P 5 and P 6. 16. What is the role of P 7 in the development of this text? 17. What is the writer's "conclusion" in this article, if any? 18. In P 9, which are the needs "we see" and which are those we "cannot see"?

B. Answer the following questions on this sheet. Select the most suitable item to complete sentences; place the letter, which marks your choice, in the space provided: 1. The author is open-heart surgery; he is heart disease. a. strongly against b. against c. in favor of d. neither against nor in favor of How did you find out? 2. Select the statement, which is a likely paraphrase of S 22: a. There are three points to summarize these dilemmas. b. I wish to add three points to these dilemmas. c. I wish to give some examples to illustrate these dilemmas. d. I can explain these dilemmas in three points.

preventing

.

3. Select the statement, which most aptly summarizes S's 32-36: . a. The facts presented by the conservatives are wrong, and so they are rejected. b. The facts presented by the conservatives are right and acceptable. c. These facts are partly right and partly wrong. d. There is a degree of skepticism that should be taken into consideration. 4. When the author says, "After all, we could do both," he is expressing a. his own opinion b. the general public opinion c. the conservative opinion 5. "These are real lives, but they are only 'statistical' lives" (S 16) implies a. That is why they should not claim much attention. b. That is why they have not claimed much attention. c. They should claim much attention. d. They have claimed much attention.

230

Categories of Comprehension in Argumentative Discourse

6. "Statistical' lives" in S 16 means . a. cannot be counted c. are probable b. can be counted d. are unknown to us. 7. The word "thus" in S31 refers to the constituent a. less humanitarian c. short-sighted b. welfare-state d. humanitarianism 8. Decide whether the following sentences are "fact" or "opinion." Write F or O before each sentence: _A. (S12) these choices are... we know who needs help. _B. (S 20) we cannot do everything we might like to the extent we might like. C. (S 28) Minimum wage laws, unemployment compensation...all these policies (and others) have been self-defeating. D. (S 25) Thus it is now common...and fewer to treatment. 9. In P 6 the main point made is ; the other points are subordinated to it. Name the type of relationship you think exists between the main point and each of the supportive points in the space provided after each sentence, except the main idea, of course. a. After all, we could do both. b. We do not have to choose between helping those in need and preventing future needs. c. We must often choose between helping identifiable lives and saving statistical lives. d. Resources are scarce.

Part II: On this sheet, write a good summary in Arabic of what you can remember of the passage you have just read. Do not exceed 80 words.

Appendix 2: Structural Analysis of the Text This passage is an argumentative text type that Levin (1990) tells us was published in the New York Times in 1980, written by Alan Wertheimer, a professor of political science at the University of Vermont, about government spending: the choice between helping identifiable lives and saving statistical lives:

Categories of Comprehension in Argumentative Discourse

231

Knowing that a large segment of his audience supports what he refers to as 'welfare-state humanitarianism,' Wertheimer uses the dilemma to force these people to recognize that the issue is complex, does not present a simple choice between right and wrong, and demands an examination of basic assumptions. (Levin, 1990, p. 250). What follows is the researcher's structural analysis of the text used, on which the test questions were based. The writer's commitment is divided into two stages: the assertion of the presence of the dilemma, the fact that we have to choose, and then the choice he actually makes. P 7 is the transition between the two stages: it sums up the previous stage, making positions clear in order to launch the attack and make a choice. Each stage consists of a claim, an opposition, a counterargument and a conclusion. The initial hypothetical situation is later turned into a probability, then into facts and testimonies inductively leading to a conclusion. The generalized predicate in S1 is hypothetical: it begins with "suppose," thus setting the concessive tone from the very start: the issue is not presented forcefully. Such a hypothetical situation at the beginning helps in mitigating the imposition of the argument by placing it a few steps away from fact. Four concrete examples are given in a row, each one mirroring the deductive superstructure of the overall argument in a syllogistic miniargument with claim (S 2), concession (first half of S 3) and counterargument/conclusion (second half of S 3). The writer's bias is carefully hidden, as part of his tactic is not to rock the boat too hard since he addresses an audience which is already biased toward saving identifiable lives. The four examples, therefore, present the situation quite impartially. "True" is another concessive and hedging tool used to avoid antagonizing the unsympathetic public. In asserting the dilemma in the first stage, the writer is not taking sides yet. He is only trying to confirm the fact that the choice is not as simple and clear as the public may think. Stage two is based on the deductive conclusion that saving more lives later (statistical lives) by spending money on prevention is better than saving less lives now (identifiable lives) by providing treatment for individual cases. In each example in P 1-4, the "true" statement is carefully balanced against the "but" statement, which gives the impression of an absence of bias. This balance is kept throughout the whole passage; both positions are right, until the very last word which slightly tips it to one side. However, there are subtle clues as to the direction in which the argument would take us, namely the use of such comparatives as "more" and "at least." Besides, if we try to map the layout of the four examples against Kaplan's (1983) rhetorical pattern of discourse focus (which is the prominence of certain information realized by surface structure syntactic means as opposed to discourse topic, which is semantic and a feature of deep structure), it is clear that the focus falls on the second half of S2, which expresses the stance the

232

Categories of Comprehension in Argumentative Discourse

writer adopts at the end of the argument, whereas the popular opposition is contained in topic position in the first half of the same sentence. The parallelism created by the repetition of four examples in exactly the same structure with the same pause positions creates a rhythm, underlines the message and also gives a feeling of relentlessness; it is an effective almost oratorical technique. P 5 introduces the claim: there is a dilemma and the choice is difficult but essential. P 6 provides an opposition to this situation, counterargues this opposition, and ends by reasserting the claim: we must choose. Then P 7 acts as a transition, summarizing the situation and introducing the parties involved. Throughout this stage, there are no explicit markers of bias. The writer is always hedging his position by using modals such as "might" and "would" and (frequency) adverbs like "sometimes" and "regrettably," by introducing other people's views ("They argue that..."), and by using the passive voice. But notice the use of some stylistic cues that pave the way for the writer's choice, such as the comparative "more" and "at least." P 8 slightly tips the balance to one side and begins the preparation for making a choice by launching the attack on welfare-state humanitarianism from a conservative point of view. Real-life factual information is used to support the counterargument inductively. It is full of negative connotations, which are immediately balanced by S 32 at the beginning of P 9 with the inclusive "we," which achieves full identification with the reader. But the argument is still in the third person. It is noteworthy that the audience is very much in the foreground in this article. The writer is hedging his argument, avoiding any strong stance in order not to provoke an already opposing public. P 9 begins with a concession and presents a counterargument to the claims made in P 8, which restores the balance to the argument once more. So far, we as readers are under the impression that both sides could be right, and this is the writer's ultimate aim: the choice is difficult. It remains for the writer only to declare his choice. The intensity of the argument is final: a "yes" comes after a rhetorical question (S 37) to declare the writer's choice concisely but emphatically. The question-answer technique, however, reduces this intensity to an acceptable degree. The passage, therefore, begins with a hypothetical "suppose" and ends with an emphatic "yes." There are several conclusions to make here: • Although the title seems to be a clue as to the writer's prospective choice, this choice is deeply embedded in the argument, hedged by several tools and surfaces only at the end due to the political and economic importance of the issue and the unsympathizing audience. • The frequency of a proposition and its paraphrases establishes its importance. The words "choice" and "choose" according to this principle becomes the key word in this passage, and S 21, which contains "choose," is the key sentence. • The writer casts his vote only at the end. It remains only a vote and a personal choice, but a difficult choice at that.

Categories of Comprehension in Argumentative Discourse

233

References Adams, R., and Brody, J. 1987. Reading beyond words. New York: Holt, Reinhart and Winston. Baker, G. P. 1896. The principles of argumentation. Boston: Ginn. Bernhardt, E. B. 1993. Reading development in a second language: theoretical, empirical and classroom perspectives. Norwood, NJ: Ablex. Brown, G, and Yule, G. 1983. Discourse analysis. Cambridge: Cambridge University Press. Connor, U. 1986. Argumentative patterns in student essays: Cross-cultural differences. In Connor and Kaplan (1986), pp. 57-71. , and Kaplan, R. B. (eds.). 1986. Writing across languages: analysis of L2 texts. Reading, MA: Addison-Wesley. Cooper, L. 1932. The rhetoric of Aristotle. New York: Appleton-Century Croft. Daif, Shawki. [1963]. Al-'asr al-Islami. Cairo: Dar al-Ma'arif. De Beaugrande, R., and Dressier, W. 1983. Introduction to text linguistics. New York: Longman. Fareh, S. I.1988 Paragraph structure in Arabic and English expository discourse. Ph.D. diss., University of Kansas. Freedman, A., Pringle, L, and Yalden, J. (eds.). 1983. Learning to write: first language/second language. London: Longman. Grellet, F. 1988. Developing reading skills : a practical guide to reading comprehension exercises. Cambridge: Cambridge University Press. Grismore, A., Markannen, R., and Steffenson, M. S. 1983. Metadiscourse in persuasive writing: a study of texts written by American and Finnish students. Applied linguistics, 4:39-69. Halliday, M. A. K., and Hassan, R. 1976. Cohesion in English. London: Longman. Harrison, A. 1983. A language testing handbook. London: Macmillan. Hatim, B. 1997. Communication across cultures: translation theory and contrastive text linguistics. Exeter: University of Exeter Press. Johnson, K. 1983. Communicative writing. In Freedman et al. (1983), pp. 247-257. Johnstone, B. 1987. Parataxis in Arabic: modification as a model for persuasion. Studies in language 11:85-98. . 1991. Repetition in Arabic discourse: paradigms, syntagms, and the ecology of language. Amsterdam: John Benjamins. Johnstone-Kock, B. 1983. Presentation as proof. Anthropological linguistics 25:47-60. Kamel, G. W. 1989. Argumentative writing by Arab learners of English as a foreign language: an empirical investigation in contrastive rhetoric. Ph.D. diss., University of Pennsylvania.

234

Categories of Comprehension in Argumentative Discourse

Kane, T. S. 1983. Oxford guide to writing: a rhetorical handbook for college students. New York: Oxford University Press. Kaplan, R. 1966. Cultural thought patterns in intercultural education. Language learning, 16:1-20. . 1972. The anatomy of rhetoric: prolegomena to a functional theory of rhetoric. Philadelphia: Center for Curriculum Development. . 1983. Contrastive rhetoric: Some implications for the writing process. In Freedman et al. (1983), pp. 139-161. . 1986. Cultural thought patterns revisited. In Connor and Kaplan (1986), pp. 9-21. Al-Khawli, Amin. 193l.Al-balagha al-'arabiya wa-athar al-falsafa fiha . Cairo: Al-Jam'iya al-Jughrafiya al-Malikiya. Levin, G. 1990. Prose models. San Diego: Harcourt Brace Jovanovitch. Longacre, R. 1979. An anatomy of speech notions. Lisse, Netherlands: Peter de Ridder Press. Maccoun, W. 1983. On the acquisition of argumentative discourse from a comprehension point of view. M.A. thesis, University of California at Los Angeles. MacDonald, D. 1983. The language of argument. New York: Harper and Row. Maria, M. 1990. Reading comprehension instruction: issues and strategies. Parkton, MD: York Press. Meyer, B. J. F. 1985. Prose analysis: purposes, procedures, and problems. In B. K. Britton and J. B. Black (eds.), Understanding expository texts. Hillsdale, NJ: Lawrence Erlbaum, pp. 11-64. Miller, C. J. R., and Kintsch, W. 1981. Readability and recall of short prose passages: a theoretical analysis. Journal of experimental psychology: human being and memory, 6:335-354. Perelman, C. L., and Olbrechts-Tyteca, L. 1969. The new rhetoric: a treatise on argumentation. Notre Dame, IN: University of Notre Dame. Raphael, C. B., and Newman, E. G. 1983. A rhetorical reader for ESL writers. New York: Macmillan. Ricento, T. K. 1987. Aspects of coherence in English and Japanese expository prose. Ph.D. diss., University of California at Los Angeles. Renkema, J. 1993. Discourse studies: an introductory textbook. Amsterdam: John Benjamins. Sa'deddin, M. A. 1989. Text development and Arabic-English negative interference. Applied linguistics, 10:36-51. Salama, Ibrahim. 1952. Balaghat aristu bayna al-'arab wa al-yunan. Cairo: Anglo-Egyptian Bookshop.

Categories of Comprehension in Argumentative Discourse

235

El-Seidi, M. A. 1996. Rhetorical structure in English and Arabic expository prose: a cross-linguistic study. Ph.D. diss., University of Cairo. Al-Sharkawi, H. 1986. Al-jadal fi al-qur'an. Alexandria: Munsha'at alma'arif. El-Shershabi, M. A. H. 1988. Substitution and lexical cohesion in the editorial argumentative discourse of Arabic and English. Ph.D. diss., University of South Carolina. Smith, F. 1988. Understanding reading. Hillsdale, NJ: Lawrence Erlbaum. Stevens, B. K. 1983. Discovery: an inductive approach to college writing. New York: Holt, Reinhart and Winston. Van Dijk, T. 1977. Text and context: explorations in the semantics and pragmatics of discourse. London: Longman. . 1981. Review of R. O. Freedle (ed.), New direction in discourse processing (Advances in discourse processes, vol. 2). Journal of linguistics, 17:140-180. Van Dijk, T., and Kintsch, W. 1983. Strategies of discourse comprehension. New York: Academic Press. Vivelo, F. R. 1978. Cultural anthropological handbook. New York: McGraw-Hill. Wertheimer, A. 1980. Statistical lives. In G. Levin, 1990, pp. 250-252. Whateley, R. 1969. Elements of rhetoric. Carbondale, IL: Southern Illinois University Press (reprint of 1828 ed.).

This page intentionally left blank