THE USE OF MODERN STANDARD ARABIC WORD FORMATION PROCESSES BY ENGLISH-SPEAKING AND FRENCH-SPEAKING ADULT

L2 LEARNERS AND NATIVE SPEAKERS

Rabia Redouane

A thesis submiüed in confonnity with the requirernents for the degree of Doaor of Philosophy

Department of Curriculwn, Teaching and Leamhg Ontario [ n s t i ~ e fbr Studies in Educa!ïc~n of the

University ofToronto

O Copyright by Rabia Redouane 200 1

National übrary of Canada

BiMiotheque nationale du Canada

A uisitions and Acquisitions et ~ i a i o ~ r a ~ h i c Services sewicer biôiiographiques 395 Wellington Street 395, rue Weüinqton Mtawa ON KI A O N 4 OttawaON K 1 A W Caneda canada

The author has granted a non- exclusive licence aiiowing the National Lhrary of Canada to reproduce, loan, distri'bute or seli copies of this thesis in microform, paper or electronic formats.

The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts fiom it may be printed or otherwise reproduced without the author's permission.

L'auteur a accordé une iicence non exclusive permettant à la Btbliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous Ia forme de microfiche/f~ de reproduction sur papier ou sur format électronique.

L'auteur conserve la propriété du droit d'auteur qui protège cette thèse. Ni la thèse ni des substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation.

THE USE OF MODERN STANDARD ARABIC WORD FORMATION PROCESSES BY ENGLISH-SPEAKING AND FRENCH-SPEAKING ADULT

L2 LEARNERS AND NATIVE SPEGKERS

Rabia Redouane, Ph.D. of Philosophy, 200 1

Department of Cuniculum, Teaching & Leaming University of Toronto

ABSTRACT

This study investigates the use of word fomtion processes in Modem Standard

Arabic (MSA) by adult L2 leamers of mainly English-speaking background.

Forty-four L2 leamas, and 40 native speakers of Arabic took part in this snidy, and

perfonned three tasks (a production task and two comprehension tasks). L2 learners dso

pdormed a vocabuiary kaowledge test.

On the production task significant differences were revealed between L2 leamers'

and native speakers' innovations, with the L2 leamers producing fewer innovations and

leaving mon blanks than native speakers. The leamers' innovations comprised varied

compound constructions, inappropriate derived pattenis dmtiag oîher notions, as well as

formaliy impossible patterns. These innovations indicate that c m though the leamen were

making productive use of the derivational system, they still lacked knowledge of patterns, the

structural relationships between forrns and meanings in MS4 as well as the texical

constraints in this language.

Significant differences were also found between L2 leamers' and native speakers' use

of MSA word formation processes, with L2 leamers exhibiting a higher preference for

compounding while native speakers prefmed derivational processes, mainly VinE+gem, and

affixation. L2 kamers also opted for Vinf., and Vinf +gem. to a lesser extent. However,

affiation was hardly ever chosen by L2 lemers. Similar orders of preference were found in

both comprehnsion tasks. Certain acquisitionai principles (Clark, 1980a) were proposed to

account for the prefefences of the U h m : the principle of semantic transparency takes

fust priocïty; second in importance were the principles of productivity and conventionality.

An L1 effect could dso be discemed in some of the choices of ward fonnation processes.

Leamers at higher relative levels of vocabulary knowledge made significantly more

use of word formation processes than lower level learwrs to coin new meanings, suggesting

that lexical knowIedge in the targct language influenced L2 leamers' use of word formation

processes. However, the behaviour patterns exhibitcd by the higher IeveI L2 temen in

chooshg word k m t i o n p r o c e s a did not resemble those of native speakers.

The findings ofthis study provide d d e d information on the L2 leamers' use of

MSA word formation proccsses, ad insights relevant to the tcaching of MSA

3.1.1 Principles in the Acquisiticn of Word Formation Processes in LI Studies ..... . . ....................... ................*........ a) Productiwty ....... ........................................................ b) Semantic Transparency

............................................................... c) Formai Simplicity d) Conventionality. ................................................................. . . ..................................................................... 3.1.2 Ernpincal Evidence

................................................................................ 3.1.3 Summary ............................................. 3.2 Word Formation in Second Language (L2)

3.2.1 Piinciples of Acquisition of Word Formation Processes in L2 Studies .......... 3.2.2 Summary ................................................................................. 3.2.3 Tramfer as an Added Rinciple .......................................................

.................................................. 3.3 Research on the Aquisition of MSA ...................................................................... 3.3.1 Classical Studies

............. 3.3.2 Studies on the Acquisition of Arabic Word Formation Frocesses 3.4 Predictions about the Order ofMerence of Word Formation Processes

........................ to Coin Agency, Instnimentaiity, Location, and Causativity

4.1 Research Questions and Hypotheses ..................................................... . * ................................................................................. 4.2 Parttcipants ............................................................................... 4.3 Insuuments

........................................................ 4.3.1 Vacabulary Knowledge Test ....................................................... 4.3.2 T'hm Word Fomtion Tasks

................................................................... 4.3.2.1 Production Task ...................................................... 4.3.2.2 Two Comprehension Tasks ....................................................... 4.3.2.3 Background Questionnaire

.................................................................................. 4.4 Procedure ............................................................................ 4.4.1 Pilot testhg

..................................................... 4.5 Data Collection of the Main Study ............................................................................ 4.5.1 Task Order

................................................................................. 4.5.2 Scoring 4.5.3 Codingthe Data ......................................................................

.............................................................................. 4.6 Data Analysis 4.6.1 Quantitative Analyses ..................................................................

.................................................................. 4-62 Qunlitative Analyses

.................................................................... 5.1 Statisticai Procedures 5.2 Cornparison of L2 Leamers' and Native Speakers' Responses on

..................................................................... the ProQction Task ............................................................................. 5.2.1 Summery

......................................................... 5.3 of Word Formation P~OCCSS~S 5.3.1 Cornparison ofL2 leamen' and Native Speakers' Use of Word

.................................... Formation Proceses on the Production Task

5.3.2 Summaxy .............................................................................. 5.3.3 Comparison of L2 Leaniers' and Native Speakers7 Choice of Word

Formation Options on Comprehension Task 1 (Inventeci Items) and Comprqhension Task iI (Real Items) ..........................................

.............................................................................. 5.3.4 Swnmary 5.4 Cornparison between Lower, Mid, and Higher Level Learners' Use of

Word Formation Processes on the Production Task ................................. .............................................................................. 5.4.1 Summary

5.5 Comparison of iiigher Level, Lower level L2 Learnezri7 and Native Speakers' Use and Choice of Word Formation Processes ......................

5.5.1 Comparison of Higher Level, Lower Level Leamers' and Native Speakers' Use of Word Formation Processes on the Production

................................................................................ Task 5.5.2 Comparison of Higher and Lower Level L2 Leamers' and Native

Speakers' Choice of Word Formation Pptions on Comprehension Tasks 1 (Invented Items) and 11 (Real Items) .............................

5.5.3 Comparison of Higher, Lower Level L2 Leamers' and Native Speakers' Choice of Options in Comprehension Task 11 (Real Items) .........................................................................

5.6 Summary of the Quantitative Findings .............................................

6.1 Qualitative Examination of L2 Leamers' and Native Speakers' Innovations in the Production Task ...............................................................

6.1.1 L2 Leamers' and Native Speakers' Innovations ................................ 6.1.2 Types of innovations praduced by L2 lemers and native speakers .........

a4 innovative compound constructions for nominal notions ............ bl Innovative pmipbrastic constructions for causatives ...................

................................................ J Innavative derived pan enis ............................................................................ 6.1.3 Summary

6.2 Acquisitional Principles in L2 Learners Use and Choice of MSA Word Formation Processes ..............................................................

......................................................... 6.2.1 Rinciple of Producîivity ........................................... 6.2.2 Principle of Semantic Transparency

................................................. 6.2.3 Principle of Formai Simplicity ................................................... 6.2-4 Principle of Conventionality

6.2.5 Summasy... ........................................................................ 6.3 LI Effécts on the L2 Use of Word Formation Processes .......................

............... 6.3.1 L1 Effects on the Word Order of Compound Constructions 6.3.2 Suamacy ...........................................................................

CHAPTER 7: DISCUSSION AND CONCLUSIONS ................ ... ......... .......................................... 7-1 Summary and Discussion of the Findmgs

............................................... 7.1- 1 On the Production of Innovations.

Table 4 5 Participants in the pilot study. .......................................... 71

Table 4:6

Table 4:7

Table 5:l

Table 512

Table 5:3

Table 5:4

Table 5:s

Table 516

Table 517

Table 5:8

Table 5 9

Table 5:10

... Learners' scores on the vocabulary knwoledge test (max.=44).

AItematives provided in comprehension tasks 1 and II for each semantic notion. ..........................................................

Overall comparison between L2 learnen' and native speakers' ........................ responses on the production task (max.=24).

1 tests comparing L2 learners' and native speakers' responses for each semantic notion on the production task ....................

Overall comparison between L2 learners' and native speakers' mean percentage use of word formation processes on the production task (m=24). .............................................

L2 learnen' and native speakers' mean percentage use of word formation processes for each semantic notion on the production task ...........................................................

Overall comparison between L2 learners' and native speakers' mean percentage choice of word fomtion options on comprehension task 1 (invented items) ( m = 3 0 ) ..................

L2 learners' and native speakers' overall mean percentage of chice of word formation options on comprehension task If (reai items) (max.=30) ..................................................

Mean percentage of L2 leamers' and native speakers' choice of word formation options for each semantic notion on comprehension task I (ivented items ). ..............................

T tests comparing L2 lemers' and native speakers' mean - perceatage choice of word formation options for each semantic notion on comprehension II (real items) ...............................

Lower, mid, and highec level leamers' overall use of word pmesses on the production task. ......................................

Means of lower, mid, and higher level leamers' use of word processes for each semantic notion on the production task... .......

Table 5: 1 1 Means of higher level lemers' and native speakers' use of word formation processes for each semantic notion on the production task .......................................................... 99

Table 5 : 12 Ttests comparing lower level leamers' and native speakers' use of word formation processes for each semantic notion on the production task .................................................. 100

Table 5: 13 Cornparison between higher lwel L2 learners' and native speakers' mean choice of word formation options for each semantic notion on wmprehension task 1 102

.......................................................... (invented items).

Table 5: 14 Means of lower level L2 leamers' and native speakers' choice of word formation options for each semantic notion on

............................... comprehension task 1 (inventecl items). 104

Table 5:15 Ttests comparing higher level L2 lemen' and native speakers' choice of word formation options for each semantic notion on

.................................... comprehension task II (ml items). 105

Table 5:16 Lower level L2 learners' and native speakers' choice of word formation options for each semantic notion on comprehension taskII(reaiitems) ....................................................... 107

TYPOLOGICAL CONVENTIONS USED IN THE TEXT

Lexical innovations in English and other languages: Bold

English gloss (translation of Arabic and foreign words and innovations): t I

Prefixes, suffixes, and lexical items in English and other languages: ltalrcs

Root consonant:

Gernination of the same consonant:

Root vowel:

Phonetic transcription: Long vowel: Geminated consonants are doubled, Eg.,

Capital C

Capital CC

Small v

MODEM STAMIAIID ARABIC PHONETIC SYMBOLS USED EN THE TEXT

CHAPTER 1

INTRODUCTIûN

1.1 Statemtnt of the Problem

The present study deals with the acquisition of word formation processes by Iearners

of a second language (L2). More specifically, it airns at investigating the use of word

formation processes by English- and French-speaking adult L2 learnen of Modern Standard

Arabic (henceforth MSA) as they coin lexical innovations. It also aims to find out how

relevant the aquisitional principles of productivity, semantic transparency, format

simplicity, conventionality (Clark, 1980a). and transfer are in accounting for these L2

leamers' use of MSA word formation processes.

1.2 Significance of the Study

Arabic is of interest to the study of the L2 acquisition of word formation processes

because it is a Semitic language with a typological structure markedly different fiom that of

the Indo-European languages on which mom previous first language (LI) and L2 lexical

research has been done (e.g., Clark, 1980a, 1980b, 19813 1981b, 1982; Clark & Hetch,

1982, Clark & Berman, 1984; Clark & Cohen, 1984; Olshtaia, 1987; Broeder, Extra, van

Hwt, Sîromqvist & ~oionmaa, 1988; Broeder, Extra & van Hout, 1989; Broeder, & Extra,

1991; Broeder, Extra, van Hout & Voionmaa, 1993; Broeder, Extra & van Hout, 1995). The

study focuses on M S 4 a modemized form of Classicai Arabic, and one of the three forms

of Arabic defined by grammariaas and language resemhersl, and on kmers who are LI

speakers of hdo-European languages.

1 have chosen word formation processes as the focal point for this mdy oa both

theoretical and ducational grounds. From a thearetical perspective, it has been claimed that

tearning a laquage, and reachg an overall linguistic cornpetence in that language,

necessitates learning both a lexicon of well-established wotds and a repertoire of word

formation processes through which the lexicon can be expanded (Clark et al., 2984, pS43;

Olshtain, 1987, p.28 1).

Leaming how words are formed in a language goes band-in-hand with learning the

specific composition of the entire vocabulary of that language (Clark et al., 1984, p.543).

L-ng word formation processes involves understanding the formation of various types of

lexical items (Jackson & Ze Amvela, 2000, p.69), leaniing to analyse them into their various

consùtuents, and to segment them into roots and patterns. It consists a h o f learning to

identie relationships between forms and meanhgs and the various ways in which lexical

items are disthguished in the lexicon (Clark & et ai., 1984, p.543). This meaas, for example,

ttiat tearning Arabic word formation processes requires leamhg how root consonants are

combined with digereut patterns to express specific meanings such as agency,

instnunemJity, and location in n m s , or causaîivity in verbs.

Aocording UI Che@ (1969, p.34). the oldest of ibese lhree fonns of Aiabic is iht Clasid Arabic of m a i i d timcs whïch is tk languagie & pmlslamic pœtq and tk Qur78a and which graciually became ~aftercodififationTbcsecaadmoljorcis%oyofArabicistbtModtniLiiuaryorSrandard ~comm0ntoallArabicspealruig~~~.ZnWnnmbnnitisured~f9rni;rl~aadforIitaary parposcs aad is inspired by thc C l a s i d Arabic of n&cvai iimk It bas singlar morpbdogy, grammar. and s y n t a n a l l h a i ~ i t h a s ~ c d a e w I ~ ~ ~ * ~ ~ l l ~ t h i i d - p e ~ ~ ~ i r a n y v a n a t i ~ ~ ~ a r d i s i h e s p o L e n i a n g i a % c ~ ~ c a ~ d a y - y ~

In addition, learning word formation processes involves learning to identitj those

patterns that are most common in the language concemed. This means that learners of

Arabic, for instance, will have to leam that the different options available for coining

content words are derivation (ixation, gernination, and afixation), compounding, and

conversion, and that denvation is the most productive, while compounding and conversion

are less huent. Specifically, they need to leam that a cornmon way of naming people with

a parthlar profession is by means of the patterns caccaC2, e-g., /xayyat/ 'tailor'. and that

its feminine coumerpan is CaCCWa, also used to denote instrument nouns. They will aIso

have to leam that the pattern CXiC (active participle indicating the doer of the action) is

f o n d fiom the triliteral (consisting of three consonants) verbal form CaCaCa, whereas the

active participle muCaCCiC is formeci fiom the second triliteral verbal f o n CaCCaCa.

Aquiring cornpetence in word formation processes facilitates mastery of the target

Iexicon by providing generalizations that diminish the need for memorization and for heavy

dependence on contextual dues (Adjemian, 1983). From a practid standpoint, this study

should provide teachers with helpfiil insights concenring theù students' preferences for

different MSA word formation processes, and with useful information about the strategies

they use to form new words. It will also explore ways in which these strategies may be

relevant to the teachiig ofword formation processes in MSA

A fiirther motivation for this study stems from the general paucity of research on

LllL2 aquisition of Arabic and on the lack of studies specifically dealing with the

acquisition of MSA word formation processes, h W , to my knowledge, no studies about

the acquisition of MSA word formation processes as LI or L2 have previously been

conducted.

1.3 Aims of the Study

The overall objectives of this study are threefotd: (1) to compare learners' and native

speakers' use of word formation processes to express new rneanings, and their judgement of

the appropriateness of specifrc MSA word formation processes for realizing such meanings;

(2) to determine learners' behaviour patterns in selecting word formation strategies at

different relative levels of vocabulary knowledge; and (3) to find out if the acquisitional

principles of productivity, semantic transparency, forma1 simpticity, conventionality, and

transfer play important roles in aquiring MSA word formation devices, as various studies

of L1L2 acquisition of other languages have dernonstrated (e.g., Clark, 1980% 1981b; Clark

et al., 1984, Bahat, 1986; Olshtain, 1987; Broeder et al., 199 1, 1993 and 1995).

Thc six questions that guide this study are formulateci on the basis of findings of

different L1 and L2 studies (see chapter 3 for review of the* studies). These questions are

as follows:

1. What arc the quantitative and qualitative diffmences between L2 iearners and

native Arabic speakers in production of innovations in MSA expressing notions such as

agency, instnunentality, location, and causativity?

2, Compared with native Arabic speakers, what MSA word formation process(es) is

(are) most commonly selected by L2 teamers when choies are provided for naming notions

of agency, instnimentality, location, and causativity?

3. Does relative L2 vocabulary knowledge rweal differeutial effects for the

production of MSA word formation processes?

4. How do L2 learners with a higher level of vocabulary knowledge compare with

native speakers in their production and choice of word formation processes in MSA?

5. Which principle(s) (productivity, semantic transparency, fomal simplicity,

conventionality) is (are) most revealing in predicting L2 leamers' use and choice of word

formation processes in MSA?

6. Does the native language of the L2 lemer (English or French) lead to dierential

use of word formation devices in MSA?

To address these questions, production and comprehension tasks were designed and

administered to L2 leamers at three levels of vocabulary knowledge and to native speakers

of MSA

1.4 Orgrniution of tbc Thesis

This thesis is organized as follows: 'ïhe present cbpter has outlined issues to be

investigated. Chapter 2 prescrits a description of differem types of word formation processes

in English French, and MSA Chapter 3 considers the thearctical background to the study,

outlining the various principlcs that have been proposed to account for the aqukitioa of

word formation proceses, and reviewing the retevant research on wotd formation proceses

in L1 and L2 acquisition, Chaptcr 4 eicplains the methodology wad in the present study,

including the participants, the instruments, and the p r o c e h for data collection and

analysis. Chapter 5 reports the q ~ ~ ~ ~ ~ ~ * t a t i v e analysis of the hdings, followed by a qualitative

analysis in chapter 6. Fiaally, chapter 7 discusses the findings, the iimitations of the present

study, and its implications for fiture research on word fomtion processes and for the

teaching of this aspect of vocabulary in L2 classroorns,

CHAPTER 2

WORD FORMATION PROCESSES IN ENGLISH, FRENCH AND MODERN STANDARD ARABIC (MSA)

The tenn "word formation" designates that aspect of morphalogy that deals with the

formation of lexicai items (Matthews, 1974, p -38). Languages resort to different word

formation processes3 to eonstruct new lexical items and to expand their lexicon. Given the

word formation processes for a particular language, certain ways of coining new words may

be preferred over &ers by native speakers (Dresser, 1981). For example, among the

p r o c e s oflem used in English and French are denvational processes (sufixation and

prefixation). in Arabic, by cornparison, the widely favoured process is the Smitic device of

combining a (rnostly triliterd) consonantal roat with vowels, andior adding limited prefixes

to form various nominal and verbal pattms.

In this chapter, 1 do not intend to provide an exhaustive survey of word formation

processes. Rather, I attempt to present briefly word formation processes conunon to Indo-

European languages with particular reference to English and French, and those typical in

MSA, a Semitic language, in order to note similarities and differences between these two

groups of languages.

Only word focmation pmœsses that involve formnig ncw words h m exkithg wo& are deait with hae. For orha types of pmœscs tbat fonn mw worQ that havt no relationship wïth any prcMously cristing wopd 5œ Jackson & Ze Aweia, 2000. p . 4 3 4

2.1 Word Formation Processes in English and French

AfExation, compounding, and conversion are considerd the most fiequent word

formation processes in English and French (Wise, 1997; Katarnba, 1994). ïhese processes

are first examined, followed by a brief description of other word formation processes.

Examples fiom different sources for English and French are included for illustration.

2.1.1 Affiution

Affixation4 is the process of ~~nstnidiag a new word out of an existing one by

appending an affix5 (either suffix or prefix) to the stem. For example, from the English stem

tmch, such words that are formed by adding a sufix are: touching, touchable, touchability,

tm~chstone; and by adding a prefuc are untot(chnbie, untaichability. Also fiom the French

stem subi, words that can be formed by suffixation are: mbl-c!r, sabi-erie, sableur, sableux,

sabl-ier, sabl-on, sabhnnsr, s a b i m i r x , sabl-omtier, and by prefixation: en-subi-er,

en-sabl-ement, dés-en-sabl-er, dés-en-sabkment (Fieisch, 1968, p.32).

In Engiish and French, derivationat prefixes do not change the word class of the

lemma although they do change its meaning. Derivational suffixes, on the other hand,

change the syntactic category or word class of the word to which they are added. Such

suffixes are disthguished witb respect to the grammatical ctass of the word they form. There

are nominal suffixes that fonn nouns fiom an adjectival or verbal base (e-g., fuimess, and

Twotypeofaflixcsexrist:deriuarionaifi*es,which bavethcfiiwtionofmatingnnv leucai items, and inaeaional af6xs, whicb carry gramoiaticai infomtion (Wise. 1997, p.103). Tbe derivati~llalafh,~ aredealtnithbere.

payment); verbal sufixes that derive verbs Eom nouns, other verbs or adjectives (e.g,,

regdate, charter, and activate); adjectival sufixes that fonn adjectives fiom ncminai,

adjectival or verbal bases (e.g., haditionai, nurrowish, and readable); and adverbial sutfixes

that constnict adverbs fiom either adjectives or nouns (e.g., busi&).

For detailed illustration of different types of sufixes and prefixes, see Bauer, 1983,

p.223; Katamba, 1994; and Jackson & Ze Amvela, 2000, for English; and Guilbert, 1975,

p. 170; LecIerc, 1989; and Wise, 1997, for French.

2.1.2 Compounding

Another process also cornmon in English and French is compounding, known as

composition in French, which is the juxtaposition of two or more independent words to form

one lexical unit (Fromkin, Rodman, Hultin and Logman, 1997, p.51). For example, the

English compound dishwasher and the French compound portefeuille are formed from two

independent words uïsh and washer, and porte and feuille respectively. Like derived words,

compounds can be classified in many categories in terms of the word class of the

constituent elements. G e n d l y in English, the word class of the last element of the

compound determines the category of the compound. Among the various types of compound

constructions that English possesses, notable are: noun compounds (N+N, V+N, Adj+N and

Adv+N), verb compounds @kW, V+V, Adj+V, and AdviV), adjective compounds (N+Adj,

Adj+Adj, AdeAdj), and adverb compounds such as (N+Adv, V+Adv, Adj+Adv, and

AdvtAdv). French also possesses a considerable number of compound conçtnictions, e-g.,

N+N, N+Adj, N+prep+N, Adj+N, Adj+Adj, Prep+N, Adv+PP, V+N, and V+V. Whatever

their hternal structure, campounds in French are generally considered nouns, except the

compounds Adj+Adj, Adv+PP, and V+N that hnction as adjectives (Wise, 1997, p.121).

Some of these syntagms are more productive than others, for example, N+Adj, N+N, and

V+N compounds are claimed ta be more productive in French (Leclerc, 1989). (For a

detailed description and examples of each category of English & French compound

constructions, see, e.g., Bauer, 1983; pp.202313, Katamba 1994; Jackson et al., 2000:

Leclerc, 1989; and Wise, 1997). in addition, cornpounds can be distinguisheâ with respect to

the word order principle of the compound. Two types of compounds exist: Head-initial

compounds where the main word is followed by a moditjnng noun, and head-final where the

main word is preceded by a modifLing noun. English is described as a head-final language

while French is head-initial.

2.1.3 Coavenion

Another process which is tiequently used in English and French to coin lexical items

is conversion or zero-denvation (Marchand, 1969; Adams, 1973) where the base is

reassigned to a new part of speech with no change in fom. The major kinds of conversion in

Engiish are nouns into verbs, e.g., a bridgceho bridge; verbs into nouns, e.g, 10 cail/a call;

and adjectives into verbq e.g., brownlto brown. In French, one type of conversion is

convertins adjectives to nouns, e.g., une voiture mtomobilelune automobile.

Besides derivation, compounding, and conversion, English and French make use of

other word formation processes. Among these are blending, clipping, acronyms,

b a ~ k f o ~ o n , and bomwing. BeIow is a brief description of these processes.

2.1.4 Blending

Blending is defined as the creation of a new word "fiom parts of two (or possibly

more) other words in such a way that there may be no transparent anaiysis into morphs"

(Bauer.1983, p.234). Blends are also known as "telescope" or "portemanteaui~ words

(Jackson et al., 2000, p.87). Some typical examples of English and French blends are smog

fiom fsmoRe+f~); teléton fiom (téléphone+marathon); and fianglais h m

(fiançcncns3ungIais).

2.13 Clipping

Clipping refers to the process by which a lemma of two or more syllables is

shortened without changing its fonn class and tùnction (Katarnba, 1994, p. 180). Among

commonly-used English and French clipped forms are ad fiorn udwrtisement, flu h m

influenza, and télé for télévision.

2.1.6 Acronyms

Acronyms refer to the process of coining words fiom the initial letters of the words

of a phrase (Adams, 1973, p. 136). Most acronyms have been constructeci as short names for

national organizations such as in English liMCEF for United Nations Intemtional

Chifdren's Emergenc~ Fund, and in French la CEE for la Communouté économique

européenne.

2.1.7 Bacblormrtion

Backformation consists of fonning new words fiom existing words by removing

&es h m a base. For example, the English verbs p e d e and Cvpewrite were formed fiom

the nouns pedlar, and fypuriter (Marcband, 1969): and in French the adjective aristocrate

from the noun uristucratie.

2.1.8 Borrowing

Finally, a word formation process that both English and French resort to is

borrowing lexical items fiom other languages. Two types 04 borrowing exist: direct

borrowing which involves borrowing a word directly from another language, e.g., the

French word omelette into English; and indirect borrowing: that is, when a word is p a s d

indirectly fiom one Ianguage to another by undergoing phonological modification, e.g.,

Engiish coffee is originally firom hhveh after having gone through Arabic Rahva and Dutch

kofle.

2.1.9 Summ y

This examination of English and French word fonnation processes reveals that both

languages similarly resort to various processes to coin new words, but afExation

(comptising prehation and suffixation), compounding, and conversion are the most

îlequent ones. It also reveals that u d i e prefixation, suffixation changes the class of the

word. The two languages, on the other hand, dEer with respect to the order of words in

cornpound cons~~ctions. Engiish is a language with a head-final word order principle while

French is a language with a head-initial word order.

2.2 Word Formation Processes in MSA

h this MSA word formation processes will be d e s c r i i in order to point out

any similarities and/or differences between MSA Engiish and French, Most grammanans

daim that Arabie possesses only three word formation processes: ishtiqaq (derivation). naht

(compounding), and tdrib (b~rrowin~)~. Others, on the other hand, argue that other

processes exist in Arabic such as conversion (Reguigui, 1994). and the process of flih*?

ghanb ai-lughal 'revival of uncornmon language' or Ristinbàtl 'extraction' (Harnzaoui, 1975;

Reguigui, 1994). In the following paragraphs, these processes are examine4 and their

related aspects and foms are presented. Examples from different sources for MSA are

included for illustration.

2.2.1 Ishtiqaq @trivation)

One of the most distinguishing features of Arabic derivational morphology and other

Semitic languages is its system of consonantai roots7 and denved patterns. By drawing on

existing Arabic verbal rootsa, an immense array of new words are f o d through ishtiqQ.

This process is considered the most natutal method for lexical innovation and expansion of

the texicon, and is central to Arabic grammatical structure (Stetkevych, 1970, p.7; Badry,

1983). According to El-KhafaTtT (l98S), ishtiqaiq is the method:

6 "Tm s'accwdcm A raonnaim que, dans la niarphalogic (la sémantique mte d'autres voies), l'arabe dîspose seuIancnt de mis pmdd& de base, pour la constitiitioa de son vombuiairt d r n t : la dérivatim éiymdogigue (ishsiqaS,, la composition (quaii6k au xns iargc dc naht) et de l'cm de termes éaangus par arabidon (tafrb). La smiehirc de k laague n'aiitorise pas dlnarw pmM&" (Monteil, 1960, p 106).

Derivation in Arabic also uses mo haî have nominal aire meanings. In MSA lhere is d n c e tbat new words an k daived h m nominal mots. For exmpk Iht instnmmt mm ImimlaiW 'saitshaka', ad the v d ItaJhammasPl 'bc mbathd arc daived h m ihe mts h-l-hi 'sait' and 1 s h . d 'sun' rqte&cIy which have a basic mrnimi cote me*

to be preferred because, more than any other methad. it adhmes to the naturd cbannels and character of Arabic, and thus the produas of derivation are less likely to offend or repel the native user. (pp. 181 -1 82)

Arabic grammarians distinguish between three types of ishtiqaq: The tisst me,

ishtiqh ai-kabir knom also as .al-qaIb (metathesis), involves switching the order of the

consonantal mot. The second Spe is ishtiqàq al-&bar or al-ibdal, (root modification),

which consists of changing one consonant of the mot. These two types of ishtiqaq are less

used than a third type, and are considerd "marginal methods", and not tme forms of

derivation Badry, 1983, p. 16, Tarazi, 1967, p. 102, El-Khafaifl, 1985, p.72). The third type,

al-ishtiqaq a 1 - W (general), is argued to be the moa prevalent and productive process, and

"remains the charactcristic method of word-creation in Ar&icff (El-KhafSfl, 1985, p.70).

M y this third type of ishtiqiq will be discussed below. For the purpose of this study,

derivation, the commonly accepted English equivdent for ishtiqm will be used to refer to

al-ishtiqsq al-%m.

Derivation in Arabic consists of the combination of merai consonantal mots with

vowel infixes and a x e s to t o m verbai and nominal patterns. Denvation comprises three

processes: 1) Vocalic infixation aione, which invalves combining the consonantai root with

a pattern of inserted vowels; 2)Vocalic infixation and semination, wiiich consists of

i n d n g vowels in the root and doobling a mot consonant, espedly the 2" consonant; 3)

Anmation, which involva the appending of an a&, usually a prefix, to the stem (a

consonantal roor wiîh vowel infixes) (Badry, 1983, p. 17). For example, fiom the underlying

consonantai root A-I-mJ l em' , among the noms, adjectives, and verbs thaî can be fonned

by vocaîic infixation dont are: /Tâiim/ 'scholat, /%lm/ Icnowiedge', fialaml 'world'; by

vocalic infixation and gemination of the 2nd consonant are /Fallamal 'to teach'; and by

affixation are /?aFlama/ 'to inform', /rnuîalliml 'instnictof, and /mutaTallim/

'knowledgeaMe' , etc. Al1 these words antain the same root, and share the same meaning

Multiple derivation h m one root is one af the characteristics of Semitic languages

in generd, and of Arabic in particular. in Arabic, every root typically has a similar potential

for derivation in the following paragraphs. nominal and verbal derivation are descn'bed with

a particultu emphasis on the patterns investigated in this study that denote the notions of

agency, instnimentatity, location, and causativity.

A wide spectnim of simple nominal patterns in MSA are formed through the

association OF specific consonantal routs with a range of vowel infixes. According to Fück

(1951), MSA nominal patterns range from the simplest forms CvCC, increased to CvCvCv,

and fùrther augmented to other patterns fonned by doubling of the 2" consonant. In MSA,

there are many more nominal patterns than verbal patterns. According to Stetkevych (1970):

Considering the Arabic synem of word derivation as a whole, it becomes clear that the possibilities of noun derivaiion are much more numerous and divwsified than those of verbal derivation At least theoreticaily, the verbal derivation is limitai ta the standard fiftecn forms-always rnaintaining the premise of a buic verbal root as the initial point of any verbai derivative. Thus, within the purely theoreticai possibilities of derivation h m any triIiteral rwt, an extrernely smafl percentage of derived words wwld be verbs, with the rest falling imo the broad categary of the Arabic n o m . In pradce, however, verbal derivations may constiîute between 10 to 25 percent of a given rm. (p. 10)

However, nominal patterns are not as regular as the verbal patterns, and because of

their irregularity "only a few arc productive and 6equent enough to allow any productive

use" (Badry, 1983, p. 18). Among the most &quent and productive nominal pattenu in

MSA are verbal nouns (nouns indicating the action of the verb and fundonhg as regular

nouns), active participles, agent nouns, and inmument noms. For a more elaborated

description of most nominal patterns, refer to Wright (1964). Holes (199S), and Fischer

(1997). Table 2: 1 illustrates exampbs of some ofthe most fiequent patterns fonned througit

vocalic infixation alone and/or gemination of the 2* consonant.

Tabk 2:1

Frquent nominal patterns in MSA derived h m verbs by vocalic infixation alone, andlor gemination of the 2"' consonant

1 ~attems 1 Arabie Word 1 Engiish Tnnilation ( I

Vk~baENooir , CaCC 'killing'

1 I

IiutriilacitRoirir 1 CàCiC I /&ahin/ 1 'charger'

t

CaCCK 1 kali&/ 1 'mixer, biender'

A further group of Arabie nominal patterns is fomed h m tnliteral consonantal

roots through extemai afixation where either derivational suffixes or prefixes are appended

to the stem Derivational prefixes include /ta/, ltd, /ma/, /mi/ and :mu/, and derivational

suffixes are Ra/, /a/, /an/, ri/, /iya/, and Mt/. Only prefixes /ma/, /mi/, /mu/ are discussed

here as they denote agents, instruments and locatives, the three nominal notions investigated

in this study.

The prefix /ma/ aîîached to the stem Ieads to the nominal patterns maCCaC,

maCCaCa, and maCCiC indicating locative nouns.

The prefix /mi/ appended to the stem yields the nominal patterns miCCaC, miCC&,

miCCaCa denoting instrument nouns.

The prefix /mu/ attached to stem yields the nominal pattern muCaCCiC denoting

agency as well as instnimentality. This prefix also yields the following patterns muCàCiC,

muCCiC, and mustaCCiC and theu ferninine counterparts, which denote instrument noms.

These patterns are active participles of the derived verb fonns. Table 2:Z presents examples

of these prefixed nominal patterns dmoting agency, instrumentality, and location.

Table 2:2

Agent, instrument, and locative patterns in MSA hy affixation

( /mifiari! 1 'key' 1

Similar to nouns, verbs in MSA are cmstruded out of a triliteral consonantal verbal

mot. The verb roat pattern CCC becornes the first basic verbal forrn through the interposing

of short voweIs (a, i and u) betwem its consonants, for example, katabal 'to wite', lshanbal

'to dnnk, and Mura/ 'to p w big'. This first verbaI fonn is claimed to be "the most

Wently used verbal pattern in Semitic languages and is usually refend to as the basic

pattern in traditionai grammars" ( B W , 1983. P. 101.

Derived verbs are also constructed fiom the triiiteral verbs (El-Tikaina, 1982). The

first verbal form CvCvCv is modified either by tengthening vowels or consonants or by

adding an afix, usually a prefix, to form other derived verbal patterns. MSA has fourteen

derived verbs that express several semantic notions such as causativity, reflexivity,

passiveness, inchoativeness, and reciprocity (Fleisch, 1%8, p. 112). Only ten are in common

use. Table 2:3 presents the ten commonest verbal patterns in MSA and their meanings.

Table 2:3

Verbal patterns in MSA

Verbal Patterns Meaning Exampies

CaCaCa katabal 'to write'

PH

PIll

C ~ C C ~ C ~ '

CXcaCa I

PV

l 1 I

Pm 1 ?inCaCaCa 1 reflexive 1 ninqqafal 'to be or become nit off

causative

causative PIV

PVI

Jfahhamd 'to make someone understand'

reciprocal

ff aqfadd 'to make someone sit' lacCaCa I

taCaCCaCa reflexive

1 L 1

PIX 1 7iCCaCCa ( inchoative 1 R n m m d 'to be or to becorne red'

kâtabat 'to write to each other'

/takassard 'to break to pieces'

taCaCaCa

Pm

Y In addition IO its use of denoriag caiISBtiVt mcaniiis, this trived form is alsa employed lo express intcnsïvc Tbe intenriveness is its major mng. (El-TiLaii9, 1% p. 12)

reciprocai

7iCtaCaCa

PX

_I

Itawafaqd 'to come to an agreement'

reflexive

listacCaCa

/?iktatabaI 'to be registered'

1 estimative eductive

Ristahsanal 'to consider good' /?i~t~ghfara/ 'toaskforpardon'

In practice, however, the maximum number of verbal foms found in any root is

seven or eight, and most roots show fewer than this number. In the Orientakt tradition,

verbal fonns are usually referred to by the letter P for "pattern" followed by a Roman

numerai. (e.g., PI, PU and so an).

2.2.2 Summary

This brief descriptive acwunt of derivation in MSA reveals that dris process is one

of the features that distinguishes MSA firom French and English To form words, MSA uses

a root which is compose. oniy of consonants, and which carries a general idea. This root

takes different bodies by either vocalic aitemation, vocalic andfor consonantal gemination,

or by adding monosyllabic affixes, mostly prefixes. English and French, however, use a

radical cornposed of consonant and vowels which is lengîbened by prefixes andfor suffixes.

But derivation is by no means the only process that distinguishes MSA fiom English and

French. Nah is a n o k process.

2.2.3 Naht (Compouadiag)

Modem Arab phiiologistq grammarians and Iinguists, as weH as Arabists, have often

referred to the pmcess of naht in the general sense of wmpounding as known in indo-

European languagcs. A close look at this word formation process, particularly in Classical

Arabic wofks, show thai it corresponds closely to blending (e.g,, words like smog in

English). In this section, an analysis of naht as treated in classical Arabic philology and in

Modem Arabic linguistics will be presented. Its appticability and. its Iimits will also be

exarnined.

The tenn naht is derivecl fiom the triliteral verbal root ln-b-t/ which means 'to

chisel' or 'to came in a hard materiai' (El-Mouloudi, 1986). in morphoiogy, this wxd is

used by the early Arab and non-Arab philologists and grammarians to mean the formation of

one word out of two or more words as a kind of abbreviation (El-KhafaStl, 1985). ibn Faris,

one of the old exponents of dt, wnsidered that in the Arabic lexicon the rnajority of

quadrilitd and quinquititeral words (those that have either four or five consonants in their

roots) are the result of this process. Ai-Maghribi (1947, p.15) alsa suppons this daim by

saying that quadriliteral and quinquiliteral words in Arabic are the result of blending pairs of

triliterai verbs such as lhanvald 'to walk fast' fonned tiom harabal 'to flee, escape' and

/walla/ 'to nin away'. In addition, Fleisch (1968,p. 124) States that naht for Arabs means to

extract, ftom a phrase or sentence, four consonants deemed characteristic of the rneaning of

the constituents in order to form a quadriliteral verb. For example, /basmalal is forrned out

of the sentence ibismil 1-Lah ar-ra?idn ar-rahahim/ 'in the name of God the most mercifbi the

most compassionate' .

Four types of naht (blending) are distinguished by Al-Maghribi (1 947, pp. 13- 14):

al-naht al-fini (verbal naht) involves the creation of a verb, usuaily consisting of more than three consonants h m elements taken fiom words in a phrase or siion sentence as in /sam?ala/ 'to gret' formed fiom Iqala assal&mu Talaykum/ l i t d l y (he said) 'Peace be upon you'.

al-naht al-wa@ (adjectival naht) consists in blending two words to mate an adjective, e.g., dibap 'firm, physically strongt fiom /Nata/ 'to hold fast' and /dabaral 'to leapl-

ai-naht al-ismi (nominal n a ) is the formation of a noun blend fiom elements of two words, e.g, /julmÜd 'a large rack' cnared âom Ijaludd 'to beeome hardt and /jarnuda/ 'to wngeal'.

al-nah nisbi, (relationai a&) denotes the relation of something or somebody to a place, tribe, school of thought, etc, as in fiabdari/ (sotneone who is afiliated to the tribe TAbd ad- du).

Thus, naht as traditionally used in Classical works corresponds to blending in Xndo-

European languages. But, in Modem Arabic works, the usage of the t m naht has expaaded

to include the process of compounding as known in indo-European languages, that is the

juxtaposition of two or more words. Two tenns are used by Modern Arab linguists to

distinguish between blending and compounding processes in Arabic: Tarkib nahti for the

process of blending, and tarkib majd for compounding. The originality of t a r a majzi as

weii as its applicability in MSA has been a controversial issue in Modem Arabic linguistics,

and both Arab and non-Arab scholats have expressed contrasting opinions. Some consider it

an unreiiable process of Arabic word formation, and believe it to be essentially a marginal

aspect of Arabic (al-Kmali, 1932, Mus- Jawad, 1955, and Fleisch, 1968). Importantiy,

they consider Arabic to be in essence a derivational laquage. Fleisch (1968) sunimarizes

this point of view by stating that:

... aucun procédé normal de composition ne s'est établi dans la langue. L'arabe ne peut pas réunir deux ou plusieurs mots joints suivant les régles syntaxiques ordinaires pour en former un seul, ainsi que fait le fiançais; habqurr - "grêle"(habbu qwin "grain de fioid") est proche de la manière française, mais reste un très rare exemple. L'arabe ne peut réunir deux mots par une voyelle hématique comme font le latin et le grec, ni les joindre selon les composés de l'anglais ou de t'allemand. La composition n'est pas dans son génie. Ceci est un grand handicap dans la constitution d'un vocabulaire technique scientifique. (p. 124)

However, others (e-g., Al-Kawakibi, 1959; Al-Hud, 1958) have strongly wpported

the use of naht as a way of expanding the Arabic lexicon adequate for modern needs. For

example, Al-Kawahii (1959) considered naht most operative especially in translating the

vocabulary of a science such as chernistry (p.8). According to Al-Husad (1958). it is

imperative to resort to naht ta coin new words and concepts in science and technology

because the Arabic derivational system is restricted to a Iimited number of patterns and

paradigms, and because it lacks the ability to evolve, abne, the new vocabulary that is

needed to express the ever-expanding domain of human thought (Mustafa Ja* 1955,

in the current usage of Arabic, naht is not limited to science and technology, its need

and useftlness are more general. A considerabIe number of compound constructions are

found in modern usage in different dornains. Among these compound consuuctions, notable

is the idafa compound (genitive conmct or ~-in-construa)'~ that combines two nouns. The

most common are: N+N, as in Imiqyk, al-ish%V (measure + radiation) = 'radiometer', and

the AP+N construction, where the first noun is the active participle of either pattern CaCiC

or muCaCCiC, as in Idsim 7ai)abdabatl (rnarker+oscillation)= 'oscillograph'; and

/rnujassim assuwad (amplifiertpictures)= 'stereoscope'. The N-in-constnict cornpouad is

also used to denote agent, instrument, and locative nouns where the nouns /rajuU 'man',

iï&d 'machine', or /mak&n/ 'place' are combined with another noun; and the AP+N

constnrction for agent and instrument nouns (see Table 2:4 for examples).

Another type of compound is N-tM combining a noun and a modifier. This type of

compound is oonsidmd a oneword unit" (Wehr, 1943). and it is oped for in MSA whm

translating European technical terms that requue two referential components (Bateson,

1967), for example, lra7s maliyl (head + money+iy)= 'capital'.

Moateil Fmich &adalion of tians Webr (1943): peut observer, daos te bagage offiael, uac napcname mcObCrrmc & coasmrtion: I'cmboitage du qmlitaiif(ou du dCtaminaa) avec Ie nibsian9f qu'il-dgit-L'ensanbk &ant senti comm mi mil mot" Manteil, 1960,p.230).

An additional type of compound construct.ion consists of combining two nouns that

are co~ected by a preposition (N+prep+N): for example, Iqabilun li ssukna/ (acceptable

to+ habitation)= 'habitable'. This type of compowd is also used in MSA to denote

instruments (see Table 2:4 for examples).

Table 2:4

Compound constructions denoting agent, instrument and locative nouns in MSA

1 N+N campsund constructions 1

ktan+ business' 'business man'

Ralatu lliaddl 'machine+ harvesting' 'harvesting machine' (instrument)

/makanu IghasW 'place+washing' 'washing place' (locative)

APçN'(a«id'rn

tkatibun ~umümiyl 'writertpublic' 'public writer' (agent)

lN$a?atu h d q l 'extinguisher+fire ' 'fire-extinguisher' (instrument)

1 /îalatunliYasri lfakihd 'machinetfor+squeezin~juice'

I 'juicer' (instrument)

A mher type of compound in Modem Arabic is one that combines a prefixed

particle and a noun. The p& is mosrly a negatiw particle such as /Ia/ hot' 'non', Ighayrl.

Kadam/ 'without', ldiddal 'againstl- These negative particles constitute the fïrst component of

the compound and the second component is either a noun or an adjective (Drozdik, 1967,

Monteil, 1960). Examples of these compound foms are: na samV 'anti-semitic', ighayr

q&üniyl 'illegai', ididda jasiisiya/ 'wuntempionage'.

Except for this last cornpound construction, the word order principle in ail types of

compounds described above is head-initial.

2.2.4 Summary

The above analysis of naht has shown that this process has been treated as a type of

blending or abbreviation in Classical Arabic philology, in modem Arabic linguistics,

however, it has been wnsidered either as an accidental linguistic phenornenon used only in

cases of neeessity, or as a productive word formation process including blending and

compounding. The fact that naht involves both blending and compounding is one of the

differences thai distinguish Mbt in MSA fiom compounding in English and French. Another

noteworthy difference is the fact that compounding in MSA is in general lirnited to N+N

combinations, contrary to English and French where compounding wvers different

categones depending on the class of the first noun. A further distinction among the three

languages is the principal word order of compounds. Unlike English, MSA is a head initial

language, a word ordn principle that it shates, however, with French Also, Arabic, like

French, combiies nouns in a head-initiai principle by comecting head and modiwng noun

with a preposition. But this type of N+prep+N compound is lirnited in Arabic to instruments-

2.2.5 TaYrib (Borrowing)

Another word formation process in MSA is lexical borrowing12 which involves the

incorporation of foreign words into the laquage. The view of lexical bomwing among

word formation processes in Arabic differs fiom one author to uiother. Some treat it as a

topic unrelated to other word formation processes in Arabic (Drozdik, 1979, p.23), whereas

others judge it as a word formation device, but one of last reson. Hamuioui (1975) advances

... est conçu comme un procédé auquel la langue doit avoir recours aprés avoir epuisé tous les autres procédés, a savoir la résurgence et al-Tshtiqaq ... Néanmoins son emploi ne doit pas porter atteinte aux phénomènes et aux schèmes de la langue qui doivent être protégés dans la mesure du possible. Nous en déduisons que si les termes empruntés constituent des corps étrangers à la langue, voire des intrus, il n'en demeure pas moins qu'ils doivent &e soumis à des nonnes établies par les grammairiens classiques. (p.361)

MSA resorrs at times to ta7n3, mcularly in cases where there is no quivalent

Arabic word, and where an indigenous Arabic translation might require a phrase or

cumbersome explmation. MSA has a number of bomwings âom Engiish and French.

Among these are Ifalsafa/ 'philosophy', hnisilinf 'penicillin', /u@bis/ 'buses', /fayt&nïn/

'vitamin', ibilijl 'plage' and /&ma/ 'cinéma'. Borrowed words can be assimilated and

incorporateci into the Arabic lexicon if they can be made to fit one of its derivational

patterns. For example, the bomwed noun 'telephone' has in MSA, an equivalent noun

/tiiifÜn/ fiom which the quadriIat«al root /t-1-f-rd has been abstracted to produce the verb

Italfad 'he telephoned'.

2.2.6 Convtnion

Besides derivation, compounding, and tayrib, Reguigui (1994) argues that MSA

resons ta conversion. The only time that MSA uses conversion is when adjectives are

converted into nouns. According to Reguigui, nominalization of the adjective is a process

becoming more fiequent in Arabic, and is a source of numerous neologisrns (p,72). The

patterns that are most affected by conversion are active and passive participles. Among the

examples provided by Reguigui: Imuhawwil taraddud ?iliktrüaiyy/ 'electronic fiequency

converter'; and /murakicab manif li takawwun al qushiirl 'antiscale compound'.

2.2.7 lihya? gbrrïb rClugba (Revival of uncommon hnguagt)

The final source of word formation in Arabic is lihya? gharib al-lugha. This process

consists of unearthg and reviving archaic words and assigning them to new and modern

semantic tùnctions, usually related to their old ones (Hainmoud, 1982, p.139; Reguigui, 1994,

p.83). Examples of this resurgence include the word /qi@r/ 'caravan of camels' which is

assigned a new meaning in modem usage, 'train'; and the word /qa@t/ 'head of wavan of

camels' which is used in MSA to mean 'locomotive', 'engine'.

2.2.8 Summr y

In tûis descriptive account of word fonnation processes in English French, and

MSA, 1 have pointed out that in English and French, affixation, compounding, and

conversion arc the most productive word fonnation processes. MSA, on the other haad has

two major devices, ishtiqaq ?al-f&n (decivation), and naht. The former is t6e most

productive word formation process, and the latter, refming in Arabic to both btending and

compounding, is less productive. Although MSA relies heavily on these two major devices,

it is nevertheless a more varied and complex system in cornparison to indo-European

laquages. This cornplex@ is illustrated by the wide range of options that Arabic offers to

express a single semantic notion. To dernonstrate this variety and complexity, Table 2 5

presents the different options existing in MSA to express instrumentality.

Table 2:s

The different options in MSA to denote instnimentality

I ion Vocalic infixation alone

CâCiC

Derivat & 1 '

- 2

* _ 3

Compounding

càcüc Vocalic lnfixation & gemination of the 2* consanant

CaCCâC

1 .

CaCCaCa f ixat ion (prefixation)

miCCaC miCCaC miCCaCa muCaCCiC muCaCCiCa rnuCâCiC muCaCiCa muCCiC muCCiCa mustaCCiC

N-in4onstnict f f ala 'machine'+nounl

2. 1 AP+N f ' ' N+prep+N

/?ala+li+N/

CHAPTER 3

THE ACQUISITION OF WURD FORMATION PROCESSES: THEORETICAL FRAMEWORK AND EMPIIUCAL

EVIDENCE

This chapter outlines the various pnnciples that have been proposai to account for

the acquisition of word formation processes, and reviews the relevant research on word

formation in LI and L2 acquisition,

When learners need a word to express a specific meaning, they search tbeir lexicon

for an appropriate lexical item- But how do learnm go about finding an appropriate word to

convey a rneaning for which they tack a well-established item in theù lexical resources? Io

the absence of a conventional word to convey a concept or meaning, Clark (198 14 1982)

claimed that lexical innovation h m the available lexical stock is a common strategy that

leamers rely on. Lexical innovation which involves forming new lexical items is a common

phenornenon in the speech of both chiltiren and adults: novel words are created to serve the

function of filling lexical gaps in th& lexicon (Clark et al., 198 l a , pp.3ûû-30 1).

To coin new wordq kamers rely on severai word formation processes, the most

common of which are derivation and compounding. But which process(es) do leamers

chwse at an eariy stage of acquisition? Why do they pick certain processes over othtrs?

The word formation options that leamers use to form lexical coinages are said to be

constrained by formal restrictions in laquage, and influenced by certain pnnciples of

acquisition (Clark, 19804 198 la; Clark et ai., 1982; and CI& et al., 1984). In the following

pmgraphs, the general principtes that are argued to account hr the choice of word

formation processes during the cowse of acquisition are outiined and examineci.

3.1 Word Formation in Fint laquage (LI)

3.1.1 Principks in the Acquisition of Word Formation Pniccsses in L1 Studiu

Research on the gradua1 mastery of word formation processes by L1 leamers (e.g.,

Badry, 1983; Berman et al., 1982; Clark, 1980% 1980b, 1981a, 1981b; Clark et al., 1982,

1984) suggests that four general principles govem the course of acquisition of word

formation pracesses. Tbese are productivity, semantic trsnsparency, formai simplicity, and

conventionality. Evidence from data based on both observation of children's spontaneous

speech and elicitation techniques is presented in section 3.1.2 to support the claim that these

four fadors play an important mle in aquinng word formation processes.

a) Productivity

It is claimed by Clark (1981a) that when Li learners acquire new word formation

devices, they rely on the productivity of the devices in the laquage. Roductivity in this

context refen to "the degree to which a particular word formation device may be used as a

mode1 for new lexical items"(Clark, 1981% p.3 12). Word formation devices that are more

productive are more available to learners, and should therefiore be acquired earlier and used

in preference to less productive devices.

Based on this principle, leamers will deveiop a stratcgy that will help them to look

for and identifL the mst commoniy used word formation process, then incorporate it in their

repertoues and adopt it for crcating innovations (Clark, 198la; Clark. 1982, p.5). For

instance, in English, iî is predicted that children will aquire the suffix +r before -ist and - imt to express agcncy because -4r is the most productive of the thcet. Clark (I981b) argues

that this prediction can be tested "both within and across i-es by examining sets of

word formation devices aud patterns closely related in meauing to sec whcther the most

productive ones are acquired eartier when meaning is heid (relatively) constant" (Clark,

198 1 b. p.254).

b) Semantic Transprrency

Clark also claimed that a word formation device that has a one-tmne mapping with

meaning is semantically transparent and should be easier to acquire than devices that have

multiple meanings (Clark, 1981a, p.3 13). This principle predicts that new words are more

easily formed fiom elements with independent satus which have conventional meanings

already known to leamers. According to this principle, compounding (involving elements

with independent status) will be the first device opted for to coin new words because it is

most readily accessible to leamers. For example, in English, the conventional meanings of

the words cut, man, and mhim make it easier to form compounds iike nit-man for

'someone who cuts things', or cut-machine for 'something that cuts things' catiter than the

word cutter with the suffix er . The reason behind this, according to Clark et ai., (I982), is

that the words man and muchine express the notions of agency and instninientaiity more

expiicitly than the sufix +r (Clark a al., 1982, p.4) which can appear in other words

without king a suffix, e.g., corner: h m e r .

Semantic transparency also predicts that new words are more easiIy coined with

elements that use a single fonn to express a single meaning. For example, in constnicting

new agents in English, sufkes such as -ist and -im should bc acquired before -er because

the su& -er expresses instrumental as well as agentive meaning (Clark, 1981b, p.3 13).

By relying on this principle, cbldren use two strategies. The 6rst strategy is ta look

for word formation processes mmprised of wods that mark only one meaning, and use them

jointly to coin new compound words The second strategy is to find a single process in the

repertoire and use it with only one meaning to coin the required word. For instance in

English, leamers may, in fact, first use the suffix -er to form new agent nouns ody. Once

they master the agency meaning of this suffi they can expand its usage to include 0 t h

meanings such as instrumentality, comparative, etc. (Clark et al., 1984, p.547).

c) Formal Simplicity

The third principle that Clark claimed to account for the acquisition of word

formation processes is formal simplicity. Simplicity is defined in this context as the degree

of change in a fom. The less a word-form has to change, the simpler the process is heId to

be (Clark, 1980% p.11). In application, this principle predicts that simpler processes are

easier to acquire than more complex ones. Guided by this principte, the l m e r strategy

entails choosing processes that make as few changes as possible in conmcting new words

fiom old ones.

Given the word formation options of English, this principle presumes that conversion

(zero change) should be used before compounding and afXxation. Clark (1980a) proposed

that because of the interaction between semantic transparency and f o d sirnplicity,

conversion and compounding can be found to occur equaily early in children's speech. She

added that compounding may be more casily accessible to leamers and easier to use than

conversion because of the uansparmcy of meaning of its elements. For instance, the tenn

nian in the English compound noun cut-man is available to make the meaning of the

compound transparent to the leamer. The change fiom the noun bicycle to the wtb lu

bicycle although it is a formaliy simple process, is less transparent in meaning. This is

because "the precise relation bawecn this nom and verb is itself a conventioaal one that has

to be Ieamtn (Clark, 19804 p. 12).

The pnncipte of simplicity dso presunes that campounding and conversion wili be

used before affixation to win new words. Affixation in English, for example, is argued to be

an opaque proces cornpanxi to compounding and conversion. This is because affïxation cm

cause pronunciation changes in the mot, e.g. activatelacInrity in English. Thus, combining

two independent nouns or wnverting nouns to verbs is seen to be easiw than forming nwns

fiam verbs by adding sufixes (Cl* 1980a). Evcn within affkatioir, this principIe

presumes that &xes should be acquired before prefixes and infixes because suffixes

appear to be "more salient perceptually than prefixes or infixesi' (Clark et aI., 1984, p.571.

In addition, this principle presumes that affixes that requue a great n u m k of changes in the

base will be acquued later than others which undergo fewer changes. According to Clark, in

Engiish for example, the suffix mss should be acquircd earlier than the su& -i@. The

reason behind this is that the fmer does not require any change in the base to which the

suffvs is added, e.g., whitehhiteness. The latter, on the other band, "requires a shift in word-

stress and often a change in a vowel or consonant (or both) of the base as we11". For

example, coining the noun eleçtricity h m the adjective vfecme involves the shifi in stress

h m the second syllable to the third, and alsa a change 6om the the final /k/ of the adjective

to /S/ in the noun (Clark, 1980a, pp.12-13).

d) Conventiondity

The fourth p h i p i e that, according to Clark, accounts for the acquisition of word

formation processes is the principb of conventiodity. This latter States that conventional

words or word formation processes skuuid be acquired and uscd earlier-

Accurding to Clark, wbat children may implicitIy assume to be conventional at any

given time is exhiiited in their repertoire of vocabuIary and word formation processes. But

tfieir implied beliefs about which words are conventional will change as they add more

processes to their basic stock (Clark et a!., 1982, p.6). For exarnple, if compounding is the

only process in the children's basic repertoire, it is assumed that this process is the

conventional way to fonn new noms, and it is used whenever children want to express a

word. But once they corne across more specialized options, they change theu implied

presumptions about what is conventionai in the language.

An obvious strategy that learnem depend on when they want to express new

meanings is to pick up on the adult conventional device and use it in preference to other

devices. For example, English-Epeaking children would replace innovative compounds such

as car-smoke with the conventional word exhmsf, and an innovative agent noun such as

bicycler and an instrument noun as drilkr with the conventional words qclisr and &Il

respectively. In each case, the conventional well-fonned word already existing in the lexicon

of the language supplants the child's innovation that is intendeci to express that rneaning

(Clark et al., 1982, p.6).

in summary, the four general finciples (productivity, semantic transparency, fonnal

simplicity, and conventionaiity) are argued to make for early acquisition of typical and

available word formation processes, and play a roIe in children's choices of word formation

processes in constmcting new words. These general principles were tested in several studies

on the acquisition of LI word fbmation proceses by exploring children's and adult

formation of lexical innovations in different s e d c domains and in diffierent languages. In

the next section, the outcomes of some of these studies are reviewed. Examples extradeù

fiom the relevant sources are used for illustration+

3.1.2 Empirical Evidence

A review of the literature on the acquisition of word formation processes reveals that

research on children's gradua1 acquisition of L1 word formation devices is abundant in

cornparison to L2 research. L1 studies have looked at the role of these general principles in

chilâren's coinages of innovative agents, instruments, and negative vérbs within and across

languages.

Clark and associates (Clark, 1980a, 1980b. 1981a, 1982; Clark et al., 1982; Clark et

al., 1984) have contniuted a series of studies on English L1 acquisition in which they have

demonstrated that transparent, productive, simple, and conventional word formation

processes appear early in language acquisition. One noteworthy study by Clark et al. (1982)

looked at how children acquire the conventional adult word formation processes for coining

novel English agent and instrument nounq the course they foUow during acquisition, and the

factors affecthg their acquisition of word formation processes. in accordance with

predictions fiom the transparency principle, the youngest children (aged two or three),

depended initially on simple cornpounds ofknown elements and often used them for agents,

e.g., wagon-girl for 'a girl who puiis wagons', r clmner-pplt for 'a person who cleans'.

Also, combining strategies from both the transparency and productivity principles, older

children seerned to idente the -er su& as an agentive marker and use it produaively for

coining agents in preference to the two other agentive suffixes, -ist and -im. Examples of

innovative agent nouns with -er that o[der c h i l d m came up with are bitter, givtr, and

pnsenttr.

Similar trends were reporteci in Clark's (1980a) study dealing with the L1 acquisition

of French agents and instruments. In this shidy, it was found that aside tiom compounding,

the two most productive sumes -ier and -eur, were acquired earliest by young children and

were the only ones to appear in thek coinages. Some examples of French lexical innovations

carrying the sufiixes -ier and -eur that the youngest children (aged two or three) came up

with are mrmnier to designate 'a seller of chestnuts' based on mmon 'chesmut', and

cenonnitr for 'a mender of hoops' based on cerceuu 'hoop'; crêmtur 'for an eater of

cream', fiom crème 'cream', and ouvreur 'for an opener' fiom uuvrir 'to open'.

Another study by Cl& (I981b) investigated the effects of productivity in children's

coinage of negative verbs (verbs expressing reversais and misdeeds) across English, French,

and German. To convey the notion of reversai, English uses the negative prefix un- as in the

verbs utdo, unbutton, and the particles ofland mt as in pull ofor take out. English dso uses

lexicalized verbs such as come/go. To express the notion of misdeeds, English uses the

prefix mis-, as in the well-esîablished misgui&. Of al1 thcse devices, the negative prefix un-

is wnsidcred to be highly productive and is widely used to fonn new verbs expressing

reversal. French, on the other hand, relies an a single device for the expression of reversai,

the prefix dé-. As in Engiish, French aisa uses kxicalued reversal verbs such as é t e i d e

hini o f and allumer 'tum on'. Of the two devices in French, the prefix dé- is the most

productive device for coining new mersal verbs. To express misdeeds, French. like

English, uses a p r e k m'- as in mBi' 'do someone W. German, on the other han4 uses

the negative prefk ent- or a negaiive particle associated with the verbs, usuaily atcs 'out, off

or ab 'down, hm', ta denote reversai. As in English and French, Gerrnan uses lexicaiUed

reversal verbs such as kommen 'come' and gekn 'go'. None of these German devices is

productive. in addition, Gennan uses two negative prefixes, ver- and mg- to express

misdeeds and both these prefixes are somewhat productive (Clark, 198 lb, p.42).

in this study, findings showed that the pnnciple of productivity had a direct

influence on children's innovations in the three languages. Children opted for the most

productive devices in coining their innovations, and where there is no productive device,

they sirnply did not coin any. b the English data, the most productive device, the preiïx un-,

was found to be acquired relatively early by young English-speaking children, and was the

only prefix used in couiing new reversai verbs. Typical examples of these innovations were

uncapturc 'release', uncnind 'make less crowded'. English-speaking children also opted for

some negative particles.

in the French data, children as young as 2;6 relied on the single productive process,

the negative prefix &-, to express reversai of action. d o n g the innovations produced by

French-speaking children were dCbitir '10 demolish' and déchauffer, 'to make less hot'. For

misdeeds, neither English nor French-speakiag chiidren used the unproductive prefixes mis-

and mé-. hstead they used &-. Udike Ewish and French-speaking children, German-

speaking children did not coin verbs denoting reversais of action because of the absence of a

productive process denoting this semantic notion. For rnisdeeds, however, they produced

innovations with the productive prcfur ver-.

Berman and associates have contributed studies on the L1 acquisition of word

formation in Hebrew - a Scmitic language, which like Arabic has a rather different

typologïcai structure fiom Indo-European languages (e.g.. Berman & Sa@, 1981; Berman,

Hecht & Clark, 1982; and Clark a al., 1984). Of note is Berman et al.'s (1982) study which

examined the devices used by native Hebrew-speaking children to coin agent and instrument

nouns in Hebrew, and the influence of general principles in coining these notions. Based on

data from 60 children (aged three, four, five, seven, and eleven), and 12 adult native Hebrew

speakers (most of them coiiege students in their mid-twenties), Berman et al. demonstrateci

that different word formation devices were preferred at different aga.

In Hebrew, the cornmonest devices used to coin both agent and instrument nouns

are: vowel insertion, as in /tsayar/ 'paintef, /vasai/ 'regulator'; pattern CaCCan, e.g., Jssrxkanl

'actor', and Imazganl 'air conditioner'; an association of word stem+ either s u e /an-/, lm-/,

/ay/, or /jp/, as lyetsu-aal 'exporter', /sandl-ar/ 'shoemaker'. liton-ay/ 'joumalist', and

kanuk-iya/ Wanukah candlestick' ; aad conversion, which involves forming nouns fiom I

present tense participial verb forms known as benoni, e.g., /$ofeV ludges'/ 'a judge', and

fmotsetd 'sucksl/ 'a pacifier'. Two other word formation processes, used mainly for

instrument nom, are the prefixed pattern ma-CCeC, e.g., /masreV 'comb', and its ferninine

nominal form ma-CCeCa, e,g, /mamtera/ 'spruiktei. In addition, compounding in Hebrew is

not cornmon for agents and instniments. However, a number of lexicalized compounds

expressing instruments exist in the cumnt lexicon of Hebrew (Berman, et ai., 1982, pp. 18-

Findings of this study showed tbat for agents, chiidren h m age four an reiied on

semantic transpareacy by finî using the agentive su& 1 4 since it is specialized for

agents (Berman et al., 1982, p.22). The sufIix /-cm/ was also prefmed for instruments by

children as weli as adults. The /ma-/ p r e h pattern, on the other hana was not used at dl by

children for coinhg instruments. Formal simpticity was also reiied on especially by the

youngest childrm (he-year-old) d e n they used the pnsent tense participial verb form

(bemni) for agent and insîrument nouns. Compounds were rardy opted for and they were

only used by seven-year-old children for instrument nouns. Findings of this study were

compared with those of Clark et ai. (1982) in order to distinguish patterns of response that

may be language-specific fiom patterns reflecting more general and universal principles of

acquisition, and to determine if the same generd acquisitional principles apply across

languages that are stxucturally dflerent. In most cases, patterns produced by either L1

learum of English or Heôrew were shown to be similar. The only difference noted was the

late appearance of compounding in Hebrew wmpared to English. According to these

researchers, this is attributable to the low productivity, and Iack of utility of compounding in

Hebrew (Berman, et al., 1982, p. 15).

Ariother study of Hebrew LI acquisition t h supports the primacy of Clark's

aquisitional principles is Clark et al. (1984) which investigated the appiicability of these

aquisitional principles in Hebrew-speaking ciLlIdren's and aduhs' coinages of novel agentive

and instnimental nouns, and compared their pattws of acquisition to those of English-

speaking children. In the comprehension data, aü children, even the yomgest (3 year-olds),

were most likely to opt for patterns with the suffix 1 4 for agents. The iniixed pattem,

however, was less used by al1 age groups (3 through 1 I). For instnimentq patterns with the - an suffix were also used more by younger chilâren than pattmis with prefix m. The later

acquisition of the instnimental prefix /ma-/ is pcrhaps due to degree of cxposure. According

to Clark et al. (1984). younger Heùrew-speaking childrtn may not yet have heard enough

instrument nouns in the lmaCCeCI pattem (Ciark et al., 1984). In the production data as

well, Hebnw-speaking chiIdren depended mostly on the suffix /a to coin both agents and

instruments rather than the prefix lmri-II Younger children (three-yearslds) fàvoured

conversion by producing some present tense partiparticipiai verb forms for agents. Compared to

the results for English (Clark et al., 1984), compounding was produced only around age five,

with a slight peak at age seven, in contrast to English-speaking children who often relied

initially on compound f o m (at age two or ?hm). Hebrew-speaking children also gave more

compounds for instruments than Engtish-speaking children, "This evidently results ftom

language-specific asymmetries among compoiind nouns in the m e n t lexicons of the two

laquages" (Clark et ai.,1984 p.567). Hebrcw cornains many compounds for instruments

using the nouns /mexona/ 'machine', and /kli/ 'implement', while English coatains many

agent compounds like milkman and postman.

3.1.3 Summrry

in sum, the above review of L1 studies indicates that younger children, as early as

age three, possess the ability to coin new words, and this ability develops with age as

differences in preference for word formation processes are found between younger and older

children. The review alsa reveals that productive, simple, transparent, and conventional

word formation processes are acquirad eartier than unproductive, complex. unconventional,

or opaque ones. Moreovcr, it shows that L1 leamers generally rely on similar acquisitional

principles regardless of the language being acquired, but because of the peculiarities of the

specific L1, the patterns of acquisition can sometimes be different.

3.2 Word Formation in %and b g u r g e (L2)

3.2.1 Rincipks of Acquisition of Word Formation Procemes in L2 Studk

Research on word formaîion in L2 is rare compared to such tesearch in LI. The few

studies that haw been coaductcd on the acquisition of L2 word formation processes by

children and adults are Kennedy-fonkn. (19û4); Bahat (1986); Olshtain (1987); Broeder

(1991); and Broeder et al. (1988, 1989, 1991, 1993, 1995). The followhg paragraphs wiU

review some of these studies,

One case study of note (Kennedy-Jonker, 1984) investigated whether the

acquisitional pnnciples discussed above obtained in the lexical innovations of two bilinguai

children in English and Dutch. and wbether these children used the same strategies in both

languages. It was found that pmductivity, semantic tfansparency and conventionaiity

principles guided the two bilingual children in producing innovative agents in English and

Dutch. in their English innovations, both children adhered to the principles of productivity

and conventionality by relying on the suffix +r for coining agents, e.g., smokcr for

'somebody who smokes cigars', brbyminder for 'sornebdy who minds babies'. in the?

Dutch innovations, they relied on both semantic transparency and productivity pnnciples.

Compounds with either meneer or wouw, and the suffix sr were used equally in i~ovating

agents. Examples of compound agents were pastmeneer 'postman' for somebody who

bnngs post, and automcvrouw 'automman' for somebody who fixes cars, and with the

suffix -er are harcnwwaer 'haimasher' for somebody who washes hair, and brkkcr

'baker' for wmebody who sells bread.

In another study, Olshtain (1987) investigated the acquisition of Hebrew word

formation processes by intermediate and advancect ad& L2 leamers of different L1

backgrounds (Englisù, Spanish, and other Ianguages) and by native speakers. Findings of

this study showed that the choice and the prcfmnce order ofthe L2 leamers in Hebrew was

subject ta some of the acquisitional principles. in accordance with the prïnciple of sernamic

transpamcy, compounding was found to be used more fiequentiy than derivation by

advanced leaniers. intermediate learaers, on the other hami, showtd a sironger prcfêrence

for afnxation Sufnxes such as /-ani and /-4y1' were especially preferred for agents.

According to Olshtain, the preference for fixation over compounding by intermediate

lemers was attniutabIe to their exposure to, and tàmiliarity with, this process as a resuh of

instructional emphasis. f ixat ion had been covered in their course, and intermediate

learners had spent time practising it (Olshtain, 1987, p.229). The findings aiso showed that

advanced learncrs' behaviour patterns in selecting word formation strategies were similar to

native Hebrew-speakers', indicating that the advanced L2 learnen possessed a level of target

laquage cornpetence that enabled them to produce innovations in Hebrew in ways that

approximated native speakers' responses (Olshtain, 1987, p.23 1).

Broeder in collaboration with other researchers has also contributeci to research on

L2 word formation with various longitudinal cross-linguistic studies on the acquisition of

word formation processes in coining lexical innovations in différent semantic domains such

entities, agency, instnimentality, location, kinship, and possession (e.g., Broeder et al., 1989,

1941, 1993, d 1995). The following paragraphs bighlight somc findings tiom these

studies.

in a longitudinal cross-linguistic study, Broeder et al. (1993) ewmined the use of L2

Dutch, Engiish and Swedish watd formation devices by adult native speakers of different

languages (Arabic, Finnisù, Itaiian, Punjabi, Spanish and Turkish) in coining innovative

agents, instruments, places, inhabitants, eountries, etc. Tbis study revealed that in the second

languages studii @utch, English, and Swedish), al1 lemer groups showed a high

preference for mmpounding at an early stage of acquisition They al1 made creative and

innovative use of a variety of compounding constnictions. N+N compounds, including

N+prep+N and X+N+N (having complex modifia), were by far the most fiequent Wlthin

these categories, head-final compounds dominated over head-initial compounds For ail

learner groups, even those for whom the source language is head-initial (e.g., Moroccan

Arabic lemers). Examples of innovative agents fiom the three lemer groups are provided

below.

In L2 Dutch data, Turkish- and Moroccan-speaking leamers produced head-final

compound constnictioas combining either mens 'person', man 'man', vrouw 'woman'. baa

'boss', mensen 'people',etc. to denote agents. Among the compounds produced by

Moroccan lemers were: boer-menwn 'farmer-people', bakkwman 'bidcer-man', vrucht-

vmuw 'foetus-wife', boer-mdsje 'whore-girl*; and by Turkish leamers, buittnland-

mensen 'foreign-people', koning-vrouw 'king-girl', and braad-bras 'bread-boss'.

In English LZ, among the N+N compounds ptoduced by Punjabi leamers were

brtad-man for 'baker', housaman for 'husband', and restaurant-+gaffer for 'restaurant

manager' while the M a n leamers produced the innovative Engiish compounds the

manager the shop, and the boss the shop. in Swedish L2, head-final N+N compounds

produced by L2 Finnish-speaking leamers included am-rnranea 'sale-manT, bibpolis

'car-police', and by Spanish leamers huvet-poüsen 'head police'.

Compared to compounding, derivation was rarely used in a productive way by the

L2 leamers. A few lemers occasionally attempted to coin derived innovations at a late stage

of acquisition. For example, in L2 Dutch, one of the Moroccan leamers used the Arabic

ferninine gender s u e (4 with the standard ûutch kinship term twrn (male teference

uncle) mm-a, (unclc-a) to refer to 'aunt', and with the word dortobi to rdér to a fernale

doctor (Broeder et al., 1993, pp. 56-57). The same leamer uxd non-agentive h k c ~ j

'bakery' to denote the agent W r 'baker'. Aiso in English L2, there was only one

occurrence of derivationaf innovation whwe one of the ïtalian speakers used the suffix -er

for an agent noun, e.g., blaektr 'black man'.

3.2.2 Summa y

The above review of L2 studies supports the claim that Clark's general acquisitional

principles account for learners' use of word formation processes in L2 Hebrew, Dutch,

English, and Swedish. In most cases, the semantic transparency principle was found to guide

children and adult L2 leamers' coinages when they made use of campounding a! an early

stage of acquisition, In addition, this review reveals that in relation to Broeder's study, head-

final compounds were prwalent even for learners with a head-initial LI background. This

mggests that L2 leamers were influenced by the target language word order ptinciple of

compounds. Derivation, however, was rare and appeared late.

3.2.3 Transfer w an Added Principlt

In the acquisition of L2 word formation processes, besides the four acquisitional

principles (productivity, semantic transparency, focmal simplicity, and conventiodity), L 1

tmsfer can be an added factor. "Cl transfer usually refm to the incorporation of features of

the LI into the knowledge systems of the L2 which the learncr is üying to build" (EUis,

1994, p.28). Although it is widely assumed that language transfer plays an important role in

L2 acquisitioa, its nature and signiticance rernain controvmial. Transfer may manifest itseIf

as m n (the foais of early studies), and impede the acquisition of L2. On the other hand,

transfer can be positive, and thus hcilitate the acquisition of L2. In his review of the

phenornenon of transfer, ûdün (1989) dehed it as "the influence resulting ftom the

similarities and diReances betweea the target language and any other Ianguage that has

been previously aquired" (p.27).

Evidence for transfer in ail aspects of language - phonology, syntax, semantics, and

pragmatics - is found in abundance in the L2 acquisition litmature ( eg , Gass, 1979, 1980;

Olshtain, 1983; Kelleman, 1979; Schitchter & Rutherford, 1979; and Odlui 1990).

Howwer, in considering the acquisition ofword formation processes in L2, transfer 6om LI

has not been given much attention, and the few studies that have dealt with this phenomenon

(tg., Olshtain, 1987; Broder, 1991; Broeder et al., 1993; and Broder et ai., 1995) have

revealed conflicting results.

In Olshtain's study, the advanceci learners consisted of students of various langage

backgrounds, but two groups of lmers (3 1 Engiish speakers and 28 Spanish speakers)

made up large enough sub-groups to investigate M e r th& responses for L1 effect on the

preferences for L2 word formation F e s s e s . Although the L2 leamers had different L1

backgrounds, a similar pretérence for compounding, especidly on an elicitation tasic, was

found, Engiish speakers, like Spanish speakers, exhibitecl a high preference for

compounding (45% and 44% respectiveiy). According to Olshîain, these two sub-groups of

leamers are representative of ai i advanced iearncrs because overd results showed 46% for

compounding arnong all th& advanced learners. Also, these sub-groups' choice of othcr

word formation devices was similar for the two large groups of learnm.

However, in Broedcr (LM), and B d e r et al.5 studies (1993, 1995), there was

c h evidence of LI-based orda pretèrences in compounds. In the following paragraphs,

fùidiags fiom Broeda(l991), and B d e r et al. f 1993) art presented for illustration,

In Broeder's (1991) cross-linguistic cornparison of the L2 acquisition of Dutch and

German possessive relationships by Moroccan and Turkish learners, findings show that

Moroccan learners of Dutch fiequentiy produced the N-van-N construction which

corresponds to the spoken Arabic construction N4yaI-N. Moroccan L2 iearners used van

'of in head-initial constructions, following the word order principle typicai to their LI, e.g.,

h k van hem 'book of him', and bock van mij 'book of me'. However, Turkish leamers,

for whom there are no such simihrities between L1 and L2, show evidence of LI

prefwences in later stages of the acquisition process with the van-Pro-N construction. At an

early stage, L2 Tutkish leamers of Dutch tended to use the construction Pro-N, e.&, mijn tas

'my bag', ziJn w m ' his woman', But, in the later stage of acquisition, they coined

compound innovations with the preposition van 'of, following the orda of their L1 word

order principle head-finai, e.g., van hem fiets 'of him bike', van hem familie 'of his family',

and van mijn tas 'of my bag'.

Also in Broeder et al.'s (1993) crossiiiguistic study, results show evidence of LI

influence in the production of innovative compounds. Moroccan Iearners tended to constmct

N+prep+N with a head-initial word order pattern, where Standard Dutch would favour head-

final word order. hovative compounds produced by Moroccan learners of ûutch included:

ktrk van Mmkko 'church of Morocco', kltren van baby 'clothes of baby', and bras van

winkel 'boss of shop', where ûutch speakers would say moskee 'moque', babyklerert 'baby-

cIothesl, winkeleigenmr 'shopkeeper' respectively. Also in the Swedish L2 data, the Spanish

learners appeared to be influenced by the cornpounding pattems of theu Lt (8roeder et

al.,1993). They produced N+prep+N constnictions in Swedish using the Spanish preposdion

de 'ofhm' that expresses both possession and location in their L1 (Brder a al., 1993, p.54).

h o n g the compounds produced by Spanish leamers is tabletter de chokolrd 'bars ofin

chocolate' for choklad-tablette 'chocolate bars'.

Even though L1 influence on L2 word formation appears to be minor, and restncted

to the coining of compounds in adult L2 acquisition, it should be taken into consideration

and added to the general acquisitional principles which play a role in the acquisition of word

formation processes.

To conclude, this review of Ll and L2 studies demonstrates that in extending their

lexicon, both LI and L2 leamers of different languages (except Hebrew L1 leamers)

sirnilarly make creative and innovative use of a wide range of compounding constructions in

eariy stage of acquisition. According to Broeder, this similarity is a remarkable indication of

universal processes of language acquisition (Broder. 1993, p.71). Additionally, in L2

studies, leaners rely both on targd and source language devices to coin compaunds.

3.3 Rcscvcb on the Acquisition of MSA

3.3.1 Classical Studics

Most past research on Arabic has dealt with Arabic grarnmar and presented

synchronie or diachronie descriptions of Arabic morphology and syntax. ïhese classical

studies have paid hardly any attention to the acquisition of Arabic. Recently, however, a new

reseatch trend has started to deal with the acquisition of Arabic. Lately, a few empirical

studies have been conducted on the acquisition of MSA Arabic and its varietics as an L 1iL2.

ïhese studies have investigated different areas of Arabic and various aspects of its

acquisition includig the role of instruction and learning strategks, communication

stmtegies and 0 t h factors that may influence its acquisition (e.g., Fakhii, 1984; lrshled &

Whelan, 1988; Khaidieh, 1991; Suleiman, 1991; and Aweisq 1993).

3.3.2 Studia on the Acquisition of Anbie Word Formation Processes

Studies dealing with the acquisition of word formation processes in Arabic are scarce

compared to studies on indo-European laquages, and to my knowledge none concemed

with the L2. Iearning of MSA word formation has yet been conducted. The only study that

has dealt with the acquisition of Arabic word formation processes is by Badry (1983) who

investigated the acquisition and organization of Moroccan Arabic derivational patterns in

LI. Forty Arabic-speaking Moroccan children canied out six tasks designed to assess their

ability to use Semitic lexical derivation productively and receptively at different age levels,

and ta investigate the types of strategies they followed, and the general principles that

acwunt for their use of derivational processes to coin verbal and nominal pattern.

The findings of this study show that the children were able to use the derivational

processes to coin new words fiom as early as age 3 5 , and this ability increased irnponantly

around age five @dry, 1983, p.165). Because both verb production and comprehension

taslrs examined wtiich of the three different verbal forms (causatives, reciprocal, medio-

passive, and basic form) was the most Eequently used, and because the purpose of the

present sîudy is to investigate the use of different word formation options available to

express the same semantic notion, discussion of the findings of these two v&-focused

tasks is irrelevant here. In Badry's noun production and comprehension tasks, the notions

invesîigatd were agents and instruments. Findings fiom the noun production task indicated

that in their innovations, al1 children in al1 age groups adhered to the principles of semantic

t raqamcy and productivity by depending excIusively on the most productive and

semantically transparent pattern CeCCaC in naming agents. Among the agentive innovations

produced by these children were nebbasl 'dresser' and Ineqqad 'jumper'. The patterns

CiCeC and muCaCCiC were not used to name agents ostensibly because they are both less

productive in Morocwi Arabic. In addition, the pattnn CiCeC lacks transparency in that it

has both agent ive and participial fiinctions (Badry, 1983, p. 158). In naming instruments,

children also relied on produdivity by opting only for the most productive ferninine pattern

CeCCaCa, e.g., Ilebbasd and /neqqaza/. Neither the dcrived pattern rniCCaCa nor the

compound constructions N-in-construct and N+prep+N were used in naming instniments.

In the aoun comprehension task, children were given the three options (CeCCaC,

CZCeC and miiCCaCiC) for the agentive nouns, and the two options (CeCCaCa and

miCCaCa) for the instrumental nouns, and were asked to describe the activity canied out by

selecting one of the options. Findings fiom this task showed that the children's choices were

abject to some of the acquisitional principles. In accordance with the pmdudivity principle,

the CeCCaC fom was mostly pre fed to express both agent and instment notions in al1

age groups. ibis high prefetcnce was manifested by the children's tendency to transfocm

other patterns imo the pattern CeCCaC. According to Badry (1983), "this conversion

suggests that children may have developed a strong relation betwan the notion of doer of an

action, either human or machine, with the pattern CeCCaC, which is the most productive

pattern in the adult systemn (p.174). The use of the pattern CeCCaC especially for agents

was dso evidence of the one-twne mapping strategy between form and meaning. The

pattem CiCeC was not opte- tOt at al1 apparently because it expresses two semantic notions:

active participle and ztgentive meaning.

3.4 Prtdictions about the Order of Prefertnce of MSA Word Formatioa Processes to Coin Ageaey, Insîrumentrlity, Location, and Causrtivity

Since Arabic makes heavy use of derivation in the coining of new content wordq

leamers would be more liely to discover the pmductivity of the derivational systern first

and overgeneralize it, in preference to the 0 t h principles of semantic transparency and

formai simplicity. Leamers in early stages of acquisition might function with the hypothesis

that al1 words are formed througb derivation, and once they have abstracted the d e that

their lexical system is based on derivation they would tend to use it werywhere to coin new

words.

Given the various MSA word fonnation processes for coining agents, instruments,

locatives, and causatives (sec Table 2:4), and based on (a) the results of studies on L1 and

L2 Hebrew (a Semitic laquage with a derivational system quite similar to MSA), (b)

Badry's predictioas (1983, pp.91-95) about Moroccan Arabic word fonnation processes, and

(c) the judgment of some native speakers, a number of the predictions about the order of

preferences of MSA word fonnation processes in coining these semantic notions are

formulateâ. Some of these preâictions conflict.

First, in accordance with Eve Clark's (1982a) productivity principle and with Badry's

predictioa (1983), one might expcct that for agent nouns in MS4 leamers would rely more

on the very productive pattern CaCCZC than on the less productive patterns CZCiC,

muCaCCiC, and N-in-constnict (/rajul/+ N). With respect to instrument nouas, the very

productive pattern CaCCiCa could bc rerelied on before the less productive pattm CZCiCa,

and rniCCaCa For locatives, the prefhed pattern rnaCCaCa, being the only derive. partmi

to express location, would likely be rnost eequcntly used than the N-inansmict

(/makW+N). For causatives, because of its high productivity7 the verbal pattern CaCCaCa

should be used before the verbal pattern 7aCCaCa and the periphrastic construction with

/jaTaid 'to make someone (do something)'.

Secondly, the principle of semantic transparency would predict that devices with

one-twne mapping between forrn and semantic notion wouid be used earlier by leamers

tban devices with two meanings. Thus, N-in-constnict compounds which consist of

combining either the word /rajui/ 'man7, /?ala/ 'machine', or / m a W 'place' with a noun

would be the easiest to identiîy and to use as these words express the notions of agency,

instnimentality, and location explicitly. Next for agents, leamers might be expected to use

the pattem mCaCCiC having only the agentive meaning, then the pattem CaCC6C which

a n have either agentive or instrument meaning, and pattern CaCiC which can have either a

masculine agentive or active participle meaning, For instruments, beyond compounding,

they would be expected to use the pattem miCCaCa with only instrumental meaning, and

then CaCCiCa and CiCiCa, both of which can have either a ferninine agemive or

instnimental meaning, For locatives, next a e r compounding, it would be the pattern

maCCaCa having only the locative meaning. Also, for causatives, semantic transparency

could mean that the periphrastic construction with tjafalai 'to makc someone (do

somcthing)' will be easier to produce and to idem* than the derived verbal forms

CaCCaCa and 7aCCaCa The periphrastic construction is wnsidered semanticalty

minsparent because of the idea of causativity that the verb /jaTala/ conveys. This assumption

is borne out in diercnt studies of the L1 acquisition of English (Bowerman, 1974; Cl*

1982) where pesiphrastic causatives with 'make' and 'get' are found to appear prior to other

causaiive verb forms. In Arabic, this principle confiicts with productivity.

Third, in line with Eve Clark's formal simplicity principle (Clark, 1980a), one rnight

expect that stem-ex&mal f i e s would be relied on befbre, and used in preference to,

infixes- This means that for agents, leamers would use the prefixed pattern rnuCaCCiC

before the infixed andfor geminated patterns CaCCiC and CZCiC. With regard to

instruments, learners might be expected to use the prefixed pattern miCCaCa more

fiequently than the infixed andor geminated pattmis, CaCCàCa and CàCiCa. For locatives,

the prefixed pattern maCCaCa would be used before any other option. Finally, for

causatives, the prefixed pattern 7aCCaCa would be used before the infixed CaCCaCa. This

principle conflicts with productivity in Arabic.

Fourth, the conventionaiity principle as defined by Clark a al., (1983) would predict

that parteni CaCCiC for agents and pattem CaCCiCa for instruments, being most often

conventionally uscd in the MSA lexiwn, and pattern maCCaCa for locatives, would be

opted for and ovenised. For causative verbs, the conventionality principle could also mean

that L2 leamers woufd chose the pattern CaCCaCa and overuse it to coin lexical

innovations.

In addition to these four acquisitional pn'nciples, the -fer principle wouId predict

that there will bc signifiant diffcrences between French-speaking and Engiish-speaking

learners in coining lexicai innovations due to LI influence. Differences in the word order of

compounds in French and in English, for example, may diffmntially privilege the choice of

certain orders ovcr otùers in Arabic-

Mormver, it is predicted that L2 v d u l a r y knowledge in the target language will

reveal diienntial effects for the use of word formation processes. The more advancd a

lemer iq the more words thcre are in his or hcr rcpertoire, and the more innovations are

Iikely to be produced (Broeder et al., 1995). Also the more advanced the learner is, the

higher the [evel of word formation knowledge they reach, Ttiis knowledge will enable them

to produce and judge innovations in ways that approach native speakers' performance

(Olshtain, 1987). Table 3:l surnmarizes the predictions for each principle.

Table 3: 1 '

Surnmary of the predictions made by each principle

CâCiC

Instrument Locative Causative

makan+N j aFala+Pro+ Adj maCCaCa CaCCaCa

?aCCaCa

maCCaCa ?aCCaCa CaCCaCa

maCCaCa CaCCaCa

CHAPTER 4

RESEARCH METHODOLOGY

L2 leamers divided into three levels of relative vocabulary knowledge ('lower',

'rnid' and 'higher'), as well as native speakers of MSA, participateci in this cross-sectional

study. The study attempts to: a) examine leamen' and native speakers' production of lexical

innovations in MSA word formation processes and their judgment of innovations based on

MSA word fonnation processes; (b) detemine their preferences for word formation

processes at different levels of vocabulary knowledge; and (3) investigate the influence of

the aquisitionai principles of productivity, semantic transparency, fonnal simplicity,

conventionaiity, and transfer on the Iearnets' use and choice of word fonnation processes in

MSA. These questions were investigated both productively, when participants were asked to

construct innovative fonns to denote new meanings not expresseci in the language, and

receptively, when they were askcd to judge for both invented and real items which of the

given alternatives was the most suitable form.

4.1 Ruearcb Questions rad Hypothesu

The research questions and hypothcses of this research have been fonnulated on the

basis of theoretical views on the acquisition of word formation processes, research findings

on this aspect of lexicon (discussed in chapter 3) and the characteristics of MSA word

€0-on processes (discussed in chapter 2).

The six research questions guiding this research were originaliy stated in Chapter 1.

They are reiterated here for reference in relation to the hypotheses of the study.

Research Question 1: What are the quantitative and qualitative differences between

L2 lemers and native Arabic speakers in production of innovations in MSA expressing

notions such as agency, instrurnentality, location, and causativity?

Hypotbuia la: There will be a quantitative difl'erence between L2 lemers and

native speakers. L2 leamers will provide fewer lexical innovations, and give more thquent

non-responses than native speakers.

Hypotbesh lb: There will be a qualitative difference between L2 lemers and native

speakers in their production of innovations. L2 learners will demonstrate more variability of

patterns and forms in their production of innovations.

Ruearcb Question 2: Cornpared with native Arabic speakers, what MSA word

formation procesies) is (are) commonly selected by the leamers when choices are provided

for naming notions of agency, insuumentality, location, and causativity?

Elypothesis 2: L2 le8fners will tend to produce and chwse compounding andfor

piphrastic constructions more often than native speakers in naming agency,

instnimentality, location, and causativity.

Rcstrrch Quutioa 3: Does relative L2 vocabulary knowledge reveal differential

effects for the production of MSA word formation processes?

Eypotbuis 3: Level of vocabulary knowledge will infIuence the use of L2 lemers'

word formation praccsses, in that learners with a lower Ievel of vocabulary knowledge will

find it harda to make creative use of word formation processes tban higher proficiency

learners.

Research Question 4: How do L2 learners with a higher level of vocabulary

knowledge compare with native speakers in their produdon and choice of word formation

processes in MS A?

EypoihesW 4: H i g k level L2 Ieamen' production and choice of word formation

processes will be more like those of native Arabic speakers than bwer level learners' will.

RcJurcb Question 5: Which principle(s) (productivity, semantic transparency,

formal simplicity, and conventionality) is (are) most reveding in predicting L2 learners' use

and choice of word formation processes in MSA?

EiypothuW 5: L2 Lemers will tend to use productive, transparent, simple, and

conventional patterns in preference to unproductive, opaque, cornplex, and unconventionai

ones to express agents, instruments, locatives, and causatives.

Rcscrrch Question 6: Daes the native language of the L2 learner (English or

French) lead to differential use of word formation devices in MSA?

Hypothais 6: Therc will be transfer from LI to the L2 at al1 levels of vocabulary

howledge in the use of MSA word formation processes.

4.2 Participants

When I planned the study, my goal was to eonduct the research entirely in Canada,

and have larger mimbers of adult participants in two LI pups : English-speaking and

French-speaking clasaoom leamers of MSA as a fbreign ~angu;ip'~. It was impossible to

achieve this for the kliowing reasons, la most Canadian rui;versities that offet Arabic, the

level offered was not sutIiciently advancai, or toio many of the students that wcre enrollcd

came h m Arabic-speaking homes. I was able to colIect data in only one Canadian

university that satisfied my needs. Because the number of participants was too low for the

purposes of the research, 1 tunied to the United States, and after many attempts, managed to

enlist Wher participants tiom two different American universities. 1 also looked to France

for French-spesking leamers, but unfortunately rny attempts there were unsuccessful.

The initial leamer m p l e thus consisted ofa total of 70 adult L2 leamers korn thm

North Arnerican universities where MSA is tau@ as a second language. These adult L2

leamers were enrolled in the second year'4 of a MSA language program. The Arabic

program in the three universities differed with respect to the matends and the teaching

methods used by the instzucîors. In one of the universities, the textbaak used was Madem

St&d Arabie Intennediuie Level by Abboud, Altama, Emin, McCarus, and Rammuny

(1971). In this book grammar and derivational rules are introduced gradually in each text.

The instructional approach placed an emphasis on form ratber than meaning. Students were

introduced to stnicturai md lexical niles graâuaily and separately in each teaching unit with

explicit explanation of patterns and theu forma1 canstraints. Through drills and rote practice,

snidents had to memotize the rules and forms. The 0th- two universities, on the other hanû,

use a different textbook entiîled Tamm aiugha lfarabiyya by Brustaci (1997). The content of

this book foaises on the development of communicative campetence, and stresses context

and meaaing much m m than fonn Students are taught vocabulary in different contexts

with the aim of allowing them to wmmunicate effkctively and in a mannet appropriate to a

context rather than gaining and mastering grammatical accwacy.

Table 4:l

Distribution of L2 leamers by vocabulary knowledge level, gender, age, L1 background, and dominant language

Males

Females

18-20

2 1-24

24-30

3 1-35

36 & up

English

French

German

Italian

Persian

Somali

Spanish

Urdu

English

French

There was also a cornparison group of 40 native speakers of MSA The purpose of

including this group was to find out ta what extent the behaviour exhibiteci by L2 learners of

MSA, at various levels of vocabulary knowledge, resembled that of these native speakers

with respect to the use and choice of word formation processes. These native speakers were

university students in Morocco, and aH of them were in theu second year of an

undergraduate law program. There were more males than temales, and most of them had

some knowledge of other foreign languages (see Table 4:2 below).

Table 4:2

Distribution of native speakers by gender, age, and tanguage background

Al1 participants responded positively to a consent letter soliciting their participation

for this study and explainhg theu roles. Teachers were aiso given a consent letter asking for

theù permission to approach their students for the research (see Appendix A for both

Ietters).

4.3 instruments

The data to be analysed in the present study were collected by means of five

instruments: a vocabulary knowledge test in MSA, three Arabic word formation tasks (a

production task and two comprehension tasks), and a background questionnaire. Since there

was no appropriate L2 vocabulary knowledge test and there were no tasks on word

formation processes available in MSA, al1 tasks were designed by me. The appropnateness

of the task materials was detennined by the judgments of two L2 university Arabic teachers

and then by pilot testing of L2 leamers and native speakers of Arabic More the tests were

adrninistered to participants in the final study (see relevant section 4.4.1 on piloting).

4.3.1 Vocrrbulrry Knowledgt Tut

The L2 partkipants received a vocabulasy knowledgc test in Arabic (see Appendix

B). The purpose of tbis test was to measure L2 leamers' o v d l lexical knowledge in MSA

ui designing this test, 1 was inspued by Nation's A Vocabulary Levels Tesr (1990). The test

in Arabic included words that varied in fiequency. it WIN divided imo 5 sections

representing 5 levels of word frequency: Levels 1 & 2 consisteci of higb muency Arabic

words that were chosen respectively fiom among the 500 and l,W most tiequent words.

Level3 comprisecl words at the 2,000-wd level that are lower in fiequency than words in

tevels 1 & 2. Level4 consistai of words at the 3,000-word level, and level 5 contained even

lower frequency words at the 5,000-word level. The specific words in the test were selected

from the foliowing sources. Words used in levels 1-4 were seltxted h m two receiit Arabie

word hquency books: Ambic keyworcls by D. Quitregard, (1994), and D.E. Kouloughli's

(199 1) Basic Lexicon @Modem Stcmdwd Arabic, which consists of the 3,000 most fiequeat

MSA words. Because these two baab only cover words at levcls 1 to 4 (500-3,0), lower

Frequency words at the 5,000 were selected fiom Landau's (1959) A Word Count ofkforiem

Arabic Prose wntaining words at the 5,000 level and up. A more recent Arabic source

dealing witb Iower frquency words at 5,000-word level was impossible to find. Besides

fiequency criteria, my teaching m e n c e was dso celied on in selecting words that are

usudly taught early and that most participants potentidly know. The se1ected items consist

of content wards (nouns, vhs, and adjectives) that arc mostiy derived i3om verbal

consonantal roots, and represent the t h derivational processes in MSA (infixafion,

gemination and axation. For each level o f the test, thete wefe three questions, each one

dealing with one of the word classes (verô, adjective, or noun).

In each case the learner had to choose hee words fiom a list of six possible words to

match tfiree given synonyms, and write the lener of that word next to its meaning.

The folIowing is a sample verb question:

to 'A -1

to permit r L j ( - -+

to begin f & - G

t* try J J L -&

to declare & i -E

to faIl & -C

As part of the development process for this test, a check on validity was conducted.

Besides myself, two independent aduit test takers, who were native speakers of Arabic and

who were blind to the pwpose of this study, were given instrucîions to answer the

vocabulary knowledge test. There was 1W/o agreement in responses beîween these two.

A h , to assess the reliability of âhis vocabuiary knowledge test, an item analysis was

conducted on the final L2 leamer sample (N(N=4 The item-test reiiability yielded a value of

Cronbach's alpha of -90. Apart €tom its validity and reliabili, this type of test also met

p d c a i needs, being easy to mark anci not time-consuming to adrninister. Individual scores

on the terit were categonted as bigher, mi4 or lower as an indication of the relative

vocabulary knowledge lewl of leamers (see seciion 45.2 for scoring system).

43.2 Tbree Word Formation Tasb

The production and receptive choice of word formation devices in MSA was

examined using three tasks that were administered both to the L2 learners and to the native

speakers of Arabic.

43.2.1 Production Task

The first task was a production task that required participants to coin innovations in

MSA for agent, inshument, locative, and causative concepts (see Appendir C). It was

designed to investigate the types of innovations produced by L2 learners and native speakers

for concepts unnamed in the conventiod lexicon of MSA, and to analyse their productive

use of MSA word formation devices. The production task was administered before two

comprehension tasks in order to ensure that responses were not influenced by the patterns

provided in the two multiple chaice comprehension tasks.

There were 24 items in the produetion task eliciting innovations for four semantic

notions: agent, instrument, locative, and causative. Each notion was elicited via six items,

with al1 items presemed in random ordw. The choice of the three nominal notions (agent,

instrument, locative) was based on the Olshtain (1987) study in Hebrew (see chapter 3,

section 2). This was in order to provide additional data &om MSA which would allow

testing of the proposed genaal principles that are believed to govern the acquisition of word

formation processes. In the preseat study, 1 added a verb notion in order to see if leamers

would opt for the same formal options and follow the sarne strategies for both nominal and

verbal innovations. The c a u ~ i v c was the notioa chosee The reason for the choice of the

causative was that arnong the ten derived verb patterns that express différent semantic

notions, two pattern denote causativity. The two verb patterns are the infixed and

geminated pattern II CaCCaCa and the prefixed pattern IV 3aCCaCa.

For L2 participants, the items were written in either English or French for the

respective Iearner groups. A glossary was providecl in a separate list, giving translation

quivalents in MSA for Engiish or French vocabufary used in the items (see Appendix D).

Howwer, for native Arabic speakers, the task was given in Arabic with no glossary. In each

question, the stimulus to the lexical item was in bold, and participants were called upon to

produce novel words in MSA

For example, to elicit an agent innovation, one of the items was:

- in Arabic, how would you cal1 a penon who sets lim?

To elicit instrument innovations, an example item was:

- in Arabic, how would you cal1 a macbiac tbat hugs people?

4.3.2.2 Two Comprcheasioa Tasiu

The two cornprehension tasks were in multiple choice format. The aim of both

wmprehension task 1 based on invented items (Appendu E) and wmprehension task II

bsed on real items (Appendu F) was to provide insight into the L2 leamers' and native

speakers' choice of alternatives that they judged the most appropriate for labelling either

existing or unnamed agents, instruments, locatives and causatives. The r d items

(comprehension task a) were included in this study to estabfsh the validity of native

speakers' use of word formation proeesses, and to detamiae if these native speakers would

choose the same processes for boîh inventad and nal items. The two tasks focussed on the

same four semantic categories: agent, instrument. locative, and causative. In each test, there

were 30 items presented in tandom order: 8 agents, 8 instruments, 7 causativeq and 7

Locatives. Unfortunately, for çomprehension task B an ermr occuned in the proportional

number of questions which was noticed at the analysis stage. 1 ended up with 7 real agent

items and 8 r d causative items instead of 8 agents and 7 causatives, as was the case for

comprehension task 1.

In both tasks, the four alternatives were based on the four wwd famation processes:

v d i c infixation aIone, vocalic infixation and gemination of the 2"6 consonant, afixation,

and compounding or periphrastic constructions. The patterns ~ a ~ i c l a ) ' ~ , CaCCaC(a),

muCaCCiC(a) representing the three types of derivational processes were included as

options for agents. The wmpaund construction N-in-conmet (/rajui/+N) was inciuded to

represent the fourth alternative.

For imment items, out o f al1 the insrniment pattems presented in Table 2:4, only

the derived patterns CiCiC(a), CaCCaC(a), and rniCCaCa were represented as the t h

derivational options. The inclusion of these patterns reflects their relative productivity in

MSA Patterns CâCiC(a), and CaCCàC(a) are most îiequentty used in technical and

scientific minages in MSA (Al--1985; Holes, 1994). Pattern miCCaCa is most

fiequently used in the conta ofeducation @dry, 1983). The fourth choice cansisted of the

compound constniction N-in-constnict (/laldf N).

For lmtive notions, the preîïxed form (ma+stem), the ody productive option in

MSA to denote this notion, was pmvided as one option. in addition, a pattern CâCiC(a)

based on vocalic infixation, and a pattern CaCCaC(a) based on vocalic infiation and

gernination of the 2* consonant, were included as distracton. Also a compound construction

N-in-constnict (/makan/+N) was included to make up the fourth alternative.

For causatives, besides the intTxed and gemiaated verbal pattern CaCCaCa, and the

prefixed pattern lacCaCa, a distractor pattem was included to represent the process of

vocalic infixation alone: the verbal pattern CacaCa denoting reciprocity. A periphrastic

form (/jaiaia/+Pro+Adj.) was included as the fourth alternative,

For L2 participants, the stimulus questions were written in either Engiish or French

for the respective lemer groups. For each item, the four choices were written in Arabic with

large print and diacritics in order to promote the leamers' reading of the items, and were

numbered 1-4 representing the four alternatives. L2 participants were required to choose out

of the four alternatives (ordered in the same way for each item) the one that they judged the

most suitable for a specified concept. For netive speakers, both the questions and the choices

were written in Arabic with regular print and without diacritics.

A sample item tiom comprehension task I design4 to elicit an agent was:

- In Arabic, how would you cal1 r pema wbo mrkes bubblts with sorp?

For instruments, a sample item was:

- in Arabic, how would you cal1 a machine that breaks peaeüs?

4.3.2.3 Background Questionnaire

A questionnaire was administered to al1 L2 participants (see Appendix G). The

questions on the background questionnaire faIl into three categories: The first eight

questions asked about personal background, including gender, age group, level ofeducation,

place of birth, and L1 and dominant language; the next seven questioru dealt with

knowledge of other languages. Leamers were asked if they k m other languages, and if so,

where they leamed them, if they spoke and wrote them, and where tfiey used them. The last

ten questions askd about the Arabic language, reasons for studying it, and learning

experience. Questions also askcd &ut learncrs' self-ratcd performance in some areas, and

theu self-ratings of the levcl of diîtculty thcy had with some activitics in MSA Leamers'

responscs to four questions of the last category are presented below.

When leamers were askcd how they found Iearning Arabic, 34 considered it a

dificult language. Eight viewed it as neither easy nor dificult, while two considaed it easy.

When they were asked: "What rnakes Arabic a difficult language to lem?" out of 44

Iearners, 16 perceived it as vocabulary. 20 claimed it was gramrnar, five said that it was the

writing system, and three said it was pronunciation.

Also, when asked to rate on a three-point sale (below average, average, above

average) how well they are doing in learning Arabic compared to other learners, 26 leamers

claimed that their learning of Arabic was below average, while 15 leamers claimed bat it

was average, and only three Iearners that it was above average. Moreover, when asked to

rate on a four-point scale (same-better-worse-can't say) their performance in speaking,

Iistening, writing, reading, and spelling compared to the othm leamers, the majonty

answered negatively that they pnforrned worse than the otl~cr learners. Table 4:3 presents

the percentage of learners' responses for each skill which show that rnany of these leamers

perceive themselves as performing poorly in each skill.

Tabk 4:3

Leamers' self-ratings of their performance in some Arabic skills relative to classrnates

1

Listeniag 1 11.3 145.)rr) l I

Speaking 1 10.7 1 55.3

I I 1 I

Reading 1 7.0 ( 59.9 1 11.1 1 22.0 1

Furihermore, whm asked to rate on a fwr-point sale (unable to do, with much

difficulty, with linle difficuity, without difficulty) their communicative performance in

Arabic of some activities such as talking to a native speaker, watching and unchstanding a

23.0 11.0

TV program, reading newspapas and magazines, writing a letter, and understanding a radio

program, the majority saw themselves as performiq the activities with dificulty and o f h

as unable to do them Table 4:4 lists the L2 leamers' self-ratings for each activity.

TaMe 4:4

Learners' self-ratings of their performance in sonte activities

Unable to 1 With much 1 With link 1 Witbort

Talk to a native speaker

Watch and understand a TV 1

magazines 1 I I I I

program

Read newspapers and

I 4 1 1

I ' Understand a radio pmgrarn 1 60 1 39.2 1 0.6 1 0.2 I

5 2

56

An examination of the above questions reveds that the majority of learners

perceiveci leamhg Arabic to be dificult, and found vocabulary and grammar to be the

aspects that malce it a difficult language. The examination also reveals the lemers' Iimited

knowledge of MSA mrding to ttmeir low perceiveci pedbrmance in various skiils and

activities. A cornparison baween the responses of leamers expoied to a communicative

approach to Ianguage teaching and the ones taught throi~gh a more formai instructionai

methd rweals no différence in self-ratings of performance in Arabic skills and activities.

Native Arabic speakers were also given a version of the puestionnaire seeking information

on gender, age, and general education (see Table 42).

69

39.9

39.4

23.4

6.1

4

2

0.6

7.6 O

4.4.1 Pilot testing

Prior to the main study, pilot testing was undertaken. This was a necessary step

before conducting the research to ensure that the vocabulary used and the instructions in

each task were clear and effective, and to determine the time required for administration.

The pilot study took place in the winter tenn of 1996 in a Canadian university. Ten students

were involved: 4 males and 6 fernales. They were in their second t m of a begimer MSA

class. Information on participants' gender, age, LI background, place of bir& and level of

educaîion is presented in Table 4:s.

Tabk 4:s

Participants in the pilot study

Agt LI background Place of bidh Lcvd of univtrsity ducation

4 M d a 23 Englis h Britain

I I

6 Fcmaia ( 19 1 Arabic 1 Canada [ ;Y* Y

20

20

22

23

32

Engiish

French

French

Englis h

indonesian

Canada

Canada

Canada

Kenya

Indonesia

1" year

1" year

p - . . . - -. . - -

Graduate student -

Graduate student

The experimen? was canied out in the MSA classrwm in my presencq and the time

set to perfom the vocabdary knowledge test, the production ta& and the two

comprehension tasks was one hour. Only two participants out of 10 complaed al1 the tasks.

In the vocabulary knowledge test, one of the participants scored 52 out of 54 points.

and another scored 26 points. The remaining eight scored behveen 3-10 points. The

participant who scored high was eliminated from the sampie because he came fiom an

Arabic-speakuig home.

The pilot testing of the production task revealed the foUowing results. Only the

participant who scored 26 points in the vocabulary test nearly completed the task. Some of

the lexical innovations that this participant came up with were based on compounding

Among the compound consuuctions produced, N+N was by far the most productive. A few

innovations were supplied based on derivation pgiticularly vocalic infixation alone. Other

participants gave up on the task.

In the two comprehension tasks I and 4 however, most of the leamers answered at

least half of the items. Tt was revealed fiom the apparent raadomness of most responses that

leamers were only gucssing, and b t they were not tàmiliar with the derivational system of

MSA and its c o d n t s .

With ngard to the background questionnaire, the overall findings showed that the

questions in the questionaairc wete clcar to the leamers-

At the end of the piioting, the participants were asked for their fetdback and to

explain why they did not amplete the tasks. They respondcd that the Mcabulary was

diicuit and hard to understand, and that the fonnat was cunfiising. They also responded

that the tasks were too long and tirne-consuming. and the time set to perforrn al1 the tasks

was insuEcient.

In the light of these findings, the vocabulary knowledge test, and the three tasks were

modified, and simplified. The questionnaire, on the other hand, did not need any change or

modification, and remained basically intact in the main study,

The pilot version of the vocabulary knowledge test in Arabic included words fiom

the 2.000 to 10,000-word levels, and beside each word a phonetic transcription was noted to

facilitate the reading and understanding of the words. The test had to be redesigned and

simplified because it was difficult for these L2 leamers to complcte. I went lower than the

2,000-word level, and 1 eiiminated the 10,000-word level. The new version of the test

included 500-, 1,000-, 2,000-, 3,000- and 5,000-word levels. The phonetic transcription was

removed as students were not familiar with the phonetic symbols.

In the production task and comprehension tasks, most low fiequency words in the

stimuli were replaced by higher tiequency words. The glossary for the production task was

also modified. in addition, the production task was reduced to 24 questions fiom 30

questions, and the time to perfonn dl tasks was lengthened by adding another 30 minutes to

enable every participant to finish the tasks.

Preliminary piloting was carried out as well with Eve native speakers of Arabic

living in Canada (three maies and two femdes). These individuals came fiom three different

Arab countries; namely, Iraq, Lebanon, and MoroccorOCCO Four were between 21-24 years of

age, and oae was W e e n 25-30. They al1 spoke English or French besides Arabic, and they

aii had university educations. Tbcu length of residence in Canada ranged between 7-22

Y--

Contrary to the L2 leamers, the native speakers' performance showed that tbe tasks

did not seem to be difficult for them to perfonn. Al1 five native speakers cornpieteci the pilot

tasks. In the production task, they provided various lexical innovations incIuding

derivational patterns and compounding constructions. The former, however, were wideiy

favoured. In the two comprehension tasks, derivational patterns were also most 0 t h chosen

for each semantic notion.

4.5 Data CoUection of the Main Study

4.5.1 Task Oder

In the main study, the tasks were performed in the Arabic ciassroom, and were

administered by me. Each participant received a package compiled in the following order:

The vocabulary knowledge test, the production ta& the comprehension task 1 (based on

invented items), the comprehension task II (based on real words), and the background

questionnaire. With respect to the word formation tasks, the participants were infonned that

there was no "right" or "wrong"answer, that they were not being "tested" in the traditional

sense of the word, and that theü responses would provide valuable infamation on leamers'

use and choice of word formation devices. It was also made clear that in any reporthg of

fmdingq their confidentiality was ensured. They were told that they had one and half hwn

to complete the tasks, and that thcy had to work individually and could work at theu own

pace. They were also encouaged not to lave blanks and respond ta al1 items. In general,

participants were able to finish aii tasks in one hour and meen minutes. As they cornplad

al1 tasks, they retumed tkm to me where they were immediately placed h a box

4.5.2 Seoring

Only the vocabulary knowledge test was scored for accuracy. To score this test, one

point was awarded to each correct answer matching a word to its meaning. This yielded a

maximum of 9 points per kquency level, and of 45 points for the whole test. Nation (199û),

in his guidelines on the use of his test as an evaiuation instrument, suggests that a score of

12 or below a! a particular level (out of a maximum of 18) indicates that the Ieamer needs

M e r work et that particular lexical level. In detennining the launers' level of vocabulary

knowledge in the present study, an adjustment was made to the scoring because the

maximum score at each level was 9 points instead of 18. Therefore, leamers were

categorized as higher, mici, or lower on the basis of the scores obtained in combining ievels.

Leamers who scored less than 15 points for levels 1, 2 and 3 wen categorized as low

scorers. Learaers in this category did not answer comectly any question in levels 4 and 5 ,

Lamers who scored behkreen 15-26, at Ievels 1, and 2 and up to 26 points for al1 levets

combined were considered mid scorers, and those who scored 27 and up for al1 leveh

combined were considerd to be high scorers. These latter were at the 3000-word level.

Most of them obtaincd 9 points out of 9 for the first level and 6 points out of 9 for each of

levels 2, 3, 4, and 5. B a d on this categorization, there were 10 lower, 19 mi4 and IS

higher scorers (sec Table 4 6 for the scoring).

Table 4:6

L2 leaniers' scores on the vocabulary knowledge test (m=44)

Lower scorers (N=10)

Mid scorers (N= 19)

L Higher scorers (N=15)

les than 15 points at levets 42 and 3 combined

1 leamer scored 5

9 leamers scored between 7- 14

At least 15 points a? levels 1 and 2 cornbined, and up to 26

points inchding 1 to 5 combined

9 learners scored between 15-20

10 learners scored between 21-26

27 points & up at levels 1 , 2,3,4 and 5 combined

2 learners scored between 27-30

12 leamers scored between 3 1-40

1 leamer scored 43

4.5.3 Coding the data

In coding the data for the production tasic, the participants' responses were grouped

into three main categories:

- Categoy 1: lexical innovations. This category includes derived nominal and verbal patterns as weU as wmpounding a d o r periphrastic constructions.

- Catego y 2: suppletivcs, that is cxisting or r a i words supplied by padcipants instead of lexical innovations.

- Catcgory 3: non-responses where blanks wne left by pamcipants. This categoty also included cases whm the whole input question was repeated. There were a few instances in the data where L2 lamers answered by repeating al1 words bolded in the question.

in order to investigate the participants' use of word formation processes, only

category 1 (lexical innovations) was consideral. Lexical innovations were classified into

four subcategones for the four word formation processes: vocalic infmation alone, vocalic

infixation and gemination of th 2* consonant, aftixatiaq and compoundinglar periphrastic

constructions.

-Subcrtegory Ir: derived patterns coimi through vocalic infwation alone.

-Subertegoy ib: derived patterns formed via vocalic infixation and gemination of the 2* consonant.

-Subcrtegory lc: derived patterns b a d on affixation.

-Subategory Id: compounding consmictions, namely N-in-constnict, N+N, AP+N, N+N+N, N+prep+N, and V+N, as well as periphrastic consmictions that consist of combining the periphrastic verbs /jaTaid, /sabbabd,or /dafaTa/ with either pronoun, adjective or preposition, and noun.

For comprehension tasks i and II, participants' responses were tallied based on the

alternatives given on the two tasks that repmted the four word formation options

investigated in this study for each semantic notion. Table 4:7 displays the four alternatives

for each semantic notion.

Tabk 4:7

Alternatives provided in comprehension tasks 1 and II for each semantic notion

2/P attem CaCC&C(a)

Vocalic infixation done

Vocalic infixation and gernination of the 2* consonant

Compounding

Vocalk infxation and gemination of the Zrn consortant 1 fixation I Compounding 1 Vocalic infixation alone

Vocalic iafixation and gemination of the 2" consonant 1 tifkation

Compounding

Vocalic infixation alone

V d i c infixation and geminatioa of the 2"' consonant 1

4.6 Data Annlysis

In order to investigate the learners' and native speakers' production of lexicai

innovations, and their use and choice ofword formation processes in MSA, both quantitative

and qualitative data analyses were c h e d out.

46.1 Quantitative Analyses

The quantitative analyses of the data dlow us to draw tentative conclusions on the

L2 learners' preferences in MSA word formation An initiai quantitative analysis was

carried out with the data fiom the production task in order to compare L2 lemers' and

native speakers' use of imovations, as opposed to theu use of suppletives and non-

responses. Further analyses were carrieci out on the production task, and both

comprehension tasks 1 and 11 to compare L2 leamen' and native speakers' use and choice of

word formation processes according to the semantic notions investigated (agency,

instnimentality, location, and causativity). Another analysis was a graup compariscn among

the lower, mid, and higher level L2 leamers to see if the level of L2 vocabulary knowledge

revealed diffcrential effécts for the use of MSA word formation devices. Fidly, an analysis

was canied out to compare higher level L2 leamers' and native speakers' use and choice of

word formation processes to se if the former were able to make creative use of word

formation processes that was comparable to those of native speakers.

4.62 Quiilitive Anrlyscs

Qualitative anaiyses were carried out to examine the leamen' production of

innovations and their serategks and pattenu in using MSA word formation processes, and to

assess the applicability of the principles of productivity, semantic transparency, fornial

simplicity, and conventiodity in expiaking the leamers' use of different word formation

processes for each semantic notion. The L2 leamers' use of word formation processes was

aiso examined in order to find out if the LI had any effect on the choice and use of these

processes in the L2.

The following chapter presents the quantitative findings.

Chapter 5 reports the results of the quantitative analyses. It is organized as follows:

First, statistical procedures used to analyse the data collectai by means of the production

ta& comprehension task I (based on invented items), and comprehmsion task U (based on

real items) are describecl, Secondly, a quantitative cornparison between L2 leamers' and

native speakers' responses in the production task is presented, followed by more detaiIed

quantitative cornpuisons between L2 leamers' and native speakers' use and choice of word

formation processes in coining i~ovations. Nexf, the lower, mid, and higher level L2

leamas' production of MSA word formation processes is compared, Finaily, cornparisons

are made between higher b e l L2 leamers and native speakers and betwm lower level L2

leamen and native speakers with respect to use and choice of word formation processes on

the three tasks.

5.1 Sîatistiul Proceduru

The computer program for statistical data analysis, SPSS for Wùidows, was used in

di d y s e s , L2 lesniers' and native speakers' responses on the three tasks (productioq and

comprehension 1 & II) were wded inthe manner deiineated in Chapter 4 and means for their

respoases were calcuIazed.

T tests were then use& (a) ta compare the overall number of innovations, as opposed -

to suppletives and non-responses, pmvided by the leanrets and native speakers; and (b) to

compare these types of responses for each semantic notion (agency, instmrnentaiii,

locatioq and causativity).

T tests were also used to compare the processes each group used as a percemage of -

their innovations on the production task. The same anaiysis was conducted for

comprehension tasks 1 and II where percentages of the L2 leamers' and native speakers'

overail responses were calculateci and cornparcd for each word formation option in the two

tests.

The cornparison of the lower, rnid, and higher level L2 leamen' use of word

formation processes on the production task was conducted using one way-analysis of

variance (ANOVA). A pst-hoc anaiysis (the Tamhane test) was used to determine where

any significant differences lay.

Group cornparisons using 1 tests again were alsa c&ed out between higher level L2

leaniers and native speakers, and between lower lewel L2 leamers and native speakers for

the three tasks. In every case, for a merence to be considemi statisticdiy sig@cant, the

level ofsignificance was set at ps.05 for al1 tests.

5.2 Cornparison of U LAirncn' and Native Speaken' Responau on the Roductioa Twk

nie first hypathesis la predicts that there will be a quantitative difference b e n

L2 teamen and native speakers in thaî L2 leamers will provide fewer lexical innovations

and give more fiequent non-responses. To test this hypothcsiq learners' overall responses

as well as theû responses for each scmantic notion on the production task were compared

statistically via t tests with those of the native speakers. Responses were grouped h o three

categories (innovations, suppletives, and non-responses), and the means were caIculated and

compared. Tables 5:L and 5 2 display the means, standard deviations, and level of

significance ofL2 Iearner versus native speaker differences on the production task

Ta blt 5: 1

Overali cornparison between L2 leamers' and native speakers' responses on the production task (max.=24)

The overall results in Tabie 5:l show statisticaily significant differences between the

two groups for suppletives and non-responses. Native speakers resorted more oflen to

suppletiveq and L2 leamers sometimes gave non-responses, while no native speakers did so.

The lexical innovations category had the higàest means for both goups compated to

suppletives and non-responses, and the 1 test did not reveal any significant difference

between the two groups with respect to this category.

As shown in Table J:2, the comparative analysis of the two groups' responses for

each semantic notion (agents, instruments, locatives, and causatives) rwealed the

followiag trends: With respect to the innovation category. the differencc was

statisticdly significant between L2 learners and aative speakers for agents, iastniments

and lmtives, Native speakers produced more innovations than L2 leamen for a g a s and

ins~rments. However, for locatives, L2 leamers produced significantly more innovations

(M=3.77) than native speakers &f=2.97, p=.011). W i regard to suppletives, the

Innovations

Suppletives

Non-ruponsa

- P ns.

-000

Lzkrrntn

M - 17.18

1.22

- t value 1.34

8.73

SI 2.60

1.99

5.60

- M 18.32

5.68

5.70

SI 4.94

2.60

0.00 .O0 -6.50 .O00

difference was statistically significant for instruments, locatives, and causatives, with

native speakers resorting to significantly more suppletives than L2 leamers for these

notions. For locatives and causatives, native speakers' mean use for suppletives was much

higher than the L2 leamers' (3 ,O3 compared to .7 1 for locatives, and 1-78 compared to 29

for causatives). For the non-response category, the difference was statistically significant

for al1 four notions. L2 leamers left some blanks for al1 four notions while native speakers

Table 5:2

1 tests cornparhg L2 leamers' and native speakers' responses for each semantic notion on the production task

5.2.1 Summa y

The above findings in the production task partially support the hypothesis la. AS

predicted, L2 learners supplieci kwer innovations than native speakers for agents and

instruments, and udike native speakers gave some non-responses for al1 four semantic

notions. Contrary to predictions, howwer, L2 learnen p r o d u d more innovations than

native speakers for locatives, and for causatives their mean was alrnost equal to, and n a

significantly différent fiorn, that of native speakers.

5.3 Cornparison of L2 Lcrirnen' and Native Speakers' Use and Cboict o f Word Formation Procmrcs

According to hypothesis 2:

L2 leamers will tend to produce and choose compounding andlor periphrastic constructions more often than derivational processes in naming agency, instrumentality, location, and causativity.

To test this hypothesis, L2 learners' and native speakers' innovations in the

production task were broken d o m into percentage use of each process for each semantic

notion and the two groups' responses were compared using t tests. The same analysis was

conducted for coniprehension tasks 1 and II where L2 learners' and native speakers'

percentage choice of each word formation option in laklling invented items and existing

items was compared.

5.3.1 Cornpuison of U Leamen' and Native Spcr ih ' Use of Word Formation Proecssa on the Production Twk

In the production task, only the category of Iexicai innovations was considered in

orda to compare the Icamers' anci native speakers' use ofword fimation processes in

coining new words. innovations wwe broken down into four processes: 1)vocalic infixation

alone (henceforth Vi), 2)vocaiic infixation and gemination of the 2* consonant

(henceforîh Vinf.+gem.), 3)affixation (henceforth Affix), and 4)compound or periphrastic

constnictions(henceforth Comp.Const./Periph.Fm.). Tables $3 and 5 4 display the mean

percentages, srandard deviations, and level of significant differences between L2 leamers

and native speakers.

Statistically significant differences between the two groups were revealed for al! four

processes, L2 leamers opted much more tiequently for compounding constructions to coin

their innovations while native speakers most often used derivational processes, namely

Vinf.+gem. and affixation (see Table 5:3). However, L2 leamers tended to use ViM more

fiequently than native speakers.

Table 5:3

Overail cornparison between L2 leamers' and native speakers' mean percentage use of word formation processes on the production task (max.=24)

- P

.O00

.O00

.O00

.O00

M% = Mean paccntage Vint = Vocaiic infîxa1ion donc VÏ+gem = ~ocalicinfixadonandgcniiaatiaaofk2"dco1~~ mx. = ~ - 0 1 1 comp*const, = companding consmictim MiphFm =Faipbrasticfoms

l2 karEf!m.,

Vinf. , Vinf.+gem. A f f î Comp.Consb/ Ptripb.Fm.

_t value

-3.80 10.24 18.59

-1 1.39

,., Nakt spdtim . ' p4!Q

16.86 19.63 1.25

62.25

6.55 50.71 30.82 1 1.92

se 17.14 15.05 2.42 28.08

SI

5.27 12.72 9.79 8 .O0

Differences also emerged in L2 leamers' and native speakers' use of word

formation processes in coining each semantic notion (see Table 54).

Table 5 4

L2 leamers' and native speakers' mean percentage use of word formation processes for each semantic notion on the production task

For agents, t tests revealed signifiant differences between L2 leamers and native

speakers for compounding Vinf.+gern., and afturation. L2 leamers otten used

compounding w h k native speakers did not use it at dl. Cnstead, native speakets were

significaatiy more Uely to use Vinf.+gem. Native speakers also opted some of the time for

the fiation process whereas L2 lemers did not use it at dl. ihe two groups did not

difier in their use of Vinf

For instruments, t tests showed significant differences fÔr aii processes. L2 learners

used compounding more often than native speakers who were most likely to use

Vinf++gem. A h , L2 iearnms used Vinf while native speakers did not. Instead, native

speakers sometimes used affixation, while no leamcrs did sa.

Similarly for locatives, r. tests revealed significant differences for al1 four processes.

L2 learners optai for compounding much more often t h native speakers who opted

massively for afixation. L2 leamers a h uscd Vinf.+gem. and Vinf. which native speakers

did not use at dl.

Finally, for causatives, f tests again revealed significant différences for ail fwr

processes. L2 learners resorted much more fiequently to periphrastic constructions whereas

native speakers opted most commoniy for VinE+gem. Native speakers a1so opted for

affixation more &en than L2 learners. These latter made some use of Vinf. which was not

produced by native speakers.

53.2 Summa y

The above findings on the pduction task wnfh hypothesis 2, since L2 lemers

used compodng significantly more ofken than native speakers in coining the three

nominal notions (agents, instments, locatives), and periphrastic cunstnictions sisnificantiy

more often for causatives. Natin speakers, on the otha ùand, d derivationai processes

significady more often. Vii+gcm. was the option the native speakers used most

commoniy for agents instninmentq and causatives; for locatives, it was afbtion.

5.3.3 Cornparison of L2 Leamers' and Native Speakers9 Choicc of Word Formation Options on Comprehension Tisk 1 (Invcated Items) and Comprehension Task U (Red Items)

The comparison of L2 leamers' and native speakers' overall choice of word

fomation options across the four semantic notions on the two comprehension tasks (se

Tables 55 , and 56) also showed a high mean percentage choice of compounding and

periphrastic fonns by L2 leamers and a high mean percentage choice of derivational

processes by native speakers. On both tests, simcant différences were observeci for al1

four processes. Arnoag derivational processes, native speakers chose most commonly

VinE+gem. and affixation. Vinf., however, was selected more often by the L2 lemers than

by the native speakers.

T i b k 5:s

Overall cornparison between L2 lmers' and native speakers' mean percentage choice of word formation options on comprehension task 1 (invented items) (max.=30)

1 .O15

-000

.O00

Wh- c* 1,: . . , ,,

Comp.Con~eripL.Fm. 1 59.01 1 20.25 1 19.67 1 8.53 1 -1 1.79 1 -000 1

Vinf.

Vinf.%em.

Amx.

z %

8.33

42.08

29.92

M !h

13.33

27.65

-00

SD 12.53

21-10

-00

SZ)

3.92

7.98

8.0 1

- t value

-2.51

4.22

23.63

Table 5 6

L2 leamen' and native speakers' overail mean percentage choice of word formation options on comprehension task il (real items) (max.=30)

1 I l

Vinf. 1 18.01 1 8.79 1 3.48 1 2.21 1 -10.61 1 -000

With respect to the choice of word formation options for each semantic notion,

significant differences were reveaied benveen L2 Iearners and native speakers for al1 options

for each semantic notion in both comprehension tasks.

in comprehension task I (Table 57). L2 leamers selected compounding most oflen to

name inventcd agents, insmimcntq and locatives, while native speakers most often chose

Vinf.+gem. for agents and instruments, and aftïxation for locatives. A h , for causatives, L2

lemers selected the @phrastic fom (/jaTald+Ro+Adj.) more oAw than native speakers

who wcre again more likely to select Vid.+gem Uniike native speakers, L2 leamers did not

select' affixatioa at al1 to aame Uivented and real agents, instruments+ locatives, and

causatives. Instead, they selected V I more o h than native speakers for invented agents,

and for inventai instruments, locatives, and causatives, they sometimes selected Vinf

whereas anive speak& did not do ro at dl.

Afiix. .O0

Comp.Consî./Periph.Fm. ) 5 1 .O 1

.O0

14.65

37.22

16.16

9.00

9.83

26.17 .O00

-12.90 .O00

Tabk 5 7

Mean percentage of L2 Iearnen' and native speakers' choice of word formation options for each semarrtic notion on comprehension task 1 (invented items)

As in comprehension task i, there were significant differences on al1 processes in

comprehension task II. L2 leamers also selected cornpouuding most commonly for red

, 'Ukuii#r

instruments and Iocatives in comprehension task II (see Table 5:s). Only for real agas did

M - - t value

: A G m S (mui=â& CàCiC(a)

CaCCZC(a)

muCaCCiC(a)

N-in-constnict

; mm- [mu*) '

CiCiC(a)

CaCCaC(a)

miCCaCa

N-in-constmct

i mmf~*--~ CàCiC(a)

CaCCZC(a)

maCCaCa

N-in-construct

- SD

. , . wrfhrtspcrrkicrs. , .

- M - SD

4.26

24.71

.O0

71.02

[ CAVWTIVES [ILIL*- - , ,

W a C a

CaCCàCa

7 aCCaCa

Ja%la+Pro+Adj.

19.80

43.34

.O0

45.33

3 1 .25

47.19

21.56

.O0

.O01

.O00

.O00

.O00

6.10

27.53

.O0

66.37

-000

.O06

-000

.O00

-6.52

2.84

12.22

4-60

14.71

15.11

14.43

.O0

9.06

26.64

.O0

31.11

30.52

29.22

.O0

40.26

.OU7

.O00

.O00

.O23

-00

50.7 1

35.36

13 -93

3 1.05

45.06

.O0

35.83

7.13

3.23

9.45

-10.39

.O0

66.27

9.00

24.73 -

15.91

28.59

.O0

55.50

.O0

20.95

18.30

11-88

.O00

.O02

.O00

.O00

.O0

.O0

57.82

42.17

37.00

25.20

.O0

32.76

.O0

22.64

10.91

20.4 1

-3.74

7.19

5.23

-7.30

-00

.O0

18.12

18.11

-2.85

-7.52

20.19

-2.33

they, like native speakers, select the VX+gem. option more fiequently than the other

processes. For agents, Vinf was dso selected more often than compounding by L2 learners.

For instruments, locatives, and causatives, they sometimes chose Vinf while native speakers

did not so at aiî. As in comprehension task 1, aflixation was not chosen at al1 by L2 learners.

Table 5:8

T tests compating L2 learners' and native speakers' mean percentage choice of word - formation options for each semantic notion on comprehension task II ( r d items)

I I

, AGEiYT!F (mm*) CZCiC(a) 27.27 16.25 13.93 8.85 -4.73 -000 CaCCâC(a) 47.08 24.57 75.36 11.66 6.84 -000 muCaCCiC(a) .O0 .O0 10.71 12.00 5.65 -000 N-in-construct 25.65 23.78 .O0 .O0 -7.15 .O00

_MSTRtJMEWB {am.=@ CâCiC(a) 4.83 9.03 .O0 .O0 r -3.55 .O01 CaCCaC(a) 32.24 25.71 5 1.56 17.72 3.83 .O00 miCCaCa .O0 .O0 - 17.50 13.22 8.37 .O00 N-in-construct 61.64 28.97 30.94 22.46 -5.27 .O00

, W ) C A T I V E S ~ ~ * - - CâCiC(a) 10.39 10.38 .O0 -00 -6.64 .O00 CaCCiC(a) 29.22 24.44 .O0 .O0 -7.93 .O0

. maCCaCa .O0 -00 82.86 19.20 . 27.30 .O00 N-in-construct 60.39 . 25.93 17.14 . 19.20 . -8.73 . .O00

. CAUSATfVES CXaCa 29.54 20.01 .O0 -00 -9.79 -000 CaCCâCa 14.20 13.64 45.62 2 1.47 7.91 .O00 'lacCaCa -00 .O0 37.81 24.10 9.92 .O00 ja%da+Pro+Adj, 56.25 18.97 16.56 15.60 -10.51 .O00

5.3.4 Summa y

To sum up, statistically significant findings on both comprehension tasks

consistently support hypothesis 2. L2 leamers showed a high preference for compounding

over derivational processes to name the nominal notions, and for periphrastic constructions to

name causatives. Native speakers, on the other hami, exhibiteci a high preference for

Vidtgem. for agents , instruments and causatives, and affixation for locatives.

5.4 Cornparison bttween Lower, Mid, and Bigber Lcvel l.#arnen' Use of Word Formation Processes in the Production Twk

Quantitative group cornparisons were -ed out among leamers of different levels

of vocabulary know!edge (as measured by the vocabulary knowledge test) to find out whether

relative level of vocabulary knowledge in the target language would influence the use of L2

leamen' word formation processes, in keeping with the expectation that leamers with a lower

level of vocabulary knowledge will find it harda to make crcative use of word formation

processes than higher level lemers (hypothesis 3).

To test this third hypothesis, an ANOVA was wnducted to compare lower, mid, and

higher level Iearnm' overaii use of word formation processes in the production task then

their use of word formation processes for each smantic notion.

The ANOVA reveals a main effect of vocabulary knowledgc on the overall use of

each of the four pmcesses: Vinf., Vinf,+gem., affixation, and compounding (see Table S:9).

Lower level leamers produced more oompounding than mid and higher level

leasners. H i g h level leamers opted more ofkn for Vinf.+gem., Vinf,, and affixation than

did mid and lower level lemers. The Tamhane test indicated that the differences in means

were statistically significant between lower and higher (F.020) and mid and higher levei

leamers (pC.004) for Vinf., between lower and mid (g<.015), and lower and higher level

iemers (p<.000) fbr VinE+gem., between lower and higher (ec.018) for *xation, and

between mid and higher (gC.023) for compounding.

Table 5:9

Lower, mid, and higher level learners' overall use of word formation processes on the production task

The cornparison between lower, mid, and higher level leafners' use of word

formation proccsses for each semantic notion revealed the foilowing results.

1

For agents, t h was an overail main effect of vocabulary knowledge on the use of

M - - M

Vinf. 1.90 1.79

Vinttgem. .80 2.95

Affir, -00 .16

Comp.Const./Peripb.Fm. 11.50 1 1.42

f

Vinf, +gem. only. The means fm use of this process tended to increase with the level of

, FIS' I M

5.40

6.80

-53

7.53

P - 9.89

25.90

4.94

4.09

vocabulaty knowledge . For example, the lower Iwel Ieaniers' mean for Vinf.+gem. was

P I

-000

-000

,012

.O24

-40, increasing to 1.58 by mid level leamers, and to 2.87 for higher level leamers (se Table

5:lO). The Tamhane test indicated that for Vd+gem, the ciifference in means was

statistically signifiant between lower and mid, (0<.049), lower and higher w.000)

lower level learners produced Vi-t-gem significantly less often than mid level leamers and

higher b e l learners. The difference in means for Vinf.+gem. was also significant between

mid and higher level leamers 6 0 5 0 ) where mid and lower level leamers produced this

process Iess often than higher level leamers.

Table 5:10 '

Means of lower, mid, and higher level learners' use of word fonnation processes for each semantic notion on the production task

For insaumenîs, there was an o v d 1 main effecî of vocabulary knowledge on

@ormance groups for two of the p m c a s a : Vi, and compoundil~gg Lower level Icrnms

used compounding more ofken than mid and higher level leamers whereas higher level

leamers used Vinf. more often than mid and lower level learners. ïhc Tamhane test revealed

a statistically significant difference in means between lower and higher level leamers

(p<.005), and between mid and higher level learners (pc.015) for Vinf. For compounding,

the mean diffaence was significant between lower and higher level learners: where lower

level leamers used compounding more than higher level leamers (pC.023).

For locatives, the Anova showed a main effect for Vinf.+gem., and for

compounding. Highec level leaners used the Vinf+gem. process more than mid and Iower

level Iearners. Compounding was used more oflen by mid level leamers than by either lower

or higher level leamers. The Tanhane test showed a significant ciifference in means b e e n

lower and higher level learners wiîh (p<.000), anci between rnid and h i g k Ievel leamers

with (p<.ûûû) for W t g e m . For compounding, the mean différence was oniy significant

between mid and higher level leamers: mid level learners used compounding more than

higher level leamers (p<.Oû9).

Finally, for causatives, there was an o v d main e f k t for level of vocabulary

knowtcdge with respect to Vinf. and afkaiion processes. Higher level leamers used Vinf

more ofim than mid and lower leamers. Higher level leamers also used m a t i o n while mid

and lower lenl leamers did not. The Tamhane test showed the difference in means to be

statistically signifiant between lower and higher level learwrs for V i with (F.003). and

between lowcr and higher level leamers for af%tation with (pC.034).

5.41 Summary

To sum up, hypothesis 3 was partially borne out by the cornparison of the use of the

various word formation processes by L2 learners at lower, mid, and higher Ievels of

vocabuiary knowledge. Level of vocabulary knowledge revealed differential effms some of

the time: e.g., with respect to Vinf for coining instruments and causatives, Vi+gem. for

coining agents and locatives, affixation for coining causatives, and compounding for coining

instruments and locatives. For the three processes (Vid, VinC.+gem., and affixation), higher

level leamers tended to opt for them more often than mid and lower level learners. Only for

compounding did higher level lemers tend to use it less ofien than mid and lower Ievel

lemers. The significant différences in means for these processes were mostly bmeen

higher level learners and the other two pups .

5.5 Cornparison of Bigber k e l , Lower Lcvd L2 kanien ' and Native Spnken' Usc and Choicc of Wotd Formation Proeuses

The final quantitative analysis of data coiiected in t his study addresses hypothesis 4

wfuch States that:

Hi& level L2 leamas' praduction and choice of word formation processes w i l be more like those of native Arabic speakers than lower level leamers' will.

To test thW hypothesis, higher and lower level Iclltners' use of word formation

processes was cornparcd to that of native speakers on the production task and on

cornprehension tasks L and U.

5.5.1 Cornparisan of Highcr, and h w c r k c t L2 Lnmtrs' and Native Speakers' Use of WoFd Formition Roeusu on the Production Task

1 tests performed on higher level leamers' and native speakers' production of the

four word formation processes to name the four semantic notions rwealed that higher level

L2 leamers diiered significantly h m native speakers in their use of W. for instruments,

îucatives, and causatives: L2 leanms relied sorne of the time on this process in mining their

innovations while native speakers did not (see Table 5: 11). Ody with respect to agents were

the means for Vinf simi1ar for the higher level leamers and the native speakers. Another

significam diffcrence was revecrled in the use of Vinf.+gem. to coin instruments and

locatives: for locatives, higher Ievei L2 leamers used it and native speakers did net. In

wntrast, for insûuments, native speakers used Vinf +gem. more oflen than the leamers.

The higher level L2 lemers and native speakers also differed in their use of

affixation for agas, instruments, locatives, and causatives: the iearncrs used affixation less

ofien t h native speakers for locatives and csusmtiveq ad for agents and instruments, they

did not use it ai dl.

In addition, hi@ level leamers and native speakers differed in their use of

compounding for agas and locatives: for agents, the Icarwrs u d wrnpounding and

native speakers did not; and for locatives L2 leamers used this process more 0th than

native speakers. in the use by h i g k level kamers of word formation processes was

comparable to that of native speakers only for Vinf. and Vinf,+gem. ptoctssts to coin

agents, and for compouading to «lin instruments.

Table 5:ll

Means of higher level leamers' and native speakers' use of word formation processes for each semantic notion on the production task

AGENTStmr.** Vinf. Vinf.+gem. MIX. Comp.Const.

;rnSTRmNTs(~~l l t .*" Vinf. Vinf.+gem.

c AEx. Cornp.Const.

;W)C"ATLVEScpuf.** V i d Vinf.+gem. AfFix. Camp-Const.

; CAuSATfVES(ii.16)* Vinf. Vinf.+gem. AtTrx. PenphIm.

I 1 I I 1

.87 2.87 -00

2.07

.79 1-76 5.80 -6.00

1.41 1 -46 .O0

1.33

1.17 3.63

.96

.O0

ns. ns. .O00 .O00

1.87 1-40 .O0

2.00

.O00

.O00

.O00 ,000

.90 1.3 1 1.04 -00

-5.85 5.20 4.95 -4.57

1.93 -73 .27

2.13

1.77 1-50 .O0 2,03

.O0 2.62 1.42 -17

1.27 1 .O3 -46

1.64

-4.09 3.63 5.34 -.49

.O0 1.56 1.28 -3 8

.O01

.O01

.O00 n.s.

.O0 3 .O0 .65

1.73

-2.75 -73 1.80 .27

1.33

.O0 1.32 .77

1 .30

-016 .O0 .O0

2.67 .30

1 .O3 1.15 -46

1-63

.O0 -00

1.21 -52

-6.08 ; .O00 10.73 -2.40

.O00

.O29

1 tests performed on lower level leamers' and native speakers' use of the four word

formation processes to name the four semantic notions showed that lower ievei leamers

also differed significantly from native speakers in their production of word formation

processes. With respect to Vinftgem, lower level L2 leamers opted less often for it than

native speakers in coining innovative agents, instmments; and for causatives, Iowa L2

leamers did not opt for this process whiie native speakers did so (see Table 5: 12).

Table 5: 12

T-tests comparing lower Ievel learners' and native speakers' use of word formation piocesses for each semantic notion on the proûuction t&

Anotber sigriificant difference was sbown in the use of &kation in coining agents,

instrumeuts, locatives, and causatives: for the four semantic notions, lower level L2 leamers

did not use it and native speakers did so. A tùrther signifiant difference was revealed in the

use of compounding for agents, instruments, and locatives, and of priphrastic constructions

for causatives: for instruments, and locatives, lower-bel L2 learners used compounding

more often than native speakers, and for agents lower leamers did use compounding and

native speakers did not. Also, for causatives, lower leamers used periphrastic consmictions

more 0 t h than native speakers.

Thus, the findings in the production task did not support hypothesis 4. Higher level

leamen as well as lower level tearaers did differ in most cases fiom native speakers in their

use of word formation processes.

5.5.2 Cornparison of Higher and Lawtr Lcvei L2 karuen' and Native Speakers' Choiec of Word Formation Options on Compnhensiaa Twk I (Inventcd Items)

The cornparison of the higher level leamers' word formation options with those of

native speakers on comprehension task 1 (see Table 5:13) revealed tint the two groups

differed significantly in their choice of Vinf. to name agents, locatives, and causatives,

where higher level teamers did not chwse this process at dl for agents, but for bath

locatives and causatives they selected it while native speakers did not choosc it ail.

Both groups also differ in their choice of Vinf.+gem. to denote agerrts instruments,

and locatives with higher level L2 leamers sclecting Vinf.+gem. Iess o h than native

speakers for agents and instruments; for locatives, oa the othcr hand, higher L2 leamas

chose it while native speakers did not. A signif~cant difference was also found between the

two groups on the choice of aftutation for ai1 four semantic notions: the bigha lewl L2

lemers, unlike native speakers, did not choose this process at dl. No significant

differences were found on higher fevel Leamers' and native speakers' choice of the Vinf.

process to coin instruments, of compounding to coin locatives, or of Vinf.+gem. to coin

causatives.

Table 5: 13

Comparison between higher Ievel L2 leamers' and native speakers' mean choice of word formation options for each semantic notion on comprehension task 1 (invented items)

in comparing lower leveI leamers' and native speakers' choice of word formation

options in the comprehension task i, it was found that lower level learners differed

significantly fiom native speakers in theu choice of compounding in naming inventai

agents, instruments, and locatives (see Table 5: 14).

For instruments and locatives, lower level leamers' mean of compounding was

significantly higher than native speaicers. For agents, however, loww level leamers often

chose compounding while native speakers did not at dl. The difference was also significant

between the two groups for afixation: Lower Ievel leamers did not select this process at al1

in naming the four semantic notions whereas native speakers did so.

Other significant differences were revealed for VinE, and Vinf.+gem. With respect to

Vinf., the significant difference was revealed for both groups in naming agents and

causatives: for the former, lower leamers did not choose Vinf. while native speakers did, and

for the latter, some lower Iearners chose it whereas native speakers did n a . With respect to

VinE+gem., a significant difference was shown bctween lower level lemers and native

speakers in naming instniments where lower level learners chose it less often than native

speakers, and in naming locatives where they sekcted it same of the time while native

speakers did not.

Table 5: 14

' t g ~ o t be compnui because the siandaid deriations of borh group~ arc 0.

Means of lower level L2 learners' and native speakers' choice of word formation options for each semantic notion on cornprehension task 1 (inventeci items)

5.5.3 Cornpuison of Higber, b e r k e l W haruers' and Native Spcrken' Choict of Options in Compnbcasion Twk II (Red Items)

AGENTS bu=û) CaC iC(a) CaCCiiC(a) muCaCCiC(a) N-in-constnict

; WSTRUkIENTS (pul4) CZCiC(a) CaCCàC(a) miCCaCa N-in-construct

LoCAmws(m8&* CiC iC(a) CaCCiX(a) rnaCCaCa N-in-coastruct

CAUSATIVES (aux.='?') CsiCaCa

In cornparhg the choice o f higher level learaers of word formation options with that

of native speakers to denote the four semantic notions on carnprchension task II, it was

. Lawert2 ïearnnr

piplra , , Native Spclikcn

- M - SD M -

ns. -000 n.s.

- SD

.O0 2.30 .O0

5.70

-1.32 -12.22

1.71

ACCXa ?aCCaCa Periph.Fm.

- t value 1

.O0 3.71 .O0

3.71

2.10 -00

2.40

2.50 3.77 1.72 .O0

n.s. -000 .O00 -000

3.38 -00

2.59

-40 1.60 .O0

6.00

.O0 1.83 .88

1.64

3.55 2.47 .97

.O00 n-S. - .O00 .O01

1.18 1.21 1-15 .O0

1.80 -5.30 -5.23 5.19

-70 2.0 1 .O0

2.3 1

1.47 1.28 -83

-13.43 -1.24 -9.45 4.85

.O0 5.30 -72

1.97

-- 5.22

-20.01 2.43

i

-00' .O0

1.28 1.28

- -001 .O00 .O32

.O0

.O0 4.05 2.95

.O0 2.70 .O0

4.30

.0Oa 1.64 -00

1 -64

.O15 3.00 2.50 2.63 .O0 .O0

found that only with respect to VinE+gem. was the higher level lemers' choice comparable

to that of native speakers' in narning instrwments (see Table 5: 15).

Tabk 5 1 5

T-tests comparing higher bvel L2 lemers' and native speakers' choice of word formation options for semantic notion on comprehension task II (real items)

For agents, higher level L2 leanins selected Vuif. more otken than native speakers,

and for instruments, locatives, and causathes, they sometimes chose Vinf while native

speakers did not. Higher level L2 Imers and native speakers a h différed in thek choice of

Vinf.+gem ta name agents, locaiives, and causatives. L2 leamers seleaed VinE+gem, more

0th than native speakers for agents; for Iocatives, they selected it while native speakers did

not, and for causatives, they selected it less ofken t h native speakers. Moreover, higher

level L2 leamers differed fiom native speakers by not choosing affixation at al1 to name the

four semantic notions. They also d i f f d in their choice of compounding to denote the four

semantic notions. For instruments, locatives, and causatives, they chose compounding more

oRen than native speakers. For agents, the higher learners chose it while native speakers did

not.

i n comparing lower level learners' and native speakers' choice of word formation

processes, findings show that there were two instances where the lower level learners were

not signiticantly different from native speakers. In these two instances, the similarity lay in

the use of V i d to coin instruments and locatives (see Table 5: 16).

Unlike native speakers, however, lower lewl L2 leamers selected Vin£ more often

than native speakers for agents, and for instruments, locatives, and causatives, they chose it

while native speakers did not. Also, lower level L2 learners selected Vinf.+gem. less often

than native speakers for agents, instruments, and causatives, but for locatives, they used it

more than native speakers. Unlike native speakers, they did not choose afixation at al1 to

name the four semantic notions. The Iower tevel learners aIso chose cornpounding

significantly more often than native speakers to name instruments, locatives, and causatives,

and for agents, they chose compounding while native speakers did not.

Table 5: id

Lower Ievel L2 leamers' and native speakers' choice of word formation options for each smantic notion on comprehension task II (real items)

Findings in both comprehension tash I and II also did not support hypothesis 4.

Higher leveI leamers' choice of word formation processes was not comparable to that of

native speakers. Lower level learners' choice of word formation processcs also did not

approach that of native speakers,

5.6 Summary of the Quantitative Findmgs

To sum up, the quantitative findings suggest partial support for hypothesis la. As

predicted, L2 learners produced fewer innovations for agents and instruments, and gave

some non-responses for al1 four semantic notions, whereas native speakers did not. But for

locatives, L2 learners produced more innovations for locatives than native speakers, and a

quite similar number of innovations as native speakers for causatives. However, complete

support was found for hypothesis 2. L2 lemers and native speakers exhibiteci different

preferences with respect to the use and choice of word formation processes. Partial support

was found for hypothesis 3 where Level of vocabulary knowledge did reveal différentia1

effects for a few word formation processes to dcnote some of the semantic notions. Finally,

hypothesis 4 was not supported in thaî highcr level L2 leamers' use and choice of word

formation processes did not resemblc that of the native speakers more closely than Iowa

level learners.

In the next chapter, a qualitative analysis of the findings will be presented to

investigate hypotheses 1 b, 5, and 6.

CHAPTER 6

QUALITATIVE F N I N G S

In this chapter, a qualitative analysis of the L2 leamers' and native speakers'

responses on the three tasks wili be provided. The purpose of this descriptive account is to

supplement the quantitative analyses, to examine more closeiy lemers' and native speakers'

lexicai innovations in the production task to assess the applicability of the acquisitiond

principks of productivity, sernantic transparency, formal simplicity. and conventionality in

explaining the Iearners' preferences for differem word formation processes, and to detemine

the eréct of L1 in the L2 lemers' production of innovations in MSA Samples of learner

and native speaker data are provided in most cases dong with the relevani test stimuli. In

each example, the opening part of the stimulus "in Arabic, how would you cal1 ..." for the

three nominal notions and "In Arabic, how would you expre S..." for causatives has been

deleteci since these are standard expressions in the production task.

6.1 Qualitative Eumination of L2 Learaen' and Native Spcrktn' Innovatians on the Production Tuk

According to hypothesis Ib, there wiU be a qualitative différence between L2

kaniers and native speakers in their production of innovations- A dose examination of

lemers' and native speakers' innovations on the production task reveds tbat leamers

exhiiited more vuiabüi i than native speakers. This variability is seen in their formation of

derived patterns, and in their production of a wide variety of innovative compound and

periphrastic constructions. Some of these innovations were semantically inappropriate in the

given context.

6.1.1 L2 Leamen' and Native Speakers' 1 unovations

Native speakers almost always coineci their innovations from a single verb conveying

the exact meaning of the verb presented in the input. For example, for agent item 14: "... a

person who draws with a pencil" the Arabic verb lrasamai 'to draw' was used to coin the

invented agent lrassam/, and for instrument item 21: "... a machine that cuts purnpkin", the

verb Iqa{aTai 'to cut', was used to coin the invented instruments IqattaTal and /q+iW. 6

Again, for locative item 15: "... a place where you burn clothes", only the verb haraqa/ that

canveys the meaning 'ta bum' was used by native speakers to constnict the innovative

locative lmaliraqal. On only one test item did two native speakers use other verbs than the

one presented in the input. This was when native speakers 12 and 24 used verbs that are

semantically related to, or extend the meaning specified by, the input verb. For agentive item

9: "... a person who spreads m o n by phone", instead of using the verb Inasharal which

means 'to spread', native speaker 12 suppIied the innovative agent IraWWajf using the verb

Idjal 'to circulate', and native speaker 24 /mu~ayyiF/ used the verb Ra@W 'to broadcast'.

In cornparison, L2 leamers, besides coining innovations from a verb conveying the

exact meaning of the verb presented in the input, also coined innovations for some items by

using other verbs semantically unnlated or only distantly related to the verb in the input. For

example, in respondiing to the agentive item 19: Y... a persan who pulls wagons", some L2

lemers gave /sUiq/ or / s a d 4 / using the verb l d g d 'to drive', and lqa?id/ or f q a d d

using the verb / q a U 'to lead' that are different in meaning fiom the verb /jarrd 'to pull'

presented in the input. Aiso in respoase to the instrumental item 23: "...a machine that stops

wind", some leamers derived their innovative instruments fiom the verb majazd 'to seize',

supplying Riàjizal or /najjaza/, and fiom the verb /manaTa/ 'to forbid' they derivecl the

innovative instruments /maniTa/ or /mannafa/ instead of /*ifal, /waqqW, or lmiwqafal

h m the verb Itawaqqafd (which signifies more precisely 'to stop' as provided in the input).

in addition, L2 learners occasiodly coined innovations kom a phonologically

similar but semantically unrelated verb. For instance, for agent item 19: "... a person who

pulls wagonsn leamer 6 gave the agentive pattern /jardyJ ' m e r ' detived fiom the verb

/jar@ meaning 'to run' instead of /jar& 'pullet' derived tiom the verb /jarra/ 'to pull'.

mer innovations were also coined by L2 leamers, but not native speakers, by raot

transformation. This means replacing one element of the mot, particularly a third consonant,

by a semi vowel /y/ in the derived pattern. Examples are /fajjây/ and harnrnâü instead of

/fajj&/ and Ihallaml Born the verbs J f a j j d and maJimal respectively by learners 7, and 14.

In addition, L2 leamers, unlike native speakers, coined innovations by metathesis which

consists of chging the order of the consonantal rock For exampie, the innovative

instrument / j a M z d h m the verb I d a j a l 'to disturb', instead of /=naja/, and th

innovative agent /farréiijl h m the verb i f a j j d 'to burst', instead of ifhjjarl by learners 19

and 26.

Further cases where leamers differed fiorn native speakers inchde imovations

coined fiom r a s that have rt basic nominal core. Sume notable exampies of these

innovations by leamers were found for agentive, instrumentai, and locative notions.

Examples include agent item 19, where the agentive noun /Fatdû/ was coined fiom the noun

Aarabid 'wagon'; instrumental item 16 where the instrumental noun lrabbâdal was

constructeci fiom îhe noun Izabdal 'butter'; and locative items 6 and 7 where the locative

nouns /malbanai and fmarbanid were coined from the nouns [laban] 'rnilk', and /murabbà/

'jam' respectively by differeut leamers.

To summarite, the above examination of leamers' and native speakers' innovations

on the production task reveals that L2 leamers, unlike native speakers, manifested more

variability in wining their innovations by using vwbs that are distantly related or

semantically unrelated to the verbs in tbe input, by transfotming or changing the order of

one of the consonant roots of the input veh , and by using rwts that have a nominal core.

6.1.2 Types of Innovations Produecd by L2 LAamt~ and Native Speakers

Although quantitatively L2 learners g e n d l y produceci fewer innovations than

native speakers (see Table 5:2), qualitatively, L2 leamen also demonstrated variability in

coining various types of innovative compound and pdphrastic constructions.

Men native speakers opted for compounding with nominal notions, they supplied

the AP+N coastniction (whcre one of the nouns is an active participle), and the N-in-

consma conmaion consisting of RW+N or /makan/+N to name instruments and

locatives. No compound constructions were produced for agents by the native speakers.

Examples of compund constructions for uistniments and locatives fiom the native speaker

data include: the AP+N compound IjSidatu naWal 'curler butter', and Iqattaiatu .. IyaqtÏÏ

'cutter pumpkias'; and the lN-inanstruct naiatu jjaW 'curling machine', and nalatu Iqa$/

for instrument item 16: "..a machine that curls buttern; and instrument item 21: "...a machine

that cuts pwnpkins" respectiveiy. For locative item 18". .a place where people fight dreams",

they also supplied /makanu lCiulm/ 'place dreaming '.

In comparison, L2 lemers supplied a wide range of compound constructions.

Besides producing AP+N and N-in-construct compounds like native speakers, L2 learners

produced N+M, N+N+N, V+N, and N+prep+N constructions. Below are typical examples

of L2 learners' use of these compound constructions.

Unlike native speakers, L2 leamers supplied the compound AP+N for agents, as well

as for instruments and locatives. For example, for agentive item 9: "...a person who spreads

rumon by phone", L2 leoiniers produced inashiru ItishaTâ~ 'spreader rumors', Inashiru

17axbatl 'spreader news', and for agent item 19: "... a person who pulls wagons", they

supplied /j& iîarabâtl 'puller wagons'. For instrument item 21: "...a machine that cuts

pumpkins", L2 leamer 36 produced: iqatibtu Lyqt,id 'cutter pumpkins'. For locative item

12: "...a place where people watch clouds", tearner 5 provided Ishahidani ssahW 'watcher

clouds' .

L2 leamers also ptovided N+M (noun+moditier) compounds, mostly for

instruments. For instance, leamer 15 came up with nala muzfijai 'disturbing machine' for

instrument item 3: "... a machine that makes a disturbing noise", and Iwaraqa malfufd

'paper strearneri for instrument item IO: " . . . a machine that makes papa streamers".

In addition, L2 leamers supplid N-in-construct wmpoundq combining the word

IrajuV 'man', Ralal 'machine', or Imakanl 'place' with another noun to denote the three

nominal notions. For example, for agentive item 5: "...a person who makes a practice of

bursting ballaons", leamers 4 ad 30 ptoduced Irajulu Ifaqàqiîl 'mm balloons', and for the

agentive item 9: "...a person who spreads rumors by phone", Irajul 7ishaW 'man news'.

For instrument item 8: " ... a machine that lights candles", among the N-in-construct

compounds producd by different L2 leaners were: R&tu shamV 'machine candie'

(learner 13), Râlatu shumÜV 'machine candles' (leamers 14 and 26), and Ralatu !?id&/

'machine lighting' (leamer 39). Also, for locative item 4: "...a place where people make soap

sculptures", learner 5 provided Imakh nnahtl 'place sculpturing'. In ail these compounds

the order of words provided is head-initiai characteristic of MSG

Mormver, L2 leamers provided N+N+N compounds where the first noua was either

/rajuV, liala/ or /makW and the second noun was an active participle for agents,

instruments, and locatives. For example, for agentive item 19: "... a person who pulls

wagons", leamers (e.g., 8, 12, and 33) supplied the compound construction Irajul jâr

I h b W 'man puller wagons'. Also, for inseniment item 16: "..a machine that curls butter",

same L2 leamen produced nâiatu jahd zabdar 'a machine curling butter', and Ralatun

shaîilm shumÜT1 'machine iightcr candles' for instrument item 23: "...a machine that lights

mdles". In addition, for locative item 4: "...a place where peuple make soap scu!pturesn, L2

leamers 9, and 15 supplied /maich nant &Md 'place sculptunng soap', and learner 29

/makhi ; M a t ssWnl . . 'place producing soap'; and for locative item 15: "...a place where

people burn clothes", tearnen 13 and 26 produced /makh harq lmalabissi 'place buniing

clothes', and lemer 43 [ma& h r q 7aûiyab) 'place burning clotfies'. (The words

nrnalabissl and /?aûiy@b/ are synonyms for the English word 'clothes').

Furthemore, L2 leamers came up with the N+prep+N wmpound constniction to

name instruments in some cases as in nala li ?idaFati shumÜT/ aaâ andnala (i 7ishWi shumÜV

'a machine for Iighting candles' for instniment item 8: "..a machine that Iights candies"; and

Rala Li jardi a a b d d for instrument item 16 : "... a machine that m i s butter" by !earners

(10.20, and 34).

Last but not least, L2 leaners produced V+N campounds. Typicai examples were:

/yajumi Iwbd 'pulls wagon' for agent item 19 "... a person who pulls wagons", and

ltajyadu zzabda/ 'curls butter' for instrument item 16: O... a machine that culs butter"

(leamer 14); and flashTa1 nnàrl 'sat fices' for agentive item 1 ".. a pnson who sets fires";

Idqab suhbl 'watched clouds' for locative item 12: ".. a place where people watch cloudsn

(leamer 18).

To slm up, this b i e f examination of native' and leamers' iruiovative compound

constnictions reveals that native speakers only supplied AP+N and N-in-constmct

compound consuuctions for instruments and locatives. L2 leanien, on the other band,

produced varied compound constnictions including AP+N, N-in-constmct, N+N+N, N+M,

and V+N fur the three semantic notions (agent, instrument, and locative notions), and

N+prep+N ody for insrruments.

b/ Imw& -pnrihrasti~ ~onstr t l~ t ion~ f~ CCZUSU~~WS

In - t m s of periphrstic consuuctions, native speakers produccd innovative

causative constructions using ody the verb /jaFald conveying the idea of causativity in

MS A Examples of such canstnictim include: ljarala hu yaqfizu mina m&da/

'he made him jump h m the window' for causative item 13: "... the action of making

somebody jump fbm the wïndow" (native speaker 12). Also, ljaralahu yajmah sydaf8tl

'he made him g a t k scashells' for causative item 22: "..- the action of making sornebady

gather seashells " supplied by the same subject. In cornparison L2 lemm produced

various innovatiw causative periphrastic c o ~ c t i o n s composed not only of the verb

/jaCaia/ 'make', but aiso lsabbabal 'cause', and /dafafa/ 'push'. Exampies include: causative

item 13: "... the action of making somebody jump fiom the window", learner 40 produced

Safala y a q h mina nnafidd 'he made jumping fiom the window'; and for the same item

learner 33 produced /dafaTa hu qafi mina nnafidal 'he pushed him jumping fiom window'.

A h , for causative item 6: "... the action of making somebody land on another planet",

learner 17 supplied the following innovative periphrastic construction lsabbaba nuzül

kawkabl 'he caused descent planet', and learner 24 produced /jaTaiahu yanzild 'he made

him descend'.

c/ Innovative &rived oarterns

Although L2 leamers produced fewer innovative derivecl patterns for each semantic

notion cornpared to native speakers (see Table 5:3), they produccd a wider range of patterns

and forms. These derived patterns, although formally appropriate in MSA, were considered

sernantically "inappropriate" in the given contexts because they expressed other semantic

notions than the oncs k i n g investigated. Examples of these derived pattern rire presented in

the foliowing paragraphs.

Among the inappropriate derived patterns that leasnets produced, notable was a

verbal noun CiCaCa of pattern 1 verbs which is associateci in MSA with professions and

aafts. This pattern was sometimes supplied for locative and causative items. Exampies of

this type are /niriata/ 'action of c a ~ n g ' coined by leamer 20 for locative item 4 instcad of

the prefixed pattern /manhata/, and 1jirnaTa for causative item 22 instead of the innovative

causative patterns IjammaW, or RajmaW.

Another inappropriate deriveà pattern offered by L2 leamers was the verbai noun

taCCiC of pattern II verbs for locatives and causatives. Examples of these occurrences (ail

supplied by learner 1 1) are /tanhW 'carving' for locative item 4 instead of the innovative

prefixed pattern Imanhata; ltaqfld 'jumping' for causative item 13 instead of the causative

verbal pattms I q a f f d or ?aqfua/; and ItajmKI 'pilling' for causative item 22 instead of

the verbs IjammaW or ff ajrnayd.

ûther inappropriate derived pattms used by L2 leamers included pattem maCCUC

(the passive participle of defived verb pattern 9, and pattern ~ C a C i c (passive participle of

denved verb pattem U). These pattems were provided mostly for locative and causative

notions. The foilowing examples were provideci by different L2 lemers. For instance,

lmat~rÜq/ 'burned' for locative item 15 produced by leamer 25 instead of lma?iraqa/, and

IrnujamiTl 'piled up' for causative item 22 supplied by lmer 16 instead of ljammah/ or

l?ajmaW.

In some cases learners inappropriately produced the simple verbal form CaCaCa, or

the ceciprocd verbal form CâCaCa ta name causatives. For exarnplq for causative item 6: "

... the action of making somebody land on anather planet", Iemers sometimes produced the

simple verb hatqd 'to descend' i a s t d of the innovaiive causatives /habbay or RahbaF,

and for causativc item 13:" ... the action of making somebody jump Eiom the window",

leamer 20 supplieci the reciprd verb I q a f d instead of /qafFazal or /?aqW.

Further inappropriate denved patterns include cases where L2 learners produced

patterns that denote relational adjectives by adding the suffix /iy/ tc? the noun. Typical

examples include [Yarabatïy] for agent item 19, and [ishali'iy] for agent item 9 where leamer

5 added the suffix /iy/ to the word [tarabal 'wagon' and fishiTa/ 'rumor' respectiveiy

instead of coining agent patterns such as Karr&/ and /shayyS/.

6.1.3 Sumrnary

To sum up, the results of native speakers' and learners' innovation responses in the

production task were in accord with hypothesis lb. A qualitative difference was found

between L2 leamers' and native speakers' production of innovations. L2 learners, unlike

native speakers, exhibited more variability in coining their innovations by using verbs that

are distantly related or semanticaily unrefated to the vehs in the input, by transforming or

changing the order of one of the consonant mots of the input verbs, and by using a wider

variety of nominal roots. The variability is alsa manifestai by L2 learners' production of a

wider range of compoundland penphrastic constructions and derived patterns denoting othw

semantic notions.

6.2 AequWiîiond Priaeipks in L2 Leamers' Use and Chaice of MSA Word Formation Proctssts

As shown in Chapter 5, section 5 3 , L2 learners demonstrateci a strong inclination in

favour of compounding in the three tasks. Vi+gem., and Vinf. were also common. Why

did leamers pick compounding more ofien tltan the other options? What principles guided

their use and choice? in this section, the general acquisitional principles of productivity,

semantic rransparmcy, formal simplicity, and conventionaiii proposed by Clark (1980a)

are assessed to h d out which of these principles guided these L2 learners of MSA in

denoting agency, instnimentality, locatios and causativity. Aiso examined is whether the

same principles that are held to account for acquisitional orders in Indo-European laquages

(EngEsh and French) and in a Semitic language (Hebrew) also apply in MSA My predicrion

was that L2 leamers would tend to use productive, transparent, simple, and conventiond

patterns in preference to unproductive, opaque, cornplex, and unconventional ones to

express agents, instxuments, locatives, and causatives (hypothesis 5). To test this hypothesis,

the L2 learners' use of word formation processes in the production task and their choice of

word formation in comprehension tasks 1 and II are examined.

63.1 Priaciplc of Producâivity

In line with Eve Clark's notion of productivity (Clark 1980a) and with findings fiom

other studies of language acquisition discussed earlier (see chapter 3), the principle of

productivity States that word formation processes that are the most productive should be the

most avdablc and should be uscd in preference to less productive processes. This means

that for agents in MSA the very productive Vinf+gem. pattern CaCCâC (Badry,l983)

should ûe more easily identüied by L2 leamers and more fiequently relied on than the Iess

productive Vid pattern CaCiC, prefixed pattern muCaCCiC, and N-in-construct compound

/rajuV+N . In the present study, the L2 leamers' preference for the productive pattern

CaCCBC for agents in comprehension task II (reai items) is in line with the prediction.

However, their more fiequent choice of compounding to coin agents in the production task

and cornpreknsion task I does not d o m to this prediction.

For instruments, the very productive V i + g e m pattern CaCCBCa would be used

more fiequently tthan the N-in-Constnict f?ala/+N, the Iess productive prefixed pattern

miCCaCa, and the pattern CaCiCa L2 Iearners' preference for N-in-constnict for

instruments in the production task and comprehension tasks I and II runs counter to the

prediction.

For locatives, it was predicted that the prefixed pattern &Caca would be most

Erequently useci, king the most productive pattern to express location. Next wouId be the N-

in-construct /makW+N, which is occasionally used in Arabic for locatives. The learners'

preference for wrnpounding both in production and in comprehension tasks 1 & [I does not

confonn to this prediction.

For causatives, the principle of productivity would predict that the very productive

Vinf.+gem, verbal pattern CaCCaCa would be used more oAen than the pattern ?aCCaCa

and the periphrastic construction with ljafdaf 'to make someone (da sornething)'. The

leamers' reliance on periphrastic wnstniaions in ait t h tasks nuis counter to this

prediction.

6.2.2 Principlt of Semantic Tnnspartncy

The principle of semantic transparency states that processes with one-teone matches

of meaning to form are easier to acquire than tbose with one-to-many or many-teone

matches (Clark, 1980a). This means that for agents, compound constructions cornbining the

word IrajuU 'man' with a noun shoirld be more easiIy idemifieci and more frequently opted

for by leamers of MSA as these words express the notion of agency most explicitly. Next

would be the pattern muCaCCiC, having only the agentive meaning, and finally the pattwns

CaCCàC and CBCiC which have various different mcanings: CaCiC can have the masculine

active participlc, agemive, and insüument meanings, and CaCCaC can demte a g d v e and

Uis~nrmentai rneaning.

The L2 learners in this study did in general seem to be operating with the one-to-one

mapping strategy. Their preference for N-in-constnrct /rajuY+N for agents in the production

task and comprehension task I reveals that sernantic transparency plays a role in their use

and choice of word formation processes. However, their use of the pattem CaCCX in the

comprehension task U mns coumer to the prediction,

For instnimentality, the sernantic transparency principle means that the word /?aIai

'machine' with a noun should be most fiequently opted for. Next would be the pattern

miCCaCa with only instrumemal rneaning, followed by patterns CaCCXa and CaiCiCa

since both denote two meanings: either a feminine agent or instrument for CaCCaCa, and

either active participle or instrument for pattern CaCCdCa in the three tasks, this prediction

was borne out as L2 learners relied most heavily on the compound R W + N to coin

instrument nouns. But their non-use of the prefixed pattern miCCaCa, and their reliance on

patterns CaCCKa, and CaCiCa nin counter to this prediction.

For location, the transparency principle would point to the N-in-consmict

lmakW+N in preference to the pattem maCCaCa This prediction was supported by L2

leamers' use of the N-in-construct /makanlfN for locatives in production task and in

comprehension tasks t and ïï. The pattern maCCaCa was not optd for at all.

For causativcs, semantic transparency would predict that periphrastic constructions

with /jaîaia/ 'to make someone (do something)' would be used more ofien than derived

intixed and prdxed patterns. The aplicit idea of causativity that the verb /jahla/ conveys

makes the construction semantidly transparent, Learners' fiequent use of periphrastic

constructions for causatives in the production ta&, and in comprehension tasks I & II

indicates tbat semantic transparency influences their choie.

63.3 Principle of Formal Simplicity

According to Clark's formal simpticity principle, it was predicted that ffixation

would be more relied on and used in preference to infixation. This means that for agents, the

prefixed pattern muCaCCiC should be used more t'requently than the infixed pattem

CaCCW. The L2 leamer findings, however, contradict the suggested strategy of reliance on

f i ~ a t i o n before infixation as evidenced by their non-use of the prefixed patterns

muCaCCiC for agents, and their use of Vinf. and Vinf+gem. processes on the three tasks.

For instruments, the principle of forma1 simplicity means that the prefixed derivd

pattem miCCaCa should be used more frequently than the infixed andor the geminated

pattem CaCCaCa, and CaCiC(a). The L2 lmns' non-use on any of the tasks of the

prefixed pattern miCCaCa also runs counter to the prediction.

For locatives too, the principle of formai simplicity means that the prefixed pattern

maCCaCa would be used more f'requentiy than the infixed pattern. L2 leamers' use of the

infixed pattern CaCiCa before the prefixed pattem mns counter to the prediction.

For causatives, the principle means that the derived verbal pattern 7aCCaCa should

be acquired and relied on before the infixed and geminated pattern CaCCaCa to coin

causatives. The L2 leamers' use of the infixed pattern CaCaCa and the infixed and

geminated pattern CaCCaCa does not conform CO the prediction.

6.2.4 Mncipk of ~o&itiondity

This pcinciple would predict that the pattern CaCCaC would be used most fkqwntIy

for agents, the pattern CaCCaCa for instnunents, the pattern maCCaCa for locatives, and the

pattern CaCCaCa for causaiives since these are the conventional devices used for the

respective notions in the MSA lexicon. Tt is worth noting that di these patterns were most

commonly used by native speakers to coin the respective semantic notions in this study. The

L2 lemers' general preference for compounding to denote agents, instruments, and

locatives in the production task and comprehension task 1, and instruments and locatives in

the comprehension task il, and also their use of periphrastic constructions in production task

and compreheasion tasks 1 & II for causatives reflect that their stmtegies are not guided by

the conventionality principle. Only the learners' preference for pattern CaCCaC for agents

in comprehension task II suggests that their strategy was influenced by conventionality.

6.2.5 Summr y

To surnmarize, hypothesis 5 was only pdally confirmed. The data on the L2

learnm' use and choice of MSA word formation processes indicate that not ail the four

acquisitional principles account equally well for their use and choice of word formation

processes. Semantic transparmcy is the ptinciple that best predicts the learners' use and

choice of word formation processes, This principle orients them towacds compounding to

coin imovative items for the different semantic notions. Ptoductivity and conventionality

also appear to have a bearing on leamers' coining of agents when they chose the Vinf,+gem.

process in comprehension task II. The formal simplicity principle, however, has not guided

lemers' we of word formation processes.

6.3 LI Effceb on the L2 Use of Word Formation Processu

This section is devoted to a discussion of the influence O f LI on the lemers' use of

word formation processes in MSA It was predicted that there would be transfm fiom LI to

the L2 in the use of these processes (ûypothesis 6).

The L2 leamen who participated in this study wnsisted of 30 Engiish-speaking

Leamers, 6 French-speaking learners, and 8 learners coming fiom different language

backgrounds (namely, German, Itaiian, Persian, Somali, Spanish and Urdu). Because the last

group contains a srnall number of representatives of each language (one for each different

language, except for Somali where there are 3 representatives), and because al1 these 8

leamers of different Lls claimed that English was their dominant language, they were not

considered in this examination. Therefore two groups of leamers were distinguished, 30

English speakers and 6 French speakers, and their responses were examined.

6.3.1 L l Effets on tbe Word Order of Compound Coastnietions

When L 1 English-speaking and L 1 French-speaking leamers' preferences for

MSA word formation processes in the three tasks were wmpared, they were found to be

generally similar. Both English and French speakers exhibited a strong preference for

compounding. N+N compounds including AP+N, N-in-constnict, N+M, and N+N/N+N+N

were by fat the most muent choices. The muent choice of compounding to coin

innovative agents, instruments, locatives, and causatives in MSA by the two Ll groups

could be seen as an influence from both Lls, But because in Braeder (1993) compounding

was not restncted to L1 influence . where even Moroccan Arabic Iearners of L2 Dutch opted

for compounding, this latter can be claimed to be an universal process of language

acquisition. In this present study, because of the structurai distinctions betwecn the source

languages (Engtish and French) and the tarsct language (Arabic) with respect to compounds,

however, it was predicted that English- and French-L1 learners would show dinerential

source ianguage effects in their coinage of innovative compounds. Thcre is some evidence

tiom the learner data that lends support to this prediction.

The effect of L1 on L2 word formation is seen in the word order of compound

constructions. In the production task, there are a few instances where L1 English-speaking

learnefs occasionally follow the English head-tinal order when forming innovative N+N

compounds in Arabic. For example, for agent 5: ".. a person who makes a practice of

bursting balloonsn Iearner 9 supplied the head-final compound constniction N+AP halünàt

aqiV 'balloons burstm' instead of head-initial compound If8qiT balUn&/ 'buster bdoons'.

A h , for instnunent item 2 ".. an instrument which makes a disturbing noise" leamer 13

produced the head-tinal compound construction M+N ImuzCijun sawt/ 'disturûing noise'

instead of a head-initial compound /sawnui rnuzfijud literally translated in English 'noise

disturbing'. Another example of the head-final is the locative item 4 ".. a place where

people rnakes a practice of bursting balloons" where English L1 leamer 13 produced the

following (head-final) constniction /sibÜn BamàtW 'soap sculptures' instead of an

innovation with head-initiai order iûamâtil dbüniyd l i t d l y 'sculptures soap'.

in contrast, LI French-speaking leamers produced only head-initial compound

constructions, an ordn typical of French as well as of the target language (Atabic). For

example, for the same item 2: "., an instrument which makes a disturûing noise" French L1

leamer 6 pruduced the hcad-initial compound canstniction N+M Isawt muzTijl 'noise

d i d i n g ' , and for instrument item 10: "... a machine that makes paper streamersn, the same

leacner produccd nawrâq malffitiil 'papers streamer'. Also, L1 French-speaking leamers

came up with the Niprcp+N type of compound contniction This is typical of Arabic

where tbis constmction is used to denote only instrument noms. Examples providecl by Li

French-speakuig leamers were nala ii qatTi lyaqtid 'machine for cutting pumpkins'; and

Rala Ii jatdi zabdd for instrument item 21: "-..a machine that cuts pumpkins", and

instrument item 16: ".. a machine that curis butter" respectively.

6.3.2 Summary

To sum up, it was found that the findings supporteci hypothesis lb. There was a

qualitative difference between L2 leamers' and native speakers' innovations in îhat L2

learnen dernonstrated more variabiiity in their responses than native speakers. Findings only

partially codhned hypothesis 5. Only productivity, conventionality, and primady semantic

transparency were found to account for L2 learnen' production and judgment of word

formation processes. Hypothesis 6, on the other han& was given some support . An LI

effect is manifested in the word order of compound constructions where a few L1 Englisti

leamers were influenced by the word order principle of th& (head-final) L1 in producing

N+N compounds.

CHAPTER 7

DISCUSSION AND CONCLUSIONS

This chapter concludes the study. First, a summary and discussion of the findings

will be provided. These findings will be discussed in relation to the six hypotheses

fomulated in chapter 4 and in the light of previous research and theory. The second section

of this chapter will outline the limitations and shortcomings of the study. Implications for

pedagogy and for future research will conciude the chapter.

7.1 Summrry and Discussion of the Main Fînding

The study examineci adult L2 learners' use of MSA word formation processes io coin

innovative agents, instruments, locatives, and causatives, and to choose invented and

existing items denoting these semantic notions. It also investigated some aquisitional

principles to account for these learners' use of MSA word formation processes to express

these f o u semantic notions, Two groups took part in this study: A group of 44 adult L2

leamers, and a group of 40 native speakers of Arabic. Ail participants were given a

production task and two comprehension tasks to test the six hypotheses formulated for this

study, as well as a background questionnaire designed to providc personal background

about the participants, @eu knowledge of Arabic, their learning expcricnce, and their seif-

ratine of some activities in Arabic. In addition, L2 lmen wcre given a vocabulary

knowledge test designed to dettrmine their lexical knowledge in MSA. Quantitative and

qualitative analyses were couducted to compare the L2 lemers' and native speakers'

production of innovations, and th& choice of MSA word formation processes in naming

both inventai and real agentive, instrumental, locative, and causative items.

The following seven main findings were drawn tiom both quantitative and

qualitative analyses :

7.1.1 On the Production of Innovations

Hypothesis la was partially supported- ïhere is evidence fiom the overall results on

the production task that quantitatively L2 leamas produced aimost as many innovations as

native speakers (M=17.18 by learners compared to M=18.32 by native speakers), and that

L2 learners oflen gave non-responses (see Table 5: 1). The L2 leamers' and native speakers'

findings for each semantic notion, howwer, revealed that L2 leamers supplied fewer

innovations for agents and Uistrumcnts, and more innovations for locatives than native

speakers. L2 leamers also produced as many innovations as native speakers for causatives.

The L2 lemers' lower production of innovations for agents and instruments in the

present study could be attniutcd to the types of items provided in some agent and

instrument questions. L2 leamcrs seemed to produce innovations only for items that

included precise and specific verbs in the questions such as the agent item 19: "... a

person who pulls wagonsn, the instrument item 8:" ... a machine that lights adles", and the

locative item 15: "... a place whae yw burn ciothes". But L2 leamers fond it difficutt to

supply innovations and most oficn le& bIanks for agent and instrument items that did not

contain specific main v d s in the question Such items are agent item 5 that contained the

expression 'makes a practice of burstiag bdlwns' rather than the main verb 'bursts', and

instrument item 3 that cuiitained the expression 'makes a d i i n g noise' rather than the

main verb 'disturbs'.

Even though L2 Iearners produced fewer innovations than native speakers for agents

and instruments, both groups showed similar patterns by supplying more innovations for

agents than for instruments and locatives. The native speakers' mean was 5.75 for agents,

5.38 for instruments, and 2.97 for locatives compareci to learners' means of 4.87 for agents,

4.64 for instruments, and 3.77 for locatives. This finding is in :ine witb Olshtain's study

where agents were the most productive type of nouns (82% innovations for agents). This

greater use of agents in the preseat snidy may be due to the fact that agents are saiient items

that are ofien introduced and taught in class.

As predicteâ by Clark et al. (1984). locative innovations were less fiequent than

agents and instruments. However, in Olshtain's study (1987), it was found that instrument

innovations were the least productive of the three semantic notions. In the present study,

what accounted for the native speakers' low mean of locative innovations was the greater

use of real-words suppletives. Native speakers supplieci suppletives mainly for six items (4,

7, 12, 15, 18, and 20), but items 12, 18, and 20 received more suppletives. Typical examples

of the suppletives produced by native speakers were ifaeafi 'space', ljabaV 'mountain',

lsamâ'll 'sky', Innafida/ 'window', /&Urfa/ 'balcony' for item 12: "..a place where people

watch clouds". Alsa, for item 18: ",.a place where p p l e fight dreams", suggested

suppletives were /fi&û/ or l d I 'bed', hlml 'dream', fia* 'life', Iyaqaeai

'awakening', /dqiF/ 'reality', Inawml 'sleep', /shuCÜr/ 'fetling', and for item 20: "..a place

where birds fly and swim at the smme timehlicited responses were fiadîqal, 'park',

hhayral 'laite', One plausible e x p I d o n to accauat for native speakers' use of

suppletives may be that becaufie narive speakers possess a rich vocabulary fiom which they

can draw, and becawe derivationai mcans ta express locatives are resuicted maialy to

afixation, with the N-in-construct wmpound alternative being less ftequent, native speakers

prefened to resort to suppletives and real words as an alternative. This p a t e r use c'f

suppletives by native speakers for twatives in this study is in line with a fùiding in

Olshtain's study (1987) where Hebrew native speakers produced more suppietives than L2

Hebrew lemers for locatives.

Even though L2 leamers p r o d u d more irinovations for locatives than native

speakers, they were mainly inappropriate derived infixed andlor gerninated patterns that

denote other semantic notions. There are also a few cases when learners, especially higher

level learners, tended to produce derived prefixed patterns wch as maCCÜC, muC&CiC, and

taCCiC. These denved prefixed pattems are considered "inappropriate" because they denote

not locatives but other semantic notions. The production of these inappropriate prefixed

pattems is evidence th! thse L2 leamers lack the knowledge to identify the appropriate

patterns to denote specific locative meanings.

Besides producing innovations for locatives, a few L2 lemers supplied suppletives

for items 12 and 18, What is interesthg is that when they produced suppletives for these

items, they supplied similar words as native speakers, for example, /sama?/ 'sky', and

/mafida/ 'window' for item 12, and lfir&sh/ and /sariri 'bed' for item 18,

For causatives, both native speakers and L2 leamen coined a considerable number of

innovations (Table 5:2). When native speakers Med to do so, they supplied real and

existing words. Among the causative items thai received real words were items 2, I l , and

13. Typical examples of ùicse exiahg wods were fiamdqal 'craziness', and flaKunf7

'violence' for item 2: ".. the action of making a blind person drive a car"; and llinti?~&/

'suicide', and Ralxawfï 'fear' for item II: ".. the 4ction of making somebody nin

backwards". Also for item 13 ".. the action of making somebody jump fiom the window",

suggested real words were / l h d b/ 'escape'.

L2 learners, on the other hand, when they failed to coin causative innovations,

mostly left blanks. This shows the learners' unfamiliarity with causative patterns.

7.1.2 On the Typa of Innavations

L2 lemers exhibited greater variability than native speakers by forming innovations

tiom different verbs than the ones provided in the input, and by producing a wide range of

innovative compounds and derived partems denoting the targeted notions, as well as

inappropriate derived pattms denoting other semantic notions.

These findings support hypothesis lb which predicted that there would be a

qualitative ciifference between L2 itaniers' and native speakers' production of innovations.

Aithough quantitatively L2 leamers produced almost as many innovations as native

speakers, quditatively the two groups demonstrateci many differences.

Uniike native speakers, L2 leamers produced a wide range of innovative compounds

to denote agency, instmmentality, and location, and imovative periphrastic constnictions to

denote causativity. It was dso found that L2 leamers produced varied detived pattern

inappropriately denoting 0th- notions, and produced t 0 d l y impossible patterns though

metathesis or tniasformaiion. This finding is in line with Badry's study (1983) where L1

Momccan-speaking children also prduced inappropriate patterns such as verbal nouns

CeCCan as in lbtnanl 'action of picking up', and taCCiC as /tqtiF/ 'action of cutting',

instead of the agentive pattem. They also produced misfonned patterns by adding a

consonant or semi-vowe! to the roog for example, ilehhawd h m the verb huh/ 'throw' for

inmumental ~ x i n hsteaci /lewWanal 'thrower' @dry, 1983, pp.162-163). The L2 iearners'

production of these varied derivai patterns suggests, on the one han& that the Leamers are

actively involved in constructive and productive use of the derivational system, and aware of

the root and pattern system in MSA word formation which requires the association of

consonantal roots with pattemed vocaiic infixes. On the other han4 it indicates leamers'

incomplete and limited grasp and knowledge of derivation as they do not know the structural

relationships between f o m and meanings in MSA. or the lexical constraints in this

language.

7.1.3 On the Use and Cûoicc of MSA Word Formation Processes

On the three tasks, the L2 leamers tended to prefer compounding over the other word

formation processes, but they also opted for Vinf., and Vinf.+gem. to a lesser extent.

Anxation was hardly ever selected, Native speakers, on the other han& tended to favour

derivation, mainly Vin£+gem. and affixation.

In the production task, compounding was found to be the preferred and primary

process used by L2 leamers to denote agents, instruments, and locatives (see Table 5:4). L2

learners at different levels of vocabulary knowledge made innovative use of cornpounds.

The most fiequent types of naminal wmpounds in the leamer data were AP+N, N+M, N-in-

constnict, and N+N/N+N+N types ofcompound constructions. L2 lemers also followed the

same strategy for causative innovations by relying on periphrastic constructions more oflen

than derivational processes.

The picture in both comprehension tasks 1 (inveated items) and II ( r d items) is

quite simiIar to the production ta& Tn comprehension task I, L2 learners also favoured

compounding for the three nominal notions and periphrastic constructions for causatives

(see Table 5:7), in comprehension task il, leamers also selected compounding most

commoniy for instruments and locatives and peiiphrastic constructions for causatives. Only

for agents did L2 learners chose Vinf-tgem. over compounding (Table 58).

The finding that compounding was the main process that adult L2 learners reiied on

to coin innovations supports the theoreticai daim first asserted by Clark in most of her LI

studies of English-speaking children (198Ia, 1981b; Clark et al., 1982; 1984) that

compounding is the first device lemers utilize in the course of acquisition of word

formation processes. Evidence fiom previous empirical studies and ftom observations of

spontaneous speech and elicitation reveals tbat chiIdren relied on compounding and that

compounding with the N+N pattern accounts for a large number of the innovations coined

by English-speaking children to denote agents and instruments. A similar early innovative

use of compounds has also been observai in L2 studies (e.g. Olshtain, 1987; Broeder et al,

1993). For example, in Broeder et aL1s study (1993), lemecs of different LIS (including

Arabic) and L2s had a strong preference for compoundiig, and N+N was the most

productive compound constmction opted for. The pderence for compounding in this study

adds to Broeder's evidence of universal processes oflanguage acquisition.

While compounding was found to be the primary choice, derivational processes,

mainly VinE+gem., and Vuif, appeared as cornmon in somc cases. L2 leamers did alsa use

infixed and geminated derived patterns to denate agents, instruments, and locatives.

Sometimei, the mean percentages of these deriveci processes outnumbered the mean

percentages of compounding as in the case of agents in comprehension task II, where mean

percentages were 27.27%, and 47.08% for Vinf. and Vinf+gem. respectively, compared to

25.65% for compounding. Similady, when L2 leamers opted for derivation to coin the

causative verbal notion, they opted for Vinf., foiiowed by Vii+gem. in the three tasks.

f ixation was not used at aü in botb cornprehension tasks for both nominal and verbal

notions. Only in the production tasic, did L2 learners supply the prefixed patterns for

- locatives and causatives, but these prefixed patterns were inappropriate in this context

because they express other meanings than the targeted notions. The L2 leamers' virtual non-

use of the atFxation process runs counter to Clark et al.3 prediction (1984) that stem-

extenial affixes should be used before infixes to coin new words. Howwer, L2 leamers'

non-use of fixation is in line with Badry's study where L1 Moroccan speaking-children

did not opt at any age group for the pretixed agentive pattern muCaCCiC or the instrumenta1

miCCaCa. One plausible explanation is that these patterns are Iess productive and less

fiequent in the Arabic lexicon.

This discrepancy in the L2 learners' use and choice of word formation processes may

have been influenced by the type of instruction and the aaiount of exposure to MSA word

formation processes the learners had had. A close examination of the learners' responses

rwealed that compounding appeared to be provideci most often by leamers who were

exposed to the communicative teaching method. These lemers supplied numerous varied

compound innovations, In coining most of these innovations, these lemers did not limit

themselves to the glosses for al1 words in the input. Instead, they used other words that did

not appear in the definition, and mrnbined hem to make up carnpounds. This indicates that

these leamers possessed enough vocabulary in their stock from which they could choase to

coin innovations. However, derivationai processes (mainly Vuif., and Vin.+gem.) were

supplitd by leamers who were exposed to formai instruction, and had ben introduced to

derivation rules and to derivcd patterns, mainly the ones based on infixation and gemination.

In mrnparison, native speakers showed a hi@ preference for derivational processes

over wmpounding, tàvounng V i + g e m for agents, instruments, and causatives, and

&&on for locatives for the thm tasks. For agents, cornpouding was not opted fbr in

the production task or eitber comptehension task A plausible expianation for native

speakers' non-use of compounding for agents is its infiequeiicy in MSA in expressing this

notion. Also, native speakers did not use Vinf. for instruments, locatives and causatives in the

three tasks. For instruments, the non-use of ViL represented by pattem CaCiC(a) is

attributed to its restnded use in a scientific domain. in this study, instrument items are not

associated with this specific domain. For locatives, the infixed paaern CiCiC(a) was not

used because it expresses the notion of agent or instrument. ln addition, native speakers did

not op for the Vinf tgern. process repre~ented by pattem CaCCaCa for locatives because it

denotes instrurnemdity and ferninine agency. Ah, for causatives, the Vinf. pmess

represented by pattern CBCaCa was not used because it is inappropriate in this context as it

denotes recipracity. Mormver, native speakers used the afixation process less ofken than

Vinf. and Vinf.+gem. processes especially for agents and instruments. Twa possible

explanations can be proposed here: One may be because affiation represented by the

patterns muCaCCiC(a) for agents and miCCaCa for instruments is inffequent in MSA, and

the usage of these prefùred patterns is limited ta items that are related to education. The other

explanation may be because these two panerris are cained fiom the derived verbal pattern II,

not fiom a simple verbal rwt. In the present study, there is one case in the comprehension

task II where the verb in the input is relateci to the educational context. For example, the

derived pattem lmimhâtl was judged as the suitable choice br the instrument item 22: "....a

machine that ensa mistakes". There are alw few cases in the the tasks where the verb in

the input is a daived v d pattern Il that reqirires the innovative derived preked p a m

and the native speakers use in these cases. For example, in camprehension task i, the pattern

ImubayyiF was chosen as the appropriate choice for the agentive item 5: "...a woman who

bleaches the floor", because the verb provided in thc question is Ibayya* a d e h d v e M

pattern II. Asa in comprehension taçk a the pattem /muwazziT/ was chosen for the agentive

item 17: " ... a p'rrson who delivers mail" as it is formed fiom IwauaTd a derived verbal

pattern II, and the instrument item IO: "... a machine that squeezes juice", /miYsara/ was .- chosen because it is derived fiom the verbal pattern II Kassara/ provided in the input. * 8

Native speakers, unlike L2 leamers, were more consistent in their use and choice of

word formation processes. This wnsistency is manifested by their preference for the same

patterns for both inventeâ and real items. This consistency is also shown by their choice of

the appropriate, most productive, and fiequent pattern, Even though L2 leiuners followed

the same order of preference for each semantic notion in the three tasks, they tended to use

and choose inappropriate patterns that denote other semantic notions for both invented and

real items, for example, their choice of agentive and insmunent pattems C&CiC(a) and

CaCCàC(a) for locatives, and the miprocal pattern CâCaCa for causatives. Leamen' use of

these pattems suggests that they are not aware of the associatioa between patterns and

meanings, and of the convernional constraints in MSA

7.1.4 On tbe Effkct of L2 Lcvel of Vocabuhy Knowlcdge on tbe Use and Choict of MSA Wod Formation Procusm

With respect to the &éct of L2 vocabulary knowledge on leamers' use of word

formation processes to min innovations, findings reveaI that higher Ievel lemers tended to

differ tiom mid, and lower level Ierrmers in using Vinf.+gem. to coin innovative agents and

locatives for which they used it more often than the two 1-1 groups. Higher level leamers

also diered ftom mid and lctwer level Learrrers in opting more often for V i to coin

innovative instrumcuts and causatives. For this latter, highcr level lemers aiso opted for

affixation while neither mid nor lower Ievei ieamers did so. F d l y , the effect of L2

vocabulary knowledge was revealed in the use of wmpounding for instruments and

locatives. For the former, lower level leamers opted more often for this process than mid and

higher level learners while for locatives, it was mid level ieaniers who relied more often on

this process than lower and higher level learners.

For derivational processes (Vinf., and Vinf.+gern.), most of the significant

differences in means lay between lower and higher level l e a m , and mid and higher level

leamers. These findings suggest that lexical knowledge in the target language did influence

the L2 learners' use of some of the word formation processes to coin their innovations.

Higher level teamers appeared to be more knowledgeable about derivation in MSA than mid

and lower leamen.

7.1.5 On the Higher Level LZ Leamen' Nativolikc Ability in the Production rad Cboice of Wod Formation Rocesses

Findings in the three tasks revealed that the behaviour exhibiteci by higher level L2

leamers of MSA did not resemblc that of the native speakers with respect to the use and

choie of word formation processes.

in the production task, the higher level L2 leamers' use of word formation processes

was quantitatively similar to native speakers' only with respect to V i d , and Vinf.+gem. to

name agents, and with respect to compounding to narne ins~rumcnts (see Table 5: 11). In al1

other cases, the t w ~ gtoups differed significantly in their use of processes.

Findings in comprehension task 1 (see Table 5:13) also show that the higher level

Iearners' choice for word formation processes differed in general h m that of native

speakers. M y for Vinf. to denote instruments, for compomding for locatives,

and for ViIlf.+gem for causatives did the higher ievei leamers' choice of word formation

approxirnate that of native speakers.

In addition, resulta in comprehension task IT show t h only for Vinf.+gern. was the

use of higher level lme r s similar to h t of native speakers to label instruments (see Table

515). For everything else, higher level leamers did differ from native speakers. These

findings contrast with Olshtain's (1987) where advanced learners, unlike in this study. did

reach a level of Hebrew cornpetence titat enabled them to produce and assess innovations in

ways chat appmached native spegkm' responses. The difference between these two sîudies

maybe due to the leaniing wntext, and to the amount of exposure to the target language. in

Olshtain's study, learners were living in Israel, and leaming Hebrew as an L2. They were

exposed to Hebrew in and outside class. Therefore, they were more experienced-

SimiIarly the lower level leacners' strategy in using and choosing word formation

processes in the present study did not approximate that of native speakers in the three tasks.

In the production taslr, Iower level L2 lmers' use of word formation processes was

quantkatively similar to native speakers' oniy with respect to Viuf. to name agents,

instniments, Imtiveg and causatives, and Vinf.+gem. to denote instruments. in al1 o k

cases, significant difikences were reveaIed between the two groups' use of word formation

processes (see Table 5: l f ) .

Findings on cornprehension task 1 show that Iower level feamers' choice of word

formation processa also d i d significantly h m that of native speakers on most response

types (set Table 514). For V i d to name insûumentq for Vinf+gem. for agents, and for

periphrast'i consauction for causatives, however, lower level leamen' choice of word

formation processes was not ~ i~ f i cant ly different fiom that of native speakers.

in cornprehension task II, tiie Iower Ievel learners' cbice of word formation

proasses was simila to that ofaative speakers oniy to label instruments and locatives (see

Table 516). For everything else, lower IeveI leamers foliowed a different straiegy in their

choice of word formation processes.

In the three tasks, then, the behaviour of both higher and lower level L2 learners of

MSA did not in general appear to resembk that of the native speakers with regard to the use

and choice of word formation processes. Neither higher nor lower level learners appear to

have reached a sufbiently high level of vacabulary knowledge in MSA to approximate

native speaker choices of word formation processes.

7.1.6 On tbe Effceb of Acquisitional Priaciplu in the Use and Choice of Word Formation h a s u

There is evidence in this study that semantic transparency played an important role in

learners' use of word formation processes. Leamers tended to rely on whole words rather

than using derivation to coin innovations. Leamers coined wmpound cons~ictions for

agents, instruments, and locatives mostly by wrnbining two words. ïhis finding is in line

with most L1 and L2 studies (e.g., Clark, 19804 1980b, 19814 1982, Clark et al., 1982.

1984; Berman et al. 1982; Brocder et d ' s 1993, 1995) that have indicated that semantically

devices were opted for by learners.

Other principtes that account for leamers' use of word formation processes in this

study are productivity and conwntionality. Learners sometimes celied on the most

productive, fkquem, and conventional process to coin th& innovations. This findîng adds

to the evidence that productivity and comntionality help to account for the acquisition

of word formation processes- Previous research on diffmnt Lls, and L2s (e.g., Clark,

1980% 1981a; Badry, 1983; Olsbtain, 1987; Brader et al., 1991, 1993, 1995) has shown

that both chiIdrcn and aduh leamers relied oa productive prüccssé~ in their language to coin

new words. For example, in Badry's sbdy (1983), LI Moroccan-speaking chiIdren tended to

rely on the productive and conventional pattern CeCCaC and CdCaCa for agents and

instruments respectively. F o d simplicity, however, did not account for L2 [earners' use

of word formation processes in this study.

7.1.7 On tbe Ll Elfeci in tbe Use of MSA Word Formation Procesw

In this study, it was alw hypothesized that besides the four acquisitional principles,

the I , 1 would influence L2 learnns' production of word formation processes (hypothesis 6).

The findings of the production task pmvide some support for this hypothesis. Some cases of

L1 transfet occurred in the word order of campound where L1 English-speaking leamers

sometimes followed the English head-final order in conmcting compaund constructions in

MSA

in coining innovations, English- and French-speaking leamers exhibitecl a high

pref'erence for wmpounding by producing more compounding constructions han innovative

derived patterns. This finding is in line with Olshtain's (1987) finding indicating h t the two

groups who consisted of 3 2 E@sh speakers and 28 S panish speakers, and who made up the

largest mb-grotrps of the advanced leaniers of 17 diffcteat Lls, similady demonstrateci a

high prefmnce tbr compciunding, The high prtfércnce for cornpounding over dnivation

was also fowd in Bmeder's (1991) mdy whae ail aduit leamers of different Lls inctuding

Moroccan Arabic opted for compounding over derivation. Therefore, even though

compoundiig is more productive and more 0th used in both target languages (English and

French) than MSA, the hi@ pretbllce fbr compouading over derivation by bath English-

and French- speaking Iearners in t h p n a t study could aot be attributed to LI influence.

However, L 1 transfer was seen in the word order of compoundmg where L l English-

speaking leamen sometimes followed the English head-final order in constmcting

compound constructions in MSA This finding corroborates previous findings in Broeder et

al. (199 1,1995) that showed that the opposite word order for Tutkish (head-final) and for

Arabic (head-initial) emerged as an L1 effect in Turkish- and Arabic-speaking Iemers'

production of compounds in Dutch L2.

7.1.8 Summa y

Thus, it may be advanced that the findings of this study have shed fiuther Iight on

leamers' ability to use word formation processes at an early stage of acquisition, and on

their preferences for MSA word formation processes. Aise, some M e r insight has been

@ed into the role of acquisitional principles in the acquisition of MSA word formation

processes. However, these findings should be considercd with caution because the study

involved some Limitations which constrain the generaiizability of the findigs.

7.2 Limitations of the Study

7.2.1 Sizt 01 the Sampie

One of the limitations can be aîtributed to the small sample of learners who

participated in îùe present study. The sample consists of 44 learners, and 40 native speakers.

Tt was difficult to have a larger ample of L2 leamers because of the low enroiment in

Arabic language classes. Most North American universities only offcr one class of Arabic,

and the numba of students enroiled in one class usualty does not excecd 10 students.

Another reason is that most leamers who enroUed in Arabic corne fiom an Arabic

background. L2 lemers in this study also represent a smdi group of adult Arabic L2

leamers because oftheir LI. This sample consisted of 38 leamers whose EngIish is their L l

or dominant tanguage, and only six whose L1 is French. Findings fiom this study do not

necessarily apply to leamers of other Lls, especially speakers of Semitic languages whose

word formation processes are close to Arabic. This sample is also Iimited because of

Iearnen' educationai background, level of L2 vocabulary knowledge, and learning

situations. What is characteristic of these leamen may not be generalizable to leamen of

different educational backgrounds in other leaming situations.

7.2.2 Instnictionrl Differences

Another limitation can be ascribed to instructional differences among the classes

within the university Arabic programs that may have affected these leamers' petforniance.

Leamers who participated in the study were exposed to two different instnictioaal methods:

One was fortnailysriented instruction and the other was a communicative teaching methd.

This factor of instruction might have been one cause for the wide variability in leamers'

coinage of lexical innovations and thcir use and choice of word formation processes.

7.2.3 Lexical Items

A f i limitation of the study is the selection of lexical items in the vocabulary

knowledge test and production task, With regard to the vocabulary knowledge test, one of

the problems encountered was a lack of recent MSA tiequency books to choose firom. For

example, for 5,000 word levels, only one old word-counts book was available, the one by

Landau (1959). Therefore, in choosing most of these low frequency words, F had to depend

on my teaching experience. Aiso, since the main purpose of the study was to examine the

use ofword formation processes in expressing new meanings, items in the production task

were carefhlly chosen to trigger innovations. In spite of making some misions in the

production task to replace most low fiequency v d s in the stimuli by higher fiequency

verbs, and to replace general verbs by more specific ones, a few problerns still lingered with

some of the items and seemed to cause difficulty for the L2 Iearners, as indicated by their

failure to coin innovations and leave blauks in some instances. Some items aiso led to

suppletives on the part of native speakers, especially locative items and causatives. It would

have been preferable if ail items selected in the production task had led to innovations. ïhis

was also a problem in Olshtain's study where instniment items Mggered real and existing

items.

7.2.4 Iiistrumcnts, Data Collection and Proctâan

The limitations of instruments, data colIection and procedure tùrther M t the

generalizability of the resutts of this present study, and necd ta be taken into accaunt. The

vocabulary kaowledge test is not a grnerai measure of L2 pmficiency, and measures only L2

learners' receptive knowledge of lexical items out of cornm. The mdy also comisted of a

production task and two comprehension tasks b a d on multiple choice items. The aim of

these tasks was to examine the leamers' use of word formation processes both productively

and receptively. They do not provide any information on how the leamers went about

forming or selectiag their innovations.

Tbe procedure undetaken to collect the data alsa limits the generaiizability of the

study. Even though tasks were performed in the classtoom under my supervision, it was

hard to assure independent individual performance. Six copies wen found containing

identicai answers. In the snalysiq I had to elminate t h m test sets (one fiom each pair),

keeping ody t h copies. It is evident that these students were &the? shanng m e r s

between them or copying fiom each other. This may bc due to the level of diiculty of the

task or to the fact that the classroorns where the tasks were conductesi were small and

crowded, and students were Sitting close to each other, and copying fiom each other.

In spite of these limitations, the findings of this study have implications for both

pedagogy and L2 research. These implications wiil be discussed in the following

paragraphs.

7.3 Implications of the Study

7.3.1 Pedagogicai Implications

What implications do the findings of the present study have for the teaching of MSA

vocabulary knowledge in general, and word formation in particular? ïhe findings have

demonstratecl that some L2 leamers were able to coin new lexical items by drawing upon

compounding, the most typid and available word formation process in their basic stock.

These leamers were the ones exposed to a communicative teaching method. The findings

also showed that some of L2 leamers possessed some knowledge and ski11 in coining

innovations through derivation, and did bettn than the other leamers . These leamers were

the ones exposed to formai instruction. Thex Iearners possess same knowkdge of the

derivational system of MSA h s e they were tau@ lexical rules systematically in the

course. An examination of the materiah used in class confiied that lemcts were in fact

taught derivation in MSA But thci ability to use this pmccss was limited to the production

of infixai andor geminated deriveci p m m only.

if we consid4 that tbis dissrrpa~cy in use ofword formation processes by leamers

in different classes is likely atüibutabie to rnethiod of instniction and amount of L2 exposure,

it can be argued that a rnahod oftachhg MSA word fomiation processes is needed. Should

it be implicit or incidentai instnccb'on w k e "learniag caa occur d e n one is using language

for communicative pwposes". and where words are exposed in a variety of contexts?

(Schmitt, 2000, p. 120) Or should it fx explicit teaching of vocabulary that focuses attention

directly on the forms?

Cunent research advances that both explicit and incidentai teaching of vocabulary

are necessary, and complement each other (Soiunen, 1997; Schmitt, 2000). Accordhg to

Schmitt (2000), most m u e n t words in English "make excellent targets for explicit

attention" (p.121). However, Mequent words are best taught through an incidental

approach. Ellis (1990) adds that through f o d instruction lemers become aware of

specific features of the target language and f o n explicit representations of what they are

introduced to and tau&

As the Arabic lexicon is largely formeû through derivational rules governing root

and pattern combinations, aquuing the well-established lexicon of Arabic involves learning

to use these des . In this snidy, the majority of learners lacked an understanding of Arabic

structural rules, theu constraintq and the association between pattern and meauings,

because, 1 would argue, they had not been introduced to them. This study suggests that MSA

word formation should be included in a teaching approach geared to providing learners with

a knowleâge of the relevant Iexicai mies and formal structures, raising theu consciousness

of these rules, increasing their linguistic awareness, improving theu word anaiysis skills, and

enabling them to attain proficiency in the target Ianguage.

In the case of MS4 word formation processes should be introduced and expiained

at an eariy stage of acquisition with the most fiequent devices and pattenu, followed

gradually by the less mueni ones. Teaching shodd first focus on forms that express one

meaning. Once Imers master one meaning of a pattern, tùey can be introduced tu other

meanings, and exceptions.

But explaining and presenting rules can sometimes lead to boredom. 1 argue that

combining structural and communicative methods of teaching word formation processes

would help leamers develop their knowledge of f o n d structures and also extend their

lexicon, interactive and communicative tasks can be as effective as formal instruction in

promoting a significant amount of noticing (Fotos, 1990, p.385). Many researchers such as

EUis (1990) and Lightbown (1992) argue that it is necessary for leamers following explicit

instruction to be exposed to communicative input that contains the target forms. This will

allow learnen to maintain proficiency gains and improve their accuracy.

Vocabulary knowledge and lexicai niles should not be taught separately from

meaning and context, According to Richards (1976)- vocabulary knowledge entails knowing

how ofien words occur, their use and appropriatemess in different contexts, and their

underlying form and derivations. Leamers have a better understanding of the meanings of

words and their use when thcy are met through a variety of acîivities and in différent

contexts (SOhen, 1997, p.241). It is argued that mrds should be taught in a way that

makes clear the rdationship between forms and meanings: for example, teaching prefixes,

suffixes, mots and ways of forming words by providing students with a set of diffcrent word

families. By learning word families "students l e m not oniy a set ofwords, but something

about those words tbat d l e s them to independently improve their word meaning store,

accomplishing two important goals at once" (Stahi & Shiel 1992-p.226).

7.33 Implications for Furtber Resnrcb

Since this is the first attempt ta study the acquisition of MSA word formation

processes by adult L2 leamas, more data is n d d to test the validity of the tïndings and

theu generalizability. There is a need for mer research to chart the order of acquisition of

diffmnt word formation processes, ad to identify fiirther principles wid strategies tht

lemers may rely on in the course of l m i n g such processes in MSA.

Thete is a need to repaît this study with adult lmers of Arabic with a wider range

of vocabulary knowledge levels. Further research is also needed to gather data fiom a

variety of languages wiih similar or different morphologies and patterns of word formation

and productivity to End out whether the same acquisitional principles also account for their

acquisition by diierent L2 leamers. This study of word formation by L2 leamers of Arabic

should be replicated with learners whose LI is a Semitic language with a typologid

structure similar or close to Arabic in order to examine in depth and pume iùrther the factor

of Ll transfer. Moreover, there is a need to investigate word formation processes via other

data collection instruments and methods such as spontaneous speech, and oral elicitation

techniques, in order to assess the & k t s of diffkrent data collection rnehâs on leamen'

respnses , to find out what ie~ll~lers do and do not h w about word formation processes at

different stages, and how they go about quiring word formation procasa. Finally, there is

a d for longhd'mal research involving more and varied semantic nom and verbal

notions in order to chart L2 leamers' orda of acquisition of différent won! fonn types in

MSA and their mastery of the conventions that govern tkir use.

7.4 Central Conclusion

This study-set out to investigate the MSA word formation processes used by L2

leamers. It also aimed to detemine which of the acquisitional principles (productivity,

semantic transparency, formai sirnpticity, wnventionality, and transfer) proved important in

these leamers' use and choice of MSA word formation devices. The quantitative and

qualitative analyses 1 have presented in this study reveal significant quantitative and

qualitative differences between L2 leamers' and native speakers' production of lexical

innovations. They also reveal differences between L2 leamers and native speakers in their

choice of word formation devices, with L2 Iemers exhibiting a highn preference for

compounding while native speakers p r c f d derivational processes, mainly Vinf+gem.,

and amxation. The study also revealed that neither the higher nor lower level L2 lemers'

use and choice of word formation processes approached that of native speakers. In addition,

these findings provided insights about L2 lemers' relative fiequency of use of MSA word

formation processes, Furthmore, they provide teachers with usefiil information about

principles that corne into play when L2 learners fom new words, and with preliminary

insights for the teaching of MSA word formation processes.

As the first atternpt at d n i n g word formation processes of MSA, a Sernitic

language, this study contributes to the domain of word formation acquisition, adding

infamtioa to Olshtain's study which looked at Hebrcw, another Semitic language, and to

Brader et al.3 research. It provides us with an initial understanding of how word formation

processes are used by leamers of Arabic L2. Findiigs kom previous MSA snidies on Ll or

L2 acquisition of word formation do not exist to support or nui counter to the resdts

obtained in this study. There is no doubt more research will be undenaken on the L2

acquisition of Arabic leicon and word formation processes.

REFERENCES

Abboud, P., Altoma, S., Envin, W., McCanis, E. & Rarnmuny, R (1971). Modem standmd Arabic intemediate level. Part 1. AM Arbor, MI: Inter-University Cornmittee for Near Eastern Languages.

Adams, V. (1973). An introductian to moclern English wordfinnation. London: Longman.

Adjemian, C. (1983). The transferability of lexical properties. In S. Gass & L. Selinker (Eds.), Lungi~age transfer in lmguage leming (pp.250-268). Rowley, MA: Newbury House.

Al-Husari, AS. (1958). Ara7 wa A htiaiflal-Lughnh wa-af-Arab. Beirut: D& ai-Tilm li-al- Malayin.

AI-Karmali, A. (1932). AgMP Lughawijym al-Aqdimih Baghdad: Matbarat al-Aytâm.

Al-Kawakibi, S. (1959). MustaIaM f i lmim. Damascus: Damascus University Press.

Al-Ma@xïM, A. (1947). Kit& al-Ishtiq* wu ut-taM (Book of derivation and arabicization) (2* ed.). Cairo: Matbaht lajnat at-ta?W wa aarjama wa nnashr.

Aweiss, S. (1993). Cognitive processes in foreign language reading: reasoning operations, comprehension monitoring, strategy use, and knowledge sources. Al-ArabMa, 26, 1- 17.

Badry, F. ( I 983). Acqisitiun of lexical deriilatiOClOI d e s in Mwoccan Arabic: implicatiom for the devefopment of standard Arabic as a secorrd Ianguuge through lireracy. Unpublished Ph,D. dissertation. Berkeley, CA: University of California-

Bahat, E. (1986). Ttse acquisition of word fiKrnation &vices in a second Imiguage in Hebrew. Unpublished Master's thesis, Tel Aviv University.

Bauer, L. (1983). English Word fbmwion. Cambridge: Cambridge University Press.

Berman, R, Hecht, B., & Clark, E, (1982). The acquisition of agent and instrument noun f o n s in Hebrew. Papers ami Reports on ChildLangisage Development, 2 21,1624.

Berman, R A, & Cohen, S.R (1984). Roductivity and memory for newIy foned words. Journal of Child Lunguage, 121,611-625.

Bennan, RA, & Sagi, Y. (1981). Word formation and innovation in young children. Hebrew Computationai Linguistrcs, 81,32-62.

Bowman, M. (1974). Learning the structure of causative verbs: A study in the relationship of cognitive, semantic and syntactic development. Pupers a d Reports on Child Development, 8,42- 178.

Broeder, P. (1991). Talking about people. A multiple case study on adult language acquisition: European Studies on Multilingualisrn, AmsterdamlLisse: Swets & Zeitlinger.

Broeder, P., & Extr& G. (1991). Acquisition of kinship reference: A study on word formation processes of adult language learners. International J o u d of Applied Litrguistics, 1 (2), 209-27.

Broeder, P., Extra, G., & van Hout, R (1989). Rocesses in the developing lexicon of adult immigrant learnen. in R. Carter, & P. Nation (Eds.), Research on vocobulary acquisition: An updae (pp.86-109). Amsterdam: Free University Press.

Broeder, P., Extra, G., & van Hout, R (1995). Word-formation processes in adult laaguage acquisition: A multiple case study on Turkish and Moroccan leamers of Dutch. In K. Sajavaara, & C. Fairweather (Eds), Apprtmches to second language acquisition (pp. 1 5-24). J ~ k y I d Cross-imqpage Shrdies 1 7, 1 5-24.

Broeder, P., Extra, G., van Hout, R Sîromqvist, S., & Voionmaa, K. (1988). Word- formation in tuiking about entities in processes in the developing lexicon. Final Report, Vo1.m European Science Foundation, Strasbourg: Tilburg and Goteborg.

Broeder, P., Exira, G., van Hout, R, & Voionmaa, K. (1993). Word formation processes in talking about entities. in C. Perdue (Ed.), Aduit language acquisition: Cross- linguisncper~pective (pp.41-72). Cambridge, UK: Cambridge University Press.

Bnistad, K. (1997). AI Kïtab fi tabllum al- Arabijya= A rexbook fot Arabic. Part ii. Washington, D.C.: Georgetown University Press.

Chejne, Anwar, G. (1%9). 7k Arabic language and its role in history. Minneapolis: University of Minnesota Press.

Cîark, E.V. (1980a). Convention and Uuiovation in acquiring the lexicon. Pupers ond R e m un C M tcmguage LkveIopment, 19.1-20.

Clark, E.V. (1980b). Productivity and child creativity in the lexicon. Unpublished paper presented at the 22nd International Congress of Psychology, Leipzig: DDR July 1980.

Clark, E.V. (1981a). Lexical innovations: How children leam to create new words. in W. Deutsch (Ed.), The child's construction of langiruge (pp.299-328). London: Academic Press.

Clark, E.V. (1981b). Negative verbs in children's speech. in W. Klein, & W.J.M. Levelt (Eds.), Cross@ the b t n ~ a e s in linguistics (pp.253-264). Dordrecht, HoUand; Boston: Reidel.

Clark, E.V. (1982). The young word-rnarker: A case study of innovation in the child's lexicon. In E.Wamer, & L-R Gleitman (Eds.), Languuge acquisition: The state of the rnr (pp.39û-425). Cambridge: Cambridge University Press.

Clark, E.V. & Berman, R (1984). Structure and use in the acquisition of word formation. ianguuge, 60 (3). 543-589.

Clark, E.V., & Hecht, BE. (1 982). Leaniing to coin agent and insrnent nouns. Cognition, 12, 1-24.

Clark, E.V., & Cohen S. R (1984). Productivity and memory for newly formed words. Jmrntrl of Child language, l I , 6 1 1-625.

Dresser, W. (1981). On word formation in natural morphology. Wiener Linguistick Garene, 26,3-13.

Droz& L. (1967). Compounding as a second-word order-formational procedure in modem written Arabic. Asian & Afican Studies,3,60-97.

Drozdik, L. (1979). Lexical innovation through bomwing as presented by scholars. Asian dC Afican Shrries. 15,2 2-29.

El-Khafaifl , H. M . (1985). The role of the Caro A c d n t y in coining Arabic scientijc tennindogy: AR hisrorid ard linginstic ewiuation Unpublished PhD. dissertation. Salt Lake City, Utah: University of Utah.

Ellis, R (1990). Tnrtrircted second I;angirage acquisition i e d n g in t k cfizssroom. Oxford: Blackwell

Eiiis, R (1994). ïïte study of second hguage acquisition. London: Oxford University Ress.

El-Mouloudi , A. B. (1986). Arabic langiuge planning: The case of lexical modernization. Unpublished Ph.D dissertation WashingtoqD-C.: Georgetown University.

El-Tikaina, 1. H. (1982). A Iexicai appuach to the Arabic verb conjugations. Unpublished PhD. dissertation. Washington, DC: University of Washington.

Fakhri, A. (1984). The use of communicative strategies in narrative discourse: A case study of a leamer of Moroccan Arabic as a second language, h g u a g e Learning, 33, (3), 15-37.

Fischer, W. (1997). Classical Arabic. In RHemon (Ed.), The semitic Iruigirages (pp. 18% 2 19). London: Routledge.

Fleisch, H. (2968). L'arabe classique: Esquisse d'une smcîure linguistique. Beyrouth: Dar El-Machreq Editeurs.

Fotos, S. (1993). Consciousness raising and noticing through focus on form: Grammar task performance versus fonnd instruction. Applied Linguistics, 14,385-407.

Fromki V., Rodman, R, Hultin, N., & Logan, H. (1997). An intruduction to language. Canada: Harcourt Brace.

Fück, J. (1 95 1). Al- iàrabiyu, Dirasutfi al-lught wa~l-Iahajat w a ~ l ~ s 4 l i b . Trmlated by TA&l-Halim al-Najcp.Cairo: M a t h fat Dur al-Kitab al- TArabi.

Gass, S. (1979). Language transfer and universal grammatical relations. Language Learning 29,327-44.

Gass, S. (1980). An investigation of syntactic transfer in adult second language learncrs. In R Scarcella, & S. Krashen (Eds,) Research in second language acquisition (pp. 132- 1 JI). Rowley, Mass: Newbury House.

Guilbert, L. (1975). La créarnrité lexrexrc41e. Paris: Larousse.

Harnmoud, M.S. (1982). Arabicization in Morocco: A case study in language planning and ~ ~ a g e poiicy dt&~S Unpublished Ph.D. dissertation. Ausin, Texas: University of Texas.

Hamzaoui, R (1975). Lra&tnie rde h langrie arabe dir Caire: Histoire et aeuvre- Tunis: l'Université de Tunis.

Holes, C. (1995). Modern Arabic saiccnnes, finctions and vmeties. New York: Longman.

Irshled, O., & Whelan, P. (1988). Exploring the dictionary: On teaching foreign leamers of Arabic to use the Arabic-English dictionary. S d i e s in Linguistic Science, 18 (l), 61- 75.

Jackson, H.,& Ze Amvela, (2000). Worak, meaning, and andbu/q: An introcfirction to modem English lexicology. London: New York: CasseIl-

Katamba, F . (1994). E~iglish worak. London: Routledge.

Kellerman, E. (1979). Tramfer and non-transfer: Where are we now? Siudies in Secotd Langucrge Acquisition 2,3747.

Kennedy-Jonker, A (1984). Bilinguism and lexical innovation. in D. Singleton, & D. Little (Eds.), Languc7ge Ieming in fmal and ijomul contexts (pp.8 1-88}. Dublin: Irish Association for Applied Linguistics.

Khaldieh, S. A (1 99 1). PhomIogical and visual processes in word recognition by American leamers of Arabic as a foreign fanpage. Unpublished PhD. dissertation. Detroit: Wayne State University.

Kouloughii, D.E. (1991). &sic lexicon of M d m S- Arabic. Paris: Editions L'Harmattan.

Landau, J. (1959). A word count of Modern Arabic prw . New York: Arnerican Council of L e d Socieîies.

Leclerc, J. (1989). Qu 'estce que lu langue? Lavai, Quebec: Mondia

Lightbown, P. (1992). 'What have we here" Some observations on the influence of instniction on L2 leaming. In R Phillipson, EKeHennan, L. Selinker, M. Sharwood Smith, and M. Swain (Eds.), Foreign lmguage perlagogy research: A commemiiw wlume for Clairs Faerch. Clevedon: Muitilingual Mattas.

Marchand, ?L (1%9). 7k categwies and types of presertl* English w o r d f 0 m . m (2nd ed.). Muaich: Be&

Matthews, P.H. (1974). Morphology: An r n w ~ c t i o n 10 the ikory of word-structure. Cambridge: Cambridge University Press.

Monteil, V. (1960). L'urabe mordente, Paris: Librairie Klincksieck

Mustafa, lawad. (1955). Al-Mabah'r al Lughawiyfi f Gr@. Cairo: Arab League, Institute of Advanced Arabic Studies.

Nation, S. (1990). A vocabulary levels test, Appendix 8. In S. Nation (Ed.), Teaching & Teahhg d iervning voc4buhry. @p.216-272). Boston, MA: Heinie & Heinle.

Odlin, t (1989). Language transfer- Cambridge: Cambridge University Press.

Odlin, T. (1990). Wordsrder transfer, metalinguistic awarness and wnstraints on foreign language learning. in B. VanPatten, & J. Lee (Eds.), Foreip lunguage leaming: A research perspective. Clevedon, Avon: Multilingual Matters.

Olshtain, E. (1983). Sociocultural competence and language transfer: the case of apologies. In S. Gass, & L. Selinker (Eds.), Lunguage transfer in fmguage lemning. Rowley, MA. : Newbury House.

Olshtah, E. (1987). The aquisition of new word formation processes in second language aquisition. Studies in Second Lunguage Acquisition, 9,22 1-23 1 .

Quitregard, D. (1994). Arabic keyworcircir Cambridge: Oleander Press.

Reguigui, A (1994). La créativité lexicale en terminologie arabe. H u m Sciences Monograph Series. Sudbury, ON: Laurentian University.

Richards, I.C. (1976). The d e of vocabulary teaching TESOL Quarterly 10(1), 77-89.

Schachter, J. & Bterford, W. (1979). Discourse fùnction and language transfer. Wwking Pupers cin Bilingualim, 19,342.

Schmitt, N. (2000). Vocabuliay in longiage teaching. London: Cambridge University Press

S h m o o d Smith, M., & Kellerrnan, E. (1986). Crosslinguistic influence in second laquage aquisition: An introduction. in E. Kellerman, & M. Sharwood Smith (Eds.), Crosslingrristic injluence in second langi~ge acquisition (pp. 1-9). Mord: Pergamon Ress.

S6kmeq A.J. (1997). Current trends in teaching second language vocabulary. In N. Schmitt, & M. Mccarthy (Eds.), Vocabulq: Description, aquisition and pedagogy (pp.237- 257). Cambridge: Cambridge University Press.

Stelkwych, Y, (1970). The M d r n Arabic literary Imguage: Lexical and stylisric rlewlopmenix Chicago: Centre for Middle Eastern Studies of the University of Chicago.

Staht, S.A., & Shiei, T,G.(1992). Teaching meaning vocabulary : Productive approaches for poor teaders. Redw ord Writing Quurter&, 8 (2), 223-24-

Suleiman, Y. (1991). AtTective and persona* factors in learning Arabic as a foreign language: A case study. AI Arabjya. 24,83-110.

Tyler, A.& Nagy, W. (1989). The acquisition of English derivational morphofogy. .Tourna1 of Memory and iunguage. 28,649-667.

Wehr,H. (1943). Entwicklung und traditionnelle Pflege der arabischen schtiftsprachender Gegenwart. ZDMG, 97 (l), 16-46.

Wise, K. (1997). The vocabulary of modern French: Origins, smchrre and ficmtim. London: Routledge.

Wright, W. (1964). A grammm of rhe Arabic language. Cambridge: Cambridge University Press.

CONSENT FORM

1 am writir~g to ask you to participate in a study that L di conduct in the Depamnent of Middle East and lslamic Studies where you are taking Arabic as a second ianguage.

I am a PhD. candidate in the Department of Curricuhim, Leamhg and Teaching, at the Ontario Institute of Studies in Education, University of Toronto. My area of interest is vocabulary learning in Arabic, and this particular study wiU aim at understanding better how leamers develop th& abiies to fom new words in Arabic.

Eyou accept to participate, you will be asked to 6ll out a background questionnaire, to do a vocabulary knowiedge test and three tasks (production task and two comprehension tasks). This will take between one hour and 1 112 hour of your time altogether. Coniidentiality wiU be en& by remohg your name tiom the task sheets and replacing it with a code known onty to my supervisor and myseif. in addition, the information wiiected wdl not be the basis of any evahiation and wiU not have any Hect on your couse grades. Momer, you are free to withdraw h m the study at any tirne.

Wyou agree to pdcipate in the study, please check ( ) the box below.

( ) Yes, 1 accept to participate in this snidy

Srudent's narile

Class

sisnature

Date

FORMULtURE D'APPROBATION

C herIC hère étudiant(e),

Je vous écris pour vous demander de participer à une enquête que je vais mener au département des langues Mvantes et Littérature dans lequel vous apprenez !'arabe comme deuxième langue.

k poursuis des h d e s de doctorat en didactique appliquée au département du "Curriculum", a lMtut d'études péâagogiques de l'Ontario, Univdté de Toronto. Mon champ d'iiérêt est l'acquisition du vocabulaire en arabe et ma recherche tente d'analyser la @on dom les apprenants développent M e t é de former des mots nouveaux en arabe.

Si vous avez l'obligeance d'accepter de prendre part a ma wherck, il vous sen demandé de remplir un questionnaire de renseignements personnels, de participer a un test de connaissance lexicale et d'accomplir trois taches (une de production et deux taches de compréhension). Cela prendra entre une heure à une heure et demie de votre temps. Je tiens à souliguer que la confidentialité sera respectée. Votre nom sera remplacé par un code connu seulement a ma directrice de thése et a moi-même. Veuiüez égaiement noter que les renseignements rassemblés ne feront partie d'aucune évaluation dans votre cours et n'ail'ecterorit en rien vos notes. U est important aussi de souligner que vous serez libre de vous retirer de la recherche à n'importe quel moment.

Si vous acceptez de participer a cette recherche, veuiiiez avoir l'amabilité de cocher la case cidessous et de donner les irifodons demandb.

Je vous prie d'agréw mes pius sincéres remerciements.

Rabia Redouane Étudiante en doctorat

[ ] Oui, ]accepte de participer a votre recherche.

Nom de Fétudiant (e)

Date

CONSENT FORM

Dear Teacher.

I am writing to ask for your permission to approach students who are taking Arabic as a second language in your department as subjects for my doctoral researck

I am a Ph-D, candidate in the department of Curriculum, Learning and Teaching at the Ontario lnstiMe of Studies in Education, University of Toronto. My area of interest is vocabulary learning in Arabic, and this particular study will aim at understanding better how leamers develop their abilities to fOnn new words in Arabic.

Students will be asked to fil1 out a background questionnaire, do a vocabulary knowledge test and b é e tasks (production task and two comprehension tasks) which will take between one how and 1 112 hour of their time dtogether.

Confidentiality will be enwed by removing shrdents'narnes fiom the task sheets and replacing them with codes known only to my supervisor and myself My supervisor and [ will be the only people with access to tfme information coilected in this study, and w infornation coilected about individual leamers will be the bais of any evaluation of individuals or of the teaching of Arabic in your department-

As a token of graîitude, you wiH have access to the final findings at the completion of the study.

if you agree to allow me to conduct the study in your department, please check ( ) the box below.

Sincereiy,

Rabia Redouane PhD. student

( ) Yeq 1 accept ta allow you to conduct your study in ow department

Tacher's name

Signature

Date

FORMIiLALRE D'APPROBATION

Cher professeur,

Je vous écris pour vous demander la permission de m'adresser aux étudiant(e)s qui apprennent l'arabe comme deuxième langue dans votre (vos) cours en vue d'obtenir leur éventuelle participation à mon projet de recherche de doctorat.

Je poursuis des études de doctorat en didactique appliquée au département du "Curriculum", à l'institut d'études pédagogiques de POntario, Université de Toronto. Mon champ d'intérêt est l'acquisition du vocabulaire en arabe et ma recherche tente d'analyser la façon dont les apprenants développent l'habileté de former des mots nouveaux en arabe.

Les étudiant(e)s devront remplir un questionnaire de renseignements personnels, participer à un test de connaissance lexicale et accomplir trois tâches (une de production et deux tâches de compréhension}. Cela prendra entre une heure a une heure et demie de leurs temps.

La confidentialité sera respectée. Les noms des éhidiant(e)s seront remplacés par des codes connus seulement a ma duectrice de thése et a moi-mime. Nous serons les seules a avoir accès aux informations rassemblées dans cette recherche, et aucun renseignement rmr les individus ou sur Penseignement de l'arabe dans votre département ferci l'objet d'une évaluation

En guise de remerciement, les résultats fimw de ma recherche seront mis à votre disposition une fois la thése est terminée.

Si vous acceptez que ma recherche soit menée dans votre (vos) cours au département, veuillez cocher la case ci-dessous.

Se vous prie d'agréer mes plus sincères remerciements.

Rabia Redouane Étudiante en doctorat

f ] Oui, vous pouvez c0duke votre recherche dans mon (mes) cours.

Nom du professeur

Signature

Date

Vocabulary Knowledge Test

In this test, you are asked to choose the appropriate w d or expression to go with each meaning. Write the letter of that word next to its meaning.

The 500-word Ievel

new public portion

director j + ~ --- patriotic *-h J - head 3 :Cd ---

to corne to permit to begin to by to declare to fa11

generaI bais recent area Part city

leader generous president nationa Est s e m t rider

* nic gfosscs are not includcd in thc onginal t a , bui thcy arc includcd hcrc hr thc rcadcr.

The 1,000-word level

letter instrument relatives

leadership vacation victory

to watcli to create to settle

family -\ --- machine ~ i - +

j - - MesSage vu& -U

\ ---- project cj+ -A report ‘j& -E counsel + -c

boliday é j ~ ! -\

; 1 ! ; 4 ---- descent la.'-. -*

2 \ L P ---- authonty ;A + -u

j + ---- succes ,+A -&

to observe &-y -i l A\ 2 ---- to engage oneself +J ! -+ 6 \ % --- to believe -1 -a ,+ ---- to establish f \ - A

to issue J u i - c to bless witli G ~ J -c

The 2,000-word level

to tire out \ -1 to infer J- ---- to accuse -+J to bring an accusation 1 9 ---- to inform +I -il

+ ' ---- to conclude + -A

to Say farewell A , -c to distress

sensation celebrity cure

The 3,000-word level

tranquil traveler murderer

joy gr o h dispute

glory Llu~- I ---- recovery

---- struggle c ---- deception

feeling honesîy

lost tourist killer calm unsuccessiül G LJ -c thet G b - c

worry & y A 4 prospenty b 2 j ! -* conflict UU-û

happiness c+ -ù ambition c, - L -e sadness 4 -C

to destroy , u -\ to give one's word La; ---- to wani J L - ~ to awaken .LLj ---- to get along ?-i ! IS

to min Y ---- to handle J-+ J -CI

to promise L ~ _ J -c to restrict -c

The 5,000-word level

balanced $ b -\ excellent * * -

\ , - . \ ---- staying behind L a 1 - - a -U

fdlingbehind + li. --- developed ? ' z 7 A - a

civilized J~~ ---- outstanding j L -&

respected ?* -E distinct J : * 7 A - P L

promotion appointment arrangement

old prisoner noble -

order nomination reduction glorification establishment explanation

jailed d+~ -1 ---- laborious -U

+ i ----hi&-minded Ju~ -&

+ ---- --Y ancient - d.ir -a eloquent J+ + - E offensive 4 -c

TEST DE CONNAISSANCE LEXICALE

Veuillez choisir le mol ou I'expression qui va avec chaque sens. Ecrivez la lettre de ce mot a cote du sens approprie.

Les mots de niveau 500

Lm mots âe niveau 1000

Les mats de niveau 2000

Les mots de niveau 3000

Les mots de niveau 5000

WORD CREATION

In this ta& you are asked to invent a new word or phrase in ARABCC for the English part in bold.

l/ in Arabic. what would you c d a pmoa who sels firrs?

21 In Arabic, how would you express tbc action of makiag a blind pcmn dmt a cafl

31 In Arabic, what would you cal1 a machine that makes a disturbing noise?

41 In Arabic, what would you cal1 a place when people make soap sculptures?

51 in Arabic, what would you cal1 a person who m a h a praetice of bunting balloons?

61 In Arabic, how would you express the action of m a k g somcbody land on anothcr phnet?

7/ ln Arabic, what would you dl a place when people taste jam?

81 in Arabic, what would you dl a machine tbat lights candlts?

91 In Arabic, what would you dl a pemn who spreads rumon by phone?

101 in Arabic, what would you cal1 a machine tbat maices paptr strearnen?

111 In Arabic, how would you express the action of making somebody nia backwards?

121 In Arabic, what would you cal1 a place when you watcb clouds?

131 In i\rabic, how would you express the action of making somtbody jump from the window?

141 in Arabic, what would you cal1 a person who draws for childrtn?

15/ In Arabic, what would you call a place where you burn clothes?

161 In Arabic, what would you cd a machine that curls butter?

171 In Arabic, how would you express tbt action of making somebody scatter buttons on the tloor?

181 in Arabic, what would you call a place whem p p k light drtams?

19/ In Arabic, what wouid you cal1 a penon wbo puus wagons?

201 In Arabic, whaî would you cal1 a place where birds fty and swîm at the srme tirne?

21/ Ln Arabic, what wouid you cal1 r machiut tbat cuts pumpkins?

173

22/ In Arabic, how would you express the action of making somtbody gather seasheh?

231 In Arabic, what would you call a machine that stops wind?

241 In Arabic, what would you call a penon who shows people h m to dream?

CREATION DE MOTS

Veuillez inventa un nouveau mot ou une phrase en ARABE pour la partie écrite en fmw eu caractères gras.

[l En arabe, comment appeleriez-vous une penonne qui diurne des incendies ?

21 En arabe, comment exprimeriez-vous l1acîioa d'amener une personne avtugit ;1 conduire une voiture ?

3/ En arabe, comment appeieriez-vous une machine qui frrit ua bruit perturbateur ?

- -

4/ En arabe, comment appeleriez-vous un endroit ou les gens font des sculptures en savon ?

51 En arabe, comment appeleria-vous une personne qui s'occupe ;I f i n crever des brlloas ?

61 En arabe, comment exprimeriez-vous llaetion d'amener qudqu'un P. rat& sur une a m planète ?

71 En arabe, comment appeleriez-vous un endroit ou l'on fait goûter de la confiire ?

81 En arabe, comment appeleriez-vous un appareil qui allume les bougies ?

91 En arabe, comment appeleriez-vous une personne qui fait counr des rumeun par tklépboae ?

101 En arabe, comment appeleriez-vws un appareil qui fait da petits rubans de papier enroulis ?

1 11 En arabe, comment exprimeriez-vous l'action de faire counr quelqu'un i reculons ?

1Y En arabe, comment appeleriez-vous un endroit ou Fon observe les nuages ?

-~ ~-

131 En arabe, comment exprimeriez-~~~~ Faction de faire sauter quelqu'un d'une fenêtre ?

141 En arabe, comment appeleriez-vous une pcnunne qui dtssint pour les enfants?

151 En arabe, comment appeleriez-vous un endroit ou l'on brûle les habits ?

161 En arabe, comment appeleriez-vous un appareil qui fait des spiraies de beurre ?

171 En arabq comment exprimeriez-vaus l'action d'amener quelqu'un A joncher le plancher de boutons ?

I8I En arabe, comment appeleriez-vous un endmit ou la gens combattent les rêvn ?

191 En arabe, cornent appeleriez-vous une penonne qui tire des chariots ?

20/ En arabe, comment appeleriez-vous un endroit où les o h u x voknt et nagent en mimt temps ?

211 En arabe, comment appeleriez-vous an appareil qui coupe les citrouilles ?

22/ En arabe, comment exprimeriez-vous I'acüon d'amener quelqu'un a accumuler d u coquillages ?

231 En arabe, comment appeleriez-vous un appareil qui arrête le vent ?

241 En arabe, comment appeleriez-vous une penonne qui montre aux gens comment rêver ?

GLOSSARY

another

balloon

bird

biind

to burn

to burst

butter

button

candle

car

children

clothes

cloud

tu curl

to cut

disturbing

to draw

to drearn

to drive

to tight

floor

ta fly

to gather

Jam

to jump h m

to land

to light

machine

to make

noise

phone

place

practice

to pull

to run backwards

to scatter

sculpture

seashetl

to set fire

to show

soap

to spread

to stop

streamer

to swim

to taste

time

wagon

to watch

wind

window

accumuler

aiiumer

amener

machine

arrêter

atterrir

autre

aveugie

ballon

bruit

brider

beurre

bouton

bougie

chariot

citrouille

combattre

conduire

confiture

coquillage

couper

courir à reculons

courir des meurs

crayon

crever

dessiner

endroit

enfants

enroulb

faire

Fenetre

gens

goûter

habits

joncher

monter

nager

nuage

observer

oiseau

papier

Pm-

perturbateur

petit

plancher

planet

rëver

ruban

sauter

scuplture

spirale

soap

t l k p hone

temps

tirer

vent

voiture

voler

COMPREHENSION TASK I

In this task, you are asked €0 choose from the four options tbCorm that you judge appropriate and suitable in ARABIC to express the meaning of the English phrase written in bold.

11 In Arabic, how would you cal1 a man who makes bubbles with soap?

21 In Arabic, how would you cal1 a machine that hugs people?

31 In Arabic, how would you express the action of making a man lock himself in a dark room?

i 88

41 In Arabic, how would you cal1 a place where you make Mickey Mouse

51 In Arabic, how would you cal1 a woman who bleaches the floor?

61 In Arabic, how would you c d a machine that kills people

71 In Arabic, how would you cal1 the action of making a woman hit her head with a hammer?

81 In Arabic, how would you cal1 a place where you dry tomatoes?

Arabic, how would d l a man who pulls a car 7'

AG -1

with a rope?

101 In Arabic, how would you cal1 a machine that breaks pencils?

111 In Arabic. how would you cal1 the action of making a man sit on nails?

121 In Arabic, how would you cal1 a place where the air is polluted?

131 In ~rabic, how would you cal1 a woman who bums food?

141 In Arabic, how wauld you cal[ a machine that knocks people down?

151 In Arabic, how would you catl the action of causing a woman to swim in icy watef?

16i In Arabic, how would you cal1 a place where lights spaikle?

17/ In Arabic, how would you cal1 a man who lifts things?

18i In Arabic. how would you cal1 a machine that masspmduces dolls?

191 In Arabic, how would you express the action of making a man slide on a rainbow?

y * ' - * j u s 1 - f -d = bJFi,-2

9 - . * k.9 1 -3

201 In Arabic, how would you cal1 a place where you wipe your feet?

211 In Arabi~, how would you cali a woman who blinks her eyes? J3 , -

Arabic, how would you cal1 a machine used to inspect things?

231 in Arabic, how would you express the action of making a woman count clouds?

241 Ir1 Arabic, how would you cal1 a place where people investigate supematural events?

251 In Arabic, how would you cal1 a man who buries people with their jewelry?

261 In Arabic, how would you cal1 a machine that smooths paper?

271 In Arabic, how would you express the action of making a woman giggle?

281 In Arabic, how wauld you cal1 a place where people sigh?

291 In Arabic, haw would you cal1 a woman who smiles at people?

301 In Arabic, how would you cal1 a machine that grabs things?

194 QUESTIONNAIRE DE COMPREHENSION 1

Veuillez choisir parmi appropriée $exprimer en caractère gras.

les quatres options la forme que vous jugez la plus en ARABE le sens donnd par la phrase franpaise &rite

II En arabe, comment appeleriez-vous un homme qui fait des bulles de savon ?

2l En arabe, comment appeleriez -vous une machine qui enlace les gens ?

31 En arabe, comment exprimeriez-vous l'action d'amener quelqu'un 2 s'enfermer dans une chambre sombre ?

* 41 En arabe, comment appeleriez-vous un endroit ou on manufacture les chemises Mickey mouse ?

51 En arabe, comment appeleriez-vous une femme qui blanchit le sol ?

61 En arabe, comment appeleriez-vous une machine qui tue les gens ?

b

71 En arabe, comment exprimeriez-vous!'action d'amener une femme a se donner des coups de marteau sur la tete ?

/ 81 En arabe, comment appeleriez-vous un endmit où on fait çecher les tomates?

91 En arabe, comment appeleriez-vous un homme qui tire une voiture avec une corde ?

101 En arabe, comment appeleriez-vous une machine qui casse les crayons?

111 En arabe, comment s'asseoir sur des clous ?

exprimeriez-vous C

l'action d'amener un homme a

% f

12l En arabe, comment appeleriez-vous un endroit ou l'air est pollue ?

13/ En arabe, comment appelerîez-vous une femme qui b i l e la nourriture ?

141 En arabe, comment appeleriez-vous une machine qui renverse les gens ?

C

151 En arabe, comment exprimeriez-vous l'action d'amener une femme a nager dans des eaux

- e

161 En arabelcomment appeleriez-vous un endroit ou les lurnieres scintillent?

171 En arabe, comment appeleriez-vous un homme qui soul&t? les choses ?

I + 198 181 En arabe, comr?ent appeleriez-vous une machine utilisee a la grande fabrication de poupees ?

191 En arabe, comment exprimeriez-vous l'action de faire glisser un homme sur un arc-en-ciel ?

\ 201 En arabe, comment appeleriez-vous un endroit ou vous essuyez les pieds ?

211 En arabe, comment appeieriez-vous une femme qui cligne ses yeux ?

2 2 En arabe, comment appeleriez-vous une machine qui examine les choses ?

231 En arabe, comment exprimeriez-vous 1' action d'amener une femme 2 compter les nuages ?

A ?4!En arabe, comment appeleriez-vous un endroit ou l'on enquete sur les evenements surnaturels ?

251 En arabe, comment appeleriez-vous un homme qui enterre les gens avec leurs bijoux ?

261 En arabe, comment appeleriez-vous une machine qui lisse les papiers ?

h

271 En arabe, comment exprimeriez-vous 1' action d'amener une femme a pouffer de rire?

k

281 En arabe, comment appeleriez-vous un endroit ou les gens soupirent ?

291 En arabe, comment appeleriez-vous une femme qui sourit aux autres ?

301 En arabe, comment appeleriez-vous une machine qui empoigne les choses ?

COMPREHENSION TASK II

In this task, you are asked to choose from the four options the f o n that you judge appropriate and suitable in ARABIC to express the rneaning of the English phrase written in bold.

I I In Arabic, how would you cal1 a man who hunts animals?

21 In Arabic, how would you cal1 a hawesting machine?

31 In Arabic, how wwld you express the action of causing a woman grief?

The phonetic tianxription is used here for tfie Rader RI=realitem

4/ In Arabic, how would you d l a place where you lock up jewelry?

51 In Arabic, how would you c d a wornan who cooks?

61 In Arabic, how would you call a machine that steams food?

7/ In Arabic, hon. \vould you express the action of making n man lie7

81 In Arabic, how wodd you call a place whcn you wash your clothn ?

fghasild a d

ij, 5 2 f g h a s ~ i d '" *ae-

irnaghsald ,\ .: f ;-3 4 s /maLaM ghasli lmalabisil , k;i , Q , -4

0 a

91 In Arabic, how wuld you cal1 a mari who govems?

101 In Arabic, how wuld you cal1 a machine that squeezes juice?

1 I I In Arabic, how would you express the action of making a man read a story over and over again?

12/ In Arabic, how wuld you cal1 a place where you swim?

131 In Arabic, how would you caIl a woman who migrates ?

141 In Arabic, howwwuld you cal1 a machine that cuts stones?

151 ln Arabic, how wutd you express the action of making a man drunk?

161 in,Arabic, how would you cal1 a place where you do an experiment?

171 In Arabic, how would you cal1 a man who delivers mail?

181 In Arabic, how wuld you cal1 a machine that sews dothes?

191 ln Arabic, how would you express the action of making a man leam?

201 In Arabic. how wu ld you cal1 a place where you display pictures?

- /maCrid un1 =< C - - 3

c f -=e 0 # * /makanu Tardi niwwan/ JA i cp+ L m - 4 0

21 1 ln Arabic, how would you cal1 a woman who writes books?

221 In Arabic, how would you cal1 a machine that erases mistakes?

231 In Arabic, how would you express the action of making a woman cry?

241 In Arabic, haw would you cal1 a place where animals are slaughtered?

251 In Arabic, how wuid you cal1 a man who works on a farm?

26/ in Arabic, how would you cal1 a machine tnat strains oil?

- -

>r - - - - /saffàyatun/ 2 ;j---d~_2

/misfatudEU 3 -5 3 - Râlatu tasfiyati zayti/ .a. .-2 --

- 9 2 C Ii .. - - ;t_ii -4

271 In Arabic, how would you express the action of causing a man to become infomed? lb&laghahu/ u H ~ r - 1 halleghahd 9 4 7 y 4 /?ablaghahd " 7c i-g- -3 jaiaiahu balighanl f ; , ,;k$ z ;-4 -- 20/ In Arabic, how wuld you cal1 a place where you- pray?

291 In Arabic, how would you express the action of making a woman look better?

3O/ In Arabic, how would you cal1 a machine that drills?

Si En Arabe. comment appeleriez-vous une femme qui fait la cuisine?

L

61 En arabe. comment appeleriez-vous un appareil qui fait cuire a la vapeur?

71 En arabe, comment exprimeriez-vous l'action de faire mentir un homme?

p* @ - * i p -1 3 . - 1,;3

,O . , 'C+ S I - 3 '9 - / & I S ' ~ - 4

81 En arabe. comment appeleriez-vous un endroit ou vous faites la lessive ?

91 En arabe, camment appeleriez-vous un homme qui gouverne?

QUESTIONNAIRE DE COMPREHENSION TASK II

f - Veuillez choisir parmi les quatres qptions la forme que vous jugez la plus appropriee a exprimer en ARABE le sens donne par la phrase écrite en caractère gras.

i l En arabe, comment appeleriez-vous un homme qui fait la chasse aux animaux?

39 \;.!.A-1 -y\ 5 ;a-2

2/ En arabe, comment appeleriez-vous une moissonneuse?

31 En arabe, comment exprimeriez-vous l'action de causer de la peine> une femme?

L

41 En arabe, comment appeleriez-vous un endroit ou vous enfermez les bijoux?

101 En arabe, comment appeleriez-vous un appareil qui extrait le jus?

1 I/ En arabe, comment exprimeriez-vous l'action d'amener un homme areiire une histoire plusieurs fois?

121 En arabe, comment appeleriez-vous un endroit & on nage?

#'

131 En arabe, comment appeleriez-vous une femme qui se deplace d'un lieu &n autre ?

3: e /

0- L A -1 J?- 0 M~

OJ C'==2 5;+lS -3

GIS ! -4 0 /

#

141 En arabe. comment appeleriez-vous une machine qui coupe les pierres? d i;h C;-1

)< . CI)' 1 ,il-; -2

151 En arabe, comment exprimeriez-vous l'action de rendre un homme ivre?

C t 161 En arabe, comment appeleriez-vous un endmit ou on fait une experience?

171 En arabe, comment appeleriez-vous un homme qui livre le courrier?

181 En arabe, comment appeleriez-vous une machine a coudre?

19f En arabe, comment exprimeriez-vous l'action d'amener un homme a apprendre?

L

201 En arabe, comment appeleriez-vous un endroit ou vous exposez des tableaux?

/ 211 En arabe, comment appeleriez-vous une femme qui ecrit des ouvrages?

221 En arabe, comment appeleriez-vous une machine qui efface les erreurs

231 En arabe, comment exprimeriez-vous l'action de faire pleurer une femme

L

24/ En arabe, comment appleriez-vous un endroit ou se fait l'abattage des animaux?

- 251 En arabe, comment appeleriez-vous un homme qui travaille a la ferme?

261 En arabe, cornpent appeleriez-vous une machine qui filtre l'huile?

271 En arabe, comment exprimeriez-vous 1' action de communiquer une information $un homme?

- 28/ En arabe, comment appeleriez-vous un endroit ou on prie?

291 En arabe, comment exprimeriez-vous 1' action d'amener une femme aparahre plus belle?

lZ &G-1 i'.; -2 i< r: ;f-s

4 , - ,-:\;î-:+4

C / f

301 En arabe, comment appeleriez-vous un appareil qui perce des trous?

BACKGROUND QUESTlONNAiRE

A. The foiiowing questions ask about your personal background.

1. Are you ..... ?

2. What is your age group? (Please check one)

3. What is your m e n t student status? (Piease check one)

a. 1st year ofunderpduate program b. 2nd or 3rd year of undérgraduate program c., graduate student d. swai shuknï e. other (please &)

4. Were you boni in C d or in the United-States?

5. If the answer is NO, what is your country of birth?

6. How old were you when you eame to Canada or to the United- States?

7. Which of these languages did you speak fint as a child?

English

French

Arabic

Other ianguage

B. The fdlowing questions ask about OTKER LANGUAGQS) that you h o w and use besides your FLRST W G U A G E .

1. Are there any languages you know other than your FIRST language?

No ifyouanwered NO please go to SECIION C (page 4)

2. if the m e r is YES, what idare thisc these lansuage(s)?

3. Where did you lem thidthese lanpage(s)? (Please check one box for each language)

language a

language b

~~~guage c

4. Can you understand thidthese laquages when WRïITEN or SPOKEN or BOTH? (Please check one box for each language)

language a

language b

~ g u a g e c

5. How oflai do you use thidthese language(s)? (Please check one box for each ianguage)

language a O language b O hguage c O O

6. Where do you use thisithese language(s)?

language r

language b

Ianguage c

At home At school At work

O O O

on trip

1. Do you speak any Arabic dialect?

2. What idare your reason(s) for taking Classicai Arabic? (Pl- check one or more as relevant)

a personal uitnest b. credit murse c. major subject for my graduate study d. have arabophone bac kground e. taking a mp E otite (please specifj.)

3. is Classicai Arabie used in your borne?

0. if you m e r NO go to question 5

4. How o f h is Arabic spoken in your home? (Please check one)

S. in whaî way(s) do y use Arabic? (Piease check as many as relevant)

6. Compared to other languagefs) that you know or you have tearrit, how do you 6nd learning Arabic? (Please check one)

a a very ~ c u l t language b. a &cult language c. neither easy nor diffiailt d. an easy language e a very easy language

7. Accarâhg ta yw, wIiat makes Arabic a diEcult language to lem? (Please check as many as-1

8. Comped to other leamers in your class, how well are ywi doing in learning Arrtbic?

9. Compared to other leanins in your dass, how would you rate your performance in the folowing areas?

10. Can you do the foiiowings in Arabic, and how w d can you do them? (Please check one box in each row)

a talk to a native speaker

b watch and understand a I-V Pwram

c read newspapers and nmpims

c understand a radio program

with much a*

A. Les quesiions suivantes se penchent sur des informations perso~elles.

1. Quel est votre sexe?

Homme 0 Femme 0

2. A quel groupe d'âge apartenez-vous? (ne cochez qu'uw seule réponse)

3. Actuellement quel est votre staftit d'étudiant(e)? (veuillez cocher qu'une d e réponse)

a le année d'un progranime de premier cycle b. 2éme ou 3éme année d'un programme de premier cycle c. étudiant(e) de 2éme ou 3émc cyck d. éaidiant(e) special(e) e. auue (veuillez préclsr)

4. Êtes-vous H e ) au Canada ou aux États-UNS?

Non Si VOUS nipondez OUI v d l e z passer à la question 7

5. Si la réponse est NON, quel est votre pays d'on*?

235

6. Quel âge aviez-vous quand vous &s arrivé(e) au Canada ou aux États-unis?

7. Laqueue de ces hgues avez-vous parié en premier lieu durant votre &ce?

Anglais

Français

Arabe

Auve langue

B. Les questions suivantes se penchent sur les LANGUES que vous connaissez et que vous utilisez a part vom P R E ~ langue.

1. Y a-t-il d'autres langues que vous com*ssez bien, en plus de votre langue pttmière?

Non 0 4 MD NON v d e z passer a la &on C page 4

2. Si la réponse est 0- qude(s) est/wrn cetteks langue(s)?

3 0ù avez-vous APPEüS cettekes ihngue(s)? ( v d e z cocher la case appropriée)

A la maisan A l'écale Autre enviromtement

hguage a

language b

tanguage c

4. Pouvez-vous mmprendre cettelm Iangue(s) quanâ de(s) atrsont PARLÉE(s) ou ÉcMXE(S) w les DEUX? (veuillez cocher la case appropriée pour chaqye langue)

parlée éaite le. deux

tanguage a

language b

language c

A la maison A l'école Au travail En voyage

C. Les questions suivantes se penchent sur la Langue ARABE: les raisons qui vous ont amermé a l'apprendre et votre expérience d'apprentissage.

1. Paria-vous un dialecte arabe?

ûui s i . ~ u s ~ d e z ~ ü ~ préciser lequel

Non

2. Pour quelie(s) raisor@) avez-vous decide d'apprendre l'arabe classique? {vmillez cocher réponse(s) appropriée(s))

a..intérèt personnel b. corn pour obtenir un crédit c. Met obligatoire pour mon programme d'études supnieures ci. origine arabophone e. voyage f autre(veuüiezspéci6a)

3. L'arabe classique est4 Ütilisé par vous a la maison?

Non 0 Si vous répondez NON veuiIlez passer à la question 5

4. Quelle est ia friqwnce d'utilisation a la maison? (veuillez cocher la case appropriie)

5. Dans queI(s) but(s) utilisez-vous L'arabe? (wuillez cocher Mes réponse@) appropriée(s))

a. pour prier à d'autres arabophones b. pour l'utiliser dans mon domaine d'étude et de recherche c. pour regarder Ia télévision d.powhelesjournaux~desmaBapnes e.pourl.hdesoewreslittérairrs E pour écoraer la radio g. autre, veuiltez préciser

6. Comparée a d'ame(s) langue(s) que vous avez apprise($, wmment trouvez-vous l 'appmisqe de Farabe? ( v d a cocher la réponse appropriée)

7. Selon vous, qu'est-ce qui rend l'arabe une langue dHcile a apprendre? (veuillez cocher la ou les réponses approptiée(s))

a grammaire b. vacabdaire c. prononciation d. le système d'écriture e. autre(s), v d e z préciser

8. En vous comparant a d'autres apprenants dans votre classe, où vous situez-vous dans vos résultats?

audesous de la moyenne a la moyenne de audessus de la moyenne de la classe la classe de la classe

9. Par rapport aw autres ehidiams de la cl- queile est, selon vous, votre pnformance dans les domaines suivants?

10. Pouvez-vous accomplir ces tâches en ARABE et comment?

a. parier a un arabophone

avec beaucoup un Peu sans de difficulté de ciifficuite di8iculté

b, regarder et comprendre un programme de TV 0

c. lire Ies journaux et les magazines O CI

d. écrire une lettre O t. comprendre uni?

émission radiophonique

incapable de fàife