UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Subject extraction from embedded clauses in standard Arabic Elesseily, Nagat Hassan 1985

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1985_A8 E44.pdf [ 2.05MB ]
Metadata
JSON: 831-1.0096473.json
JSON-LD: 831-1.0096473-ld.json
RDF/XML (Pretty): 831-1.0096473-rdf.xml
RDF/JSON: 831-1.0096473-rdf.json
Turtle: 831-1.0096473-turtle.txt
N-Triples: 831-1.0096473-rdf-ntriples.txt
Original Record: 831-1.0096473-source.json
Full Text
831-1.0096473-fulltext.txt
Citation
831-1.0096473.ris

Full Text

SUBJECT EXTRACTION FROM EMBEDDED CLAUSES IN STANDARD ARABIC by NAGAT H. ELESSEILY A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTERS OF ARTS in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF LINGUSITICS We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA OCTOBER 1985 © NAGAT H. ELESSEILY, 1985 In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree at the THE UNIVERSITY OF BRITISH COLUMBIA, I agree that the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and study. I f u r t h e r agree that permission for extensive copying of t h i s t h e s i s f o r s c h o l a r l y purposes may be granted by the Head of my Department or by h i s or her r e p r e s e n t a t i v e s . I t i s understood that copying or p u b l i c a t i o n of t h i s t h e s i s f o r f i n a n c i a l gain s h a l l not be allowed without my w r i t t e n p e r m i s s i o n . THE UNIVERSITY OF BRITISH COLUMBIA 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date: OCTOBER 1985 Abstract Standard Arabic exhibits 'that trace' e f f e c t in one instance in the extraction of the subject from an 'anna' clause while the extraction of the object and the subject of an 'an' clause may be exctracted freely in the formation of WH-question. The extraction of the subject of an 'anna' clause may not be extracted unless the extracted position i s marked by a c l i t i c on the complementizer 'anna'. If the c l i t i c appears in place of the moved NP in an 'an' clause i t renders the sentence ungrammatical. The adoption of the Government and Binding Framework, Chomsky (1981), (1982) and in p a r t i c u l a r Case Theory, Government theory and the Empty Category P r i n c i p l e (ECP) enable us to explain t h i s d i s t i n c t behavior in the extraction of the subject of an 'anna' clause and show that the appearance of the c l i t i c i s predicted by the proposed analysis. It is argued that the c l i t i c appears in the extraction of the subject of an 'anna' clause in order to properly govern the trace l e f t by the extracted subject, and so as not to violate ECP. Since verbs are proper governors in SA, extraction of the subject of an 'an' clause must apply from a governed p o s i t i o n . In fact t h i s i s exactly what our analysis predicts. Since 'an' i s not a case assigner and since we are assuming that government and case are assigned only to the right, AGR and verb preposing are obligatory in an 'an' clause to assign case to the subject NP. Therefore extraction of the subject leaves a trace properly governed by the verb. In the extraction of the subject of an 'anna' clause on the other hand, since 'anna' is a case assigner and assigns a cusative case to i t s subject, AGR and verb preposing may not apply. Thus, the extraction of the subject leaves a trace which i s not properly governed in v i o l a t i o n of ECP, and the c l i t i c must appear in order to properly govern the trace l e f t by movement. Acknowledgments: My s i n c e r e s t thanks to my c o l l e g u e s . T h e i r moral support and encouragment were i n v a l u a b l e i n h e l p i n g me complete t h i s t h e s i s . Thanks a l s o to my a d v i s o r , Michael Rochemont and the r e s t of my committee David Ingram and Hanna K a s s i s whose comments I b e n e f i t e d g r e a t l y . F i n a l l y , I want to thank my f a m i l y f o r t h e i r p a t i e n c e and t h e i r u n f a i l i n g support. i v Tabl e of Cont e nt s I. INTRODUCTION / II. AN OVERVIEW OF THE GB FRAMEWORK 4 A. THE GB FRAMEWORK 4 1. THE RULE SYSTEM 5 2. THE SYSTEM OF PRINCIPLES 7 ///. WORD-ORDER AND CASE ASSIGNMENT 18 A. VSO ANALYSIS 18 B. AN ALTERNATIVE ANALYSIS 19 IV. PRO-DROP PHENOMENA AND CLITICS. 27 A. PRO-DROP ANALYSIS 28 B. CLITICS 29 V. WH-EXTRACT ION 42 A. THE PROBLEM 42 B. THE SOLUTION 44 CONCLUSION 54 BIBLIOGRAPHY 56 INDEX 59 v I. INTRODUCTION This thesis is a study of WH-extraction from embedded clauses in Standard Arabic (SA).. The problem to be explained i s the behavior observed in the extraction of subjects from embedded clauses introduced by the 'anna' and 'an' complementizers. If the subject is extracted from an 'anna' clause, a c l i t i c must appear in place of the moved element. This is not the case when the embedded clause i s introduced by an 'an' complementizer. A c l i t i c can not be present in place of the moved noun phrase. Our, goal is to show that the adoption of p r i n c i p l e s introduced in the Government and Binding Framework (Chomsky, 1981, 1982), enable us to explain WH-extraction in SA in a uniform and natural manner. The s p e c i f i c p r i n c i p l e s needed are Government and Proper Government, the Empty Category Pr i n c i p l e (ECP) the Case Theory. It w i l l be argued that a c l i t i c appears in an 'anna' clause in order to properly govern the trace l e f t by the moved element. The extraction of an element that i s not properly governed would be in v i o l a t i o n of the ECP. Furthermore, the fact that the c l i t i c appears in an 'anna' clause and not in an 'an' clause leads us to assume that the subject in 'an' clause i s properly governed. Since only verbs and c l i t i c s are proper governors in SA (as w i l l be shown), the extraction of the subject of an 'an' clause must have applied from a governed post-verbal position. VSO, in fact, i s the word-order required by an 'an' comlementizer. This word order explains why the c l i t i c 1 2 is not required in this case, since the trace w i l l be properly governed by the verb. The presentation of the arguments for the above analysis w i l l be divided into f i v e Sections. Section One is the present Introduction. Section Two w i l l present an overview of the GB Framework, and in p a r t i c u l a r d e f i n i t i o n s of the p r i n c i p l e s that are of dir e c t relevance to our an a l y s i s . Section Three addresses the issue of word order in SA. To explain the extraction of NP's from embedded clauses, i t i s necessary to e s t a b l i s h the underlying word order of sentences in SA. At S-structure, the word order is VSO in 'an* clauses, but SVO in 'anna' clauses. We w i l l argue that the D-structure of SA i s SVO, and that t h i s becomes VSO at S-structure v i a verb movement. Two main arguments w i l l be presented to support t h i s analysis. One is based on evidence from the Binding theory. The other is based on s i m p l i c i t y measures in explaining Case assignment to the subject, since there w i l l be a c o n f l i c t the in assignment of case to the subject of an 'anna' clause. The Case F i l t e r forces verb movement so that the subject, [NP,S], can get Case. Section Four w i l l discuss C l i t i c s and Pro-Drop phenomena. Pro-drop i s t i e d to c l i t i c i z a t i o n . -Both, in fact, involve the empty nominal position, pro. We w i l l argue that the c l i t i c must appear in order to i d e n t i f y the contents of pro as well as absorb the Case assigned to that p o s i t i o n . F i n a l l y , in Section Five, I present a f i n a l set of arguments to show that the d i s t i n c t behavior in the 3 extraction of the subject of 'anna' follows automatically from the proposed analysis. I I . AN OVERVIEW OF THE GB FRAMEWORK A. THE GB FRAMEWORK The GB theory, Chomsky (1981), has developed d i r e c t l y from e a r l i e r work, in p a r t i c u l a r from the general framework of the 'Extended Standard Theory' (EST): Chomsky (1973, 1976, 1977, 1980), Chomsky and Lasnik (1977) and subsequent l i t e r a t u r e . C o n c e p t u a l l y , GB repr e s e n t s a s h i f t from a system of r u l e s to a system of simple and rather n a t u r a l p r i n c i p l e s . Although the p r i n c i p l e s h o l d u n i v e r s a l l y , languages may d i f f e r from each other with respect to t h e i r a p p l i c a t i o n . T h i s i s the theory of parameters i n which languages are the r e s u l t of parametric v a r i a t i o n s . A c l e a r example i s the pro-drop parameter f o r the n u l l s u b j e c t phenomena. Languages that allow m i s s i n g s u b j e c t s such as I t a l i a n and Spanish are set p o s i t i v e l y for t h i s parameter. E n g l i s h on the other hand, i s set n e g a t i v e l y , so the subj e c t p o s i t i o n must have p h o n o l o g i c a l content. The c e n t r a l concern of the GB theory i s to p r e d i c t the d i s t r i b u t i o n of nominal elements, and to determine a typology of empty c a t e g o r i e s . The theory p r e d i c t s i n a p r i n c i p l e d manner whether nominal elements are 1) f u l l l e x i c a l NP's, i e . a f u l l y r e a l i z e d e x p r e s s i o n , l e x i c a l anaphors or l e x i c a l pronouns, or 2) empty c a t e g o r i e s , i e . NP-traces, PRO, v a r i a b l e s (WH-traces) or pro (the m i s s i n g s u b j e c t of pro- drop languages). GB theory d i s t i n g u i s h e s between two p e r s p e c t i v e s , one emphasizes the r u l e system, 4 5 the other emphasizes the systems of p r i n c i p l e s . Both perspectives interact to achieve the goal of the theory. 1. THE RULE SYSTEM The subcomponents of the rule system according to Chomsky (1982) are: (1) A lexicon B syntax i . base component i i . transformational component C interpretive components i . Phonological form component i i . Logical form component Each of these components has a special function. The lexicon s p e c i f i e s the 'inherent' properties of l e x i c a l items, in p a r t i c u l a r properties such as thematic and se l e c t i o n a l s p e c i f i c a t i o n s . For example, a verb l i k e 'persuade' has the property of assigning a certain thematic role to each category i t subcategorizes, i e . i t s object and the clausal complement as in "John persuaded B i l l to leave." The lexicon also s p e c i f i e s properties of phonetic form and meaning that are not determined by rule. The rules of the base generate D-structures through insertion of l e x i c a l items associated with 0-roles into representations of grammatical functions (GF), such as subject, object, etc.. Only GF positions assigned 9-roles 6 are l e x i c a l l y f i l l e d at D-structure. There is also the option of phonetically n u l l PRO. D-structures are related to S-structures e s s e n t i a l l y by a general rule, 'Move a', which allows any category to move free l y . This i s feasible because the p r i n c i p l e s of the GB theory provide constraints on S-structure and on the application of Move a. An element in S-structure bearing a 0-role may move to a position that i s assigned no ©-role, leaving traces coindexed with their antecedent as i s the following example: (2) a. John seems t to have l e f t . i i b. *John wanted t to have l e f t , i i In (2)a. the subject 'John', although i t is ©-marked by the verb 'leave', must move to get Case. Movement i s possible since the verb 'seem' does not assign a ©-role to i t s subject. In (2)b. on the other hand, movement i s not possible since both the verbs 'want' and 'leave' assign ©-roles, thus both positions must be f i l l e d with arguments at every l e v e l of representation. S-structures are assigned a PF representation in the phonological component. They are assigned an interpretation in the LF component. Thus S-structure i s an association between representations of form and representations of meaning, although the mappings of S-structure onto PF and LF are independent of one another. Hence the core grammar i s 7 represented as follows, (Chomsky and Lasnik, 1977): (3) D-structure Move a S-structure PF component LF component 2. THE SYSTEM OF PRINCIPLES The perspective of the GB theory that focuses on p r i n c i p l e s contains the following subsystems: (4) (a) X - bar theory (b) 9 - theory (c) Control theory (d) Binding theory (e) Case theory (f) Government theory (g) Bounding theory Several of these subtheories w i l l be p a r t i c u l a r l y relevant to our analysis s p e c i f i c a l l y , Case theory, Government theory and the Empty Category P r i n c i p l e (ECP), as well as the Bounding theory and in pa r t i c u l a r the Subjacency p r i n c i p l e . 8 The X-bar theory T h i s theory r a d i c a l l y reduces the c l a s s of p o s s i b l e base components. I t expresses the p h r a s a l expansion of any given category by i t s s t r u c t u r a l r e p r e s e n t a t i o n . I t i s assumed t h a t , w i t h i n a maximal p r o j e c t i o n X, there i s a head X and a complement, a s t r u c t u r e of the form X n > X n - 1 , where the c o n s t i t u e n t X n must c o n t a i n as i t s head a c o n s t i t u e n t X n -1 bar. 6 - Theory The ©-theory i s concerned with the assignment of thematic r o l e s ( such as agent, theme, et c . ) to c e r t a i n p o s i t i o n s . The b a s i c p r i n c i p l e of the ©-theory i s the ©-criterion. I t r e q u i r e s that each argument bears one and only one 0 - r o l e ; and that each 0 - r o l e i s a s s i g n e d one and only one argument (Chomsky 1981:36). The ©-criterion e f f e c t i v e l y a p p l i e s to a l l three l e v e l s of r e p r e s e n t a t i o n s : D - s t r u c t u r e , S - s t r u c t u r e and LF. Movement from a ©-position to a ©-position i s blocked s i n c e the element moved would be assigned dual ©-roles ( c f . ( 2 b ) ) . The subject p o s i t i o n of a r a i s i n g verb l i k e 'seem' cannot be a ©-position s i n c e an NP can move i n t o i t as i n : (5) John seems t to be a f o o l . i i The P r o j e c t i o n P r i n c i p l e i s c l o s e l y a s s o c i a t e d with the ©-criterion. I t s t a t e s that r e p r e s e n t a t i o n s at each s y n t a c t i c l e v e l ( i e . LF, D- and S - s t r u c t u r e s ) , are 9 p r o j e c t e d from the l e x i c o n i n that they observe the s u b c a t e g o r i z a t i o n p r o p e r t i e s of l e x i c a l items (Chomsky 1981:29). I t f o l l o w s then that the s u b c a t e g o r i z a t i o n frames must be the same at every s y n t a c t i c l e v e l . In other words, the r e p r e s e n t a t i o n s at each of the three l e v e l s are p r o j e c t i o n s of l e x i c a l p r o p e r t i e s . Both the ©-criterion and the P r o j e c t i o n P r i n c i p l e t h e r f o r e c o n s t i t u t e w e l l formedness c o n d i t i o n s that must be met at a l l l e v e l s of r e p r e s e n t a t i o n s . C o n t r o l theory T h i s theory i s concerned with the p o t e n t i a l of a b s t r a c t pronominal PRO. Wi t h i n t h i s theory, e i t h e r l i n k e d to an antecedent or i t i s assign e d r e f e r e n c e . The Binding theory. T h i s theory i s concerned with the r e l a t i o n of anaphors, pronominals, names and v a r i a b l e s t o p o s s i b l e antecedents. I t provides the grammar with a p r i n c i p l e d way of determining the types of NP's that can appear. The f o l l o w i n g are the bin d i n g c o n d i t i o n s as given by Chomsky (1981:188): (6) Binding Theory (A) An anaphor i s bound i n i t s governing c a t e g o r y . (B) A pronominal i s f r e e i n i t s governing category. (C) An R-expression i s f r e e . (where X i s bound i f c-commanded by an antecedent and f r e e r e f e r e n c e PRO i s a r b i t r a r y 10 o t h e r w i s e ) . I t f o l l o w s from these c o n d i t i o n s that l e x i c a l pronominals are free i n p o s i t i o n s where anaphors are bound, as i n : (7) John saw him. i j / * i (8) John saw h i m s e l f . i i / * j In (7), s i n c e pronominals must be f r e e w i t h i n the governing category S ( c o n d i t i o n B) 'him' can not be coindexed with the s u b j e c t . In (8), on the other hand, s i n c e an anaphor must be bound w i t h i n i t s governing category S ( c o n d i t i o n A), and 'John' c-commands the anaphor, i t may be coindexed with the antecedent. Since PRO i s a pronominal anaphor, i t f o l l o w s from the bi n d i n g c o n d i t i o n s that i t must be f r e e . At the same time, however, i t must be bound l i k e anaphors s i n c e i t has no i n t r i n s i c r e f e r e n t i a l c o n t e n t . Because of t h i s c o n t r a d i c t i o n , PRO must appear only i n ungoverned p o s i t i o n s where i t has no governing category and t h e r e f o r e does not f a l l under the Binding c o n d i t i o n s . For example, (9)a. i s ungrammatical because PRO i s governed by ' f o r ' . (9)b. on the other hand, i s grammatical s i n c e PRO i s not governed and i s f r e e w i t h i n i t s governing category S. The d e f i n i t i o n of Government w i l l be given at a l a t e r p o i n t i n t h i s s e c t i o n and w i l l be e l a b o r a t e d upon i n S e c t i o n V. (9) a. * John wanted f o r [ s PRO to win ] ]. (9)b. John wanted [-[s PRO to win ] ] . GB further distinguishes between the empty pronominal PRO [+anaphor, +pronominal], and the empty pronominal pro [-anaphor, +pronominal]. Chomsky (1982) introduced the l a t t e r category to account for the pro-drop phenomena. This w i l l be discussed at length in Section IV and w i l l be used in the analysis here to represent missing subjects in SA. Case theory. This theory i s concerned with the assignment of Case, and requires that every NP with phonological content receives Case. Case i s assigned to NP's when they are in Case marking positions, for example subjects of tensed clauses, objects of t r a n s i t i v e verbs and objects of prepositions. Assignment of Case i s subject to condition of governments (see below). Case i s presumably assigned or checked at D-structure or S-structure. In PF, l e x i c a l items that are not assigned case are f i l t e r e d out by the Case f i l t e r . According to Chomsky (1981:49), the Case f i l t e r i s stated as follows: * NP i f NP has phonetic content and has no case. This theory i s of d i r e c t relevance to our proposed analysis and w i l l be referred to in the discussion of word order in Section I I I . Government Theory The notion government i s central to and pervasive throughout the GB theory. It is relevant to 1 2 subcategorization and the 0 binding theory, and t i e s ©-marking and Case theory. Government i s formally defined as follows Chomsky, 1981:250): [ / 3 . . . 7 . . . C 1 . . . 7 ] , where (a) a = X* or i s coindexed with 7 (b) where <p i s a maximal projection, i f 4> dominates 7 then 0 dominates a. (c) a c-commands 7. In th i s case a governs 7. The d e f i n i t i o n of c-command that we w i l l use here is the following: a c-commands 0 i f f (i) a does not contain 0 , and ( i i ) there i s no maximal projection dominating a that does not dominate 0 . This d e f i n i t i o n expresses that the domain of government i s the maximal projection of the governing head, and that the head must c-command the governee within the maximal projection as i n : (I0)a. X X where X i s the maximal projection of X, X i s the head of X, therefore X governs Y. 1 3 (10)b. X X z X Z V where Z i s a maximal projection, X does not govern Y. Within the theory of government is the p r i n c i p l e the Empty Category P r i n c i p l e (ECP). It requires that traces of moved elements must be properly governed. Proper government is a narrower d e f i n i t i o n of government. This p r i n c i p l e plays a c r u c i a l role in our proposed analysis. We w i l l delay discussion of i t u n t i l Section V. The Bounding theory The Bounding theory s p e c i f i e s l o c a l i t y conditions, in p a r t i c u l a r the Subjacency condition, c f . Chomsky (1977). Subjacency i s a condition on the kind of relations that may hold between antecedents and traces and as such, i t r e s t r i c t s the movements of NP. According to t h i s condition a moved element can not cross more than one bounding node. To i l l u s t r a t e : (11) Who did John say [ s f [ s t" l e f t ] ] (12) What did you say [ t' John l i k e s t" ] (13) a. I wonder [-what[Mary claimed [-(that) [John had seen]]]] 1 4 (13)b. *I wonder[what[Mary heardfthe c l a i m [ t h a t [John had s e e n ] ] ] ] ] l S MP S 5 Sentences (11), (12) and (I3) a . are perm i t t e d s i n c e the movement a p p l i e s c y c l i c a l l y without v i o l a t i o n of the Subjacency C o n d i t i o n . However ,A 1 3)b.- i s ungrammatical s i n c e the moved element crosses-two bounding nodes (an S and an NP) v i o l a t i n g the Subjacency C o n d i t i o n . Bounding nodes are subject to parametric v a r i a t i o n and they vary between languages. For example the bounding nodes i n E n g l i s h are S and NP (Chomsky, 1981), and as i l l u s t r a t e d above. The bounding nodes i n French and I t a l i a n are S and NP ( S p o r t i c h e , 1981, and R i z z i , 1978). For s i m p l i c i t y , we w i l l assume that S and NP are the bounding nodes i n SA. We w i l l not attempt to make a d i s t i n c t i o n between S and S. By using the bounding theory, we w i l l be a b l e to d i s t i n g u i s h c o n s t r u c t i o n s with moved c a t e g o r i e s from those that are base generated. In p a r t i c u l a r , we w i l l show that r e l a t i v e c l a u s e c o n s t r u c t i o n s are base generated whereas WH-formations r e s u l t from movements. Both c o n s t r u c t i o n s i n SA appear to use a resumptive pronoun s t r a t e g y . To i l l u s t r a t e : (14) r a ' a - y - t u [^ssayarat - a [ ^ l a t i [ shatama - ha a r r a j u l - u ] ] ] saw- I the car - acc that broke(3sg.m)-it the man-nom 'I saw the car that the man broke' (15) 15 man [^ra'a-y-ta sayarat-u - hu ] who saw-you(sg.m) car-nom- his 'Whose car did you see' Thus, from (14) and (15), i t is legitimate to assume that both constructions may be derived by movement as there i s no v i o l a t i o n of Subjacency. However, structures such as the following, taken from Aoun (1979), are grammatical. (16) ra'a-y-tu [ad-dubat -a. [alla5i-na [qala l i UP i s S saw-I theofficers-acc who-(pl.m) said(3sg.m)to me 1-hakim -u [.anna-hu [sajana [almutamarredin-a s s U P thegovernor-nomthat-he emprisoned(3sg.m)themutineers-acc Lalla3i-na [ satamu-hum,] ] ] ] ] ]] j s s l who-pl.m. insulted(3pl)them 'I saw the o f f i c e r s that the governor told me that he emprisoned the mutineers that insulted them.' If (16) i s derived by movement, i t would be in v i o l a t i o n of Subjacency as the moved element would have crossed more that two bounding nodes (an NP and S(orS)). But given that SA obeys the Complex NP Constraint (CNPC) (see below), subsumed under Subjacency, cf.Ross (1977), Aoun (1979), and given that examples such as (16) exist in SA, we are lead to assume that r e l a t i v e clause formation i s base generated and not generated by movement. 16 Now l e t us compare the f o l l o w i n g examples: ( I 7 ) a . wajat-tu [ l - k i t a b - a [ a l l a 3 i [ 3 a n a n - t u [ a n n a [ a l w a l a d - a N P t 3 6 a s found-I the book-acc that thought-I that the boy-acc rama-hu,] ] ] ] ] threw ( i t ) 'I found the book that the boy threw ( i t ) . ' (17)b. *man. wa j a t - t u [ a^lkitab-a [_ala3i [ s5anan-tu[ anna- (hu.) [rama-hu]]]]] s By the same l i n e of argumentation, i f we assume that WH-c o n s t r u c t i o n s are base generated, we would expect (17)b. to be grammatical and the resumptive pronoun to r e g u l a r i z e the d e r i v a t i o n . But, the d e s t r i b u t i o n of resumptive pronoun i n q u e s t i o n i s f a r more r e s t r i c t e d than i n r e l a t i v e c l a u s e s . ( T h i s i s s u e i s r a i s e d i n S e c t i o n V ) . Since t h i s resumptive pronoun behaves d i f f e r e n t l y with respect to Subjacency i e . i t obeys the CNPC and the d e r i v a t i o n i s ungrammatical, we are l e a d to assume that WH-construction i s d e r i v e d by movement and i s not base generated. In summary, we have observed that the GB framework p r e d i c t s i n a p r i n c i p l e d manner the d i s t r i b u t i o n of NP's and the domain i n which they may appear. The adoption of t h i s framework w i l l a i d us i n p r e d i c t i n g the d i s t r i b u t i o n s of nominal elements i n SA, i n p a r t i c u l a r the nominal elements 1 7 that appear in subject position. X-bar theory and ©-theory as well as the Projection P r i n c i p l e , ©-criterion and Case theory predict the d i s t r i b u t i o n of NP's in D- and and S-structures. The type of NP however i s determined by the lexicon, Case theory and Binding theory. Furthermore, s p e c i f i c theories such as the Case theory, Binding theory , and Government w i l l be of dir e c t relevance as they w i l l enter into the subsequent analysis of some aspects of the syntax of Standard Arabic. This in turn w i l l aid in accounting for what seems to be a d i s t i n c t behavior in the extraction of subjects from embedded clauses. I l l . WORD-ORDER AND CASE ASSIGNMENT In th i s section, I w i l l argue against the analysis that SA i s a VSO language and w i l l propose an alternative analysis of SA as an SVO language. A. VSO ANALYSIS The " t r a d i t i o n a l " assumption within the generative framework has been that the underlying word-order of SA is VSO, c f . Aoun (1979), Emonds (1980). This is based on the fact that when a sentence contains a verbal element, the verb precedes the subject when i t i s l e x i c a l as in ( 1 8 ) unless emphasis is on the 'agent' subject, in which case i t precedes the verb as in (19): (18) 3ahaba alwalad - u wentOsg.m) the boy-nom 'The boy went' (19) alwalad - u Sahaba ' The boy went' It has been assumed that the underlying structure i s the following according to Aoun (1979): (20) S 18 19 Within t h i s s t r u c t u r e , grammatical r e l a t i o n s are expressed by c o - s u p e r s c r i p t i n g . T h i s i s a type of indexing that i s d i f f e r e n t from the c o - s u b s c r i p t i n g , the indexing r e q u i r e d f o r b i n d i n g . T h i s indexing i n turn forms a d i s c o n t i n u o u s VP of the f o l l o w i n g form: (21 ) k p k p AGR V S O Here the s u b j e c t i s d e f i n e d as the NP coindexed with AGR and the o b j e c t i s the NP coindexed with the verb. T h i s c o i n d e x i n g expresses proper government f o r Case assignment. Thus, the NP r e c e i v e s i t s Case from i t s (proper) governor, c f . Aoun (1979). T h i s n o t i o n w i l l be e l a b o r a t e d upon in S e c t i o n V. Within t h i s a n a l y s i s , SVO word order i s d e r i v e d by T o p i c a l i z a t i o n . Hence, the i n i t i a l p o s i t i o n in the sentence i s not ©-marked to allow movement without v i o l a t i o n of the ©-criterion. B. AN ALTERNATIVE ANALYSIS I would l i k e to propose that SA i s an SVO language. F i r s t , SVO word ord e r s , i n f a c t , occur along with VSO ones i n matrix sentences as i l l u s t r a t e d i n (18) and (19), as w e l l as i n embedded c l a u s e s . The l a t t e r may be introduced by e i t h e r an 'an' or an 'anna' complementizer. The word order i n each c l a u s e , however, depends on the c h o i c e of the complementizer that i n t r o d u c e s i t . 'an' r e q u i r e s a VSO 20 word-order and assigns the Subjunctive mood (mudari{ mansub). 'anna' assigns accusative Case and requires an SVO word order. To i l l u s t r a t e : (22) arada almudarres -u [. an [ yaktub-a 1-walad-u s su want(3sg.m)theteacher-nom that write(3sg.m)-subj.theboy-nom d-darrs-a]] the lesson-acc 'The teacher wanted the boy to write the lesson' (23) arada almudarres-u [-anna [1-walad - a yaktub-u d-dars-a 3 ] s s the boy-acc write - i n d i e . 'The teacher wanted the boy to write the lesson.' These facts therefore support an SVO order as well as an VSO one. Second, the behavior of reflexive anaphors supports the claim that SA is an SVO language. Recall from Section 2 that anaphors must be bound within their governing category. For example, in (24), the anaphor "nafsa-hu" (himself) needs to be bound by "alwaladu" (the boy), i t s antecedent: (24) kalama al-walad -u nafsa - hu spoke(3sg.m) the boy-nom se l f - him 'The boy spoke to himself.' If the VSO analysis i s adopted, the anaphor i s bound in i t s governing category S according to condition A of the binding theory. The l e x i c a l NP "alwaladu", however, i s also bound 21 within the governing category S as i l l u s t r a t e d in structure (21). This i s in d i r e c t v i o l a t i o n of condition C of the binding theory, which stipulates that an R-expression must be free. Thus the subject must be higher in the structure, i e . outside the projection of the verb. Therefore, i f we assume an SVO analysis with the structure of (25), then t h i s v i o l a t i o n i s avoid. (25) S NP INFL VP V NP Further support for the SVO proposal concerns Case assignment to the subject of an 'anna' clause. In the VSO analysis, where SVO i s derived by T o p i c a l i z a t i o n , i t i s d i f f i c u l t to explain how Case i s assigned to the subject of 'anna' and why i t gets accusative Case. There w i l l be a c o n f l i c t in the Case assignment of that position since 'anna' i s a Case assigner and must assign i t s Case. At the same time the coindexing between AGR and the subject position for Nominative Case assignment w i l l also hold. Thus, within the VSO analysis there i s no obvious way to answer these questions and block Nominative Case assignment. However, in the analysis proposed here there i s a simple 22 solution as i s shown below. If we assume that SVO word order is basic in SA, we then need to show how VSO is derived. The two p o s s i b i l i t i e s are either the verb is preposed or the subject moves into VP. These two options are shown in (26) and (27) respectively. (26) Verb-preposing 23 to the verb, the verb would govern the subject but not d i r e c t l y 0-mark i t . Since t h i s movement would change the argument s t r u c t u r e of the verb, t h i s w i l l be i n v i o l a t i o n of the P r o j e c t i o n P r i n c i p l e . Furthermore, the t r a c e of the moved NP w i l l bind a l e x i c a l NP, i n v i o l a t i o n of c o n d i t i o n C of the b i n d i n g theory. Hence, the P r o j e c t i o n P r i n c i p l e as w e l l as c o n d i t i o n C of the b i n d i n g theory l e a d us to r e j e c t s u b j e c t movement. We are l e d to adopt the other a l t e r n a t i v e , verb movement - where the verb a d j o i n s to S. If we adopt t h i s a n a l y s i s , there are three f u r t h e r q u e s t i o n s we need to answer. These a r e : 1) How does the s u b j e c t get i t s Case? 2) Why i s there no verb movement in 'anna' c l a u s e and the word order i s SVO? 3) How does the s u b j e c t get case i n Matrix SVO order? To answer these q u e s t i o n s we w i l l r e l y c r u c i a l l y on Case theory. We w i l l assume that Case assignment i n SA only a p p l i e s to the r i g h t and r e q u i r e s adjacency. T h i s i s a r e s u l t of parametric v a r i a t i o n (these s p e c i f i c parameters have been d i s c u s s e d i n recent work i n Government by Sproat (1983) and Koopman (1985)). We w i l l f u r t h e r assume that the " i n f l e c t i o n a l " and b i n d i n g element (INFL) i s d i r e c t l y dominated by S and i s the head of S (Chomsky, 1981). I t c o n t a i n s the f e a t u r e s [•tense], and an agreement element (AGR). AGR i s nominal in c h a r a c t e r and has the f e a t u r e s person, number and gender. I t appears when the sentence i s f i n i t e i e . [+tense]. I t i s assumed that AGR i s the element that governs and a s s i g n s Case to the subject NP. 24 With these assumptions and Case theory, we w i l l show how the previously raised questions are answered. Since AGR is the element that assigns Case to the subject, and Case i s only assigned to the right, there must exist a rule in the grammar that moves AGR to the l e f t and adjoins i t to S. AGR is then in a position to assign Case to the subject NP to i t s right. 1 Since INFL has no f u l l l e x i c a l status, i t appears phonetically as part of a verbal a f f i x system in surface structure. Thus, verb preposing i s obligatory in order to provide a locus for AGR to c l i t i c i z e to. We w i l l assume that t h i s verb movement adjoining the verb to AGR applies prior to S-structure. This results in the word order we find in matrix sentences with verbal elements where the verb precedes the subject as in (18). The following structure i l l u s t r a t e s t h i s : (28) S 1 Another p o s s i b i l i t y is to assume that AGR originates in an i n i t i a l position in the base i . e . , the underlying word order is INFL NP VP. This p o s s i b i l i t y was rejected since there w i l l be no adjacency between the complementier 'anna' and the embedded subject NP to which i t assigns accusative case. 25 For matrix sentences with no overt verbal element such as: (29) alwalad - u f i - 1-manzel - i the boy- nom in -the-house - gen 'The boy i s in the house.' we propose that there i s an empty copula, following a suggestion of M. Rochemont. This copula functions in the same manner as a l e x i c a l verb. It must be present in order for Case to be assigned to the subject NP, otherwise the sentence would be ruled out by the Case f i l t e r . We can now apply t h i s analysis to embedded clauses. The difference between 'an' and 'anna' clause can now be explained i f we further specify that 'anna' i s a Case assigner, but 'an' i s not. As in matrix sentences, verb preposing must occur in 'an' clauses in order to assign Case to the subject of the embedded clause. Since 'anna' i s a Case assigner, however, i t assigns Case to the subject so that no verb preposing i s necessary. The result of t h i s i s the required SVO word order in 'anna' clauses. A remaining problem i s SVO word order in matrix sentences. We propose that this results from To p i c a l i z a t i o n , a movement rule. The Topic i s moved to a higher A position peripheral to S. Hence the SVO order has the following structure: 26 (30) S NP INFL VP 1 1. V NP . t. As far as the ECP i s concerned, the trace of the subject w i l l be properly governed by the verb. In summary, we have argued in this section that i f the underlying structure of SA i s SVO, not only grammatical relations are expressed in terms of their constituent structures but Case assignment can also be accounted for in a u n i f i e d way in matrix and embedded clauses. Furthermore, i t follows automatically from the proposed analysis that the subject of 'anna' is assigned accusative case. Thus, th i s analysis resolves what seemed to be a c o n f l i c t in the case assignment of t h i s position under the VSO analysis. IV. PRO-DROP PHENOMENA AND CLITICS. Up u n t i l t h i s s e c t i o n , the proposed a n a l y s i s has d e a l t with s u b j e c t p o s i t i o n f i l l e d with f u l l l e x i c a l nominal elements. To complete our a n a l y s i s , we w i l l c o n s i d e r i n s t a n c e s where argument p o s i t i o n s are f i l l e d w i t h empty nominal elements . Hence, we t a l k about missing s u b j e c t s and c l i t i c s . SA i s a n u l l argument language, with m i s s i n g s u b j e c t s or o b j e c t s appearing as in (31) and (32) r e s p e c t i v e l y . (31 ) dahaba went(3sg.m.) 'He went.' (32) kataba - hu wrote(3sg.m.) - it(m.sg.)-acc 'He wrote i t . ' In languages that allow m i s s i n g o b j e c t s such as I t a l i a n , Spanish, French and SA, a c l i t i c must n e c e s s a r i l y be p r e s e n t . In t h i s S e c t i o n , f i r s t we w i l l present the a n a l y s i s of the m i s s i n g subject property adopted here. Then we w i l l attempt t o answer the f o l l o w i n g q u e s t i o n s : 1) How are c l i t i c s analyzed? 2) What i s the r e l a t i o n s h i p between c l i t i c s and missing o b j e c t s ? 3) Why are there no s u b j e c t c l i t i c s ? 4) Why may a c l i t i c appear i n the s u b j e c t p o s i t i o n in 'anna' c l a u s e s ? 27 28 A. PRO-DROP ANALYSIS Thus far we have assumed that the subject position in a sentence i s obligatory Chomsky, 1981,1982). In Section Three, we argued that the .subject in SA i s generated outside VP. Now l e t us turn to examples where we have n u l l subject positions as in the following examples: (33) Sahaba went(3sg.m.) 'He went.' (34) 3ahab - u went(3pl.m) 'They went.' From the surface form of these sentences, we cannot t e l l whether the subject position is obligatory or not since i t is empty. However, to account for the morphological differences in the agreement marker that i s r e f l e c t e d on the verb and to s a t i s f y the 0- c r i t e r i o n and the Projection P r i n c i p l e , we have to assume that there must be an empty pronominal with which the verb agrees. I w i l l assume that th i s empty pronominal i s pro [ -anaphor, +pronominal ] following Chomsky (1982). In the discussion of the typology of empty pronominals, Chomsky (1982), argues that pro i s best suited for representing the missing subjects in pro-drop languages for 29 various reasons. F i r s t , missing subjects in pro-drop languages can never be arbitrary in reference. This i s not accounted for i f PRO i s used in this position. Secondly, in Spanish interrogatives a rule of verb fronting applies, placing the subject in a governed position, c f . Torrego ( 1 9 8 4 ) . This makes i t impossible to use PRO to represent the missing subjects, at least in Spanish. Thirdly, the use of pro makes i t unnecessary to invoke a parameter involving movement of INFL in the syntax in order to allow PRO to appear in the ungoverned position ( rule R of Chomsky, 1981). The only condition on the appearance of pro to represent the missing subject i s that i t must be ' l o c a l l y determined' by AGR. This i s only permissible i f AGR i s r i c h enough to i d e n t i f y the features of pro. For example, r i c h agreement i s re f l e c t e d on the verb as i l l u s t r a t e d in (33) and ( 3 4 ) . This contrasts with English where there is no r i c h agreement and therefore, a f u l l pronoun must appear as in He went. / * Went. B. CLITICS The analysis which we w i l l adopt of c l i t i c s i s that they are base generated as a feature on the head that they c l i t i c i z e to. It follows from the Projection P r i n c i p l e that a ©-marked position must be s t r u c t u r a l l y represented even i f i t i s empty. Therefore, the c l i t i c must be associated with an empty argument position. We further assume that since the c l i t i c appears as a feature on the head, i t c-commands 30 and governs that empty p o s i t i o n (Borer, 1984). The q u e s t i o n a r i s e s as to the type of t h i s empty category. I w i l l adopt Chomsky's (1982) a n a l y s i s and assume that i t i s pro. Thus, the s t r u c t u r e that w i l l be adopted for c l i t i c s i s the f o l l o w i n g : (35) X X - c l pro Since the f e a t u r e s of pro must be determined and s i n c e the c l i t i c i s nominal i n nature, i e . i t has the f e a t u r e s person, number and gender, we c o u l d assume that the c l i t i c i s o b l i g a t o r y present i n order to determine the f e a t u r e s of pro ( c f . Hurtado, 1985, Roberge,1985). We w i l l f u r t h e r assume that the c l i t i c i s a ' s p e l l out' of the Case f e a t u r e assigned by the head (Aoun,1979), (Borer,1984). To p r o v i d e evidence to support t h i s assumption, we need to d i s c u s s c o n s t r u c t s t a t e s t r u c t u r e s . Construct s t a t e s t r u c t u r e s are a form of complex noun phrases formed by a s u c c e s s i o n of bare nouns. They are r i g h t branching s i m i l a r in t h e i r form to the Hebrew ones (Borer (1984)). T h i s s t r u c t u r e i s r e s t r i c t e d i n i t s formation i n that the determiner can only appear a t t a c h e d to the l a s t c o n s t i t u e n t . T h i s i s i l l u s t r a t e d i n (36) - (38): (36) 31 (37) (38) bab-u 1- manzil - i door-nom the-house - gen 'the door of the house' bab- u manzil - i 1- mudarres - i door-nom house -gen the-teacher(m)-gen 'the door of the house of the teacher' *bab-u 1-manzil-i 1-mudarres door-nom the-house-gen the-teacher The construct state structure i s thus represented as: (39) i We w i l l assume, following Borer's analysis, that Genitive Case i s assigned by the head noun only i f the f i r s t node (N), which dominates that head (N) immediately 32 dominates the complement,cf. Borer (1984:48). Any expansion which r e s u l t s i n r i g h t branching of the head such as the use of a d j e c t i v e s , leads to the ungrammaticality as in (40) represented as (41). (40) * al-bab-u 1-kabir-u 1-manzil the-door-nom the-big-nom the-house (41 ) \ DET N DET N the door the b i g the house The ungrammaticality of (40) can be e x p l a i n e d as f o l l o w s . Since the f i r s t node (N) which dominates the head (N) does not dominate the complement (N), (N) can not be assi g n e d Case and the s t r u c t u r e i s r u l e d out by the Case f i l t e r . For the d e r i v a t i o n to be grammatical, a Case a s s i g n e r must be i n s e r t e d to a s s i g n Case to (N) as i n (42) r e p r e s e n t e d as (43). (42) 33 It should also be mentioned that, in construct state structures, the d i s t i n c t i o n between s p e c i f i e r s and complements i s not re f l e c t e d s t r u c t u r a l l y . Thus, (44) and (45) have i d e n t i c a l structures. (44) kitabat - u 1- bint - i writing -nom the-girl-gen 'the g i r l ' s writing' (45) kitabat - u 1- qasTdat - i writing -nom the- poem - gen 'the writing of the poem' 34 In sum, we have i l l u s t r a t e d that the head in a construct state structure assigns genitive Case to i t s complement. This follows from the Case theory. Since Case i s always re a l i z e d morphologically in SA, i t must be stipulated in the grammar that case must be realized, c f . Stowell (1981). This s t i p u l a t i o n i s essential to our proposed analysis for c l i t i c s . Now with this dicussion of construct state structures and the assumption that Case must be realized, we can proceed to show how c l i t i c s absorb case. Since c l i t i c s may appear as a feature on the head of a phrase, we expect them to appear c l i t i c i z e d to heads of construct state constructions as in (46) [ kitabat - u - ha pro ] writing -nom - h e r / i t (3sg.f) 'her writing / the writing of i t ' (47) * [ kitabat - 0 - pro ] The c l i t i c "ha" in (46) can refer to either a s p e c i f i e r as in (44), or a complement as in (45). This c l i t i c appears in order to l o c a l l y determine the features of the empty pronominal pro. A bare noun can not do so since there i s no head-complement agreement as in (47). Since the head of construct state structures assigns Case and that Case i s morphologically realized, i t i s correct to assume that the 35 c l i t i c must be present to identif y the features of pro as well as absorb the Case assigned by the head. Based upon th i s (47) i s therefore, ungrammatical since the features of pro are not i d e n t i f i e d and the Case assigned by the head noun 'kitabat' is not realized. Consider now the cases when both a s p e c i f i e r and a complement appear in a sentence as in : (48)a. kitabat -u 1-bint-i 1- qasidat- a writing-nom the-girl-gen the-poem-acc 'The g i r l ' s writing of the poem' (48) b. kitabat -u 1-bint - i l i - l - q a s i d a t - i writing-nom the-girl-gen of -the poem-gen 'The g i r l ' s writing of the poem' (49) a. kitabat-u - ha 1-qasTdat - a (49) b. kitabat-u - ha l i - 1 - qasldat-i (49) c. * kitabat-u -ha 1-qasidat In (49), the c l i t i c appears to ide n t i f y pro as well as absorb the Case assigned by the head noun. Since the verbal noun 'kitabat' may take multiple arguments and assigns two 0-roles, the complement 'qasidat' may be assigned accusative Case by the head as in (49)a., or a Case marker may be inserted and assigns genitive Case to that comlpement NP as in (49)b. The complement however does not receive the genitive Case assigned by the head of the construction since 36 the c l i t i c absorbs that case. Consequently i f the complement does not receive Case, the derivation i s ungrammatical and i t is ruled out by the Case f i l t e r as in (49)c. Further evidence to support the fact that c l i t i c s absorb Case i s provided from cases where c l i t i c s appear attached to verbs or prepositions. According to subcategorization frames and X-theory, heads of constructions require the presence of their complements. If the comlement i s empty, i e . pro, i t s contents must be i d e n t i f i e d . To i l l u s t r a t e : (50)a. (50)b. (51)a. (51)b. daraba 1- walad - a h i t (3sg.m) the- boy - acc - hu him(3.sg,m) 'He h i t the boy./ He h i t him.' * daraba pro l a ( i b a f i - 1 - hadiqat - i played(3.sg.m)in - the garden - gen -ha it(3sg.f) 'He played in the garden./He played in i t ' * l a ( i b a f i - pro 37 Verbs and p r e p o s i t i o n s a s s i g n a c c u s a t i v e case and g e n i t i v e case r e s p e c t i v e l y under the assumption that whenever Case i s a s s i g n e d , i t must be r e a l i z e d as i n (50)a. and (51)a. I t should a l s o be noted that the c l i t i c may appear a t t a c h e d to these heads. We can conclude that the c l i t i c i d e n t i f i e s the f e a t u r e s of pro as w e l l as absorbs the Case assig n e d by the head of the phrase. With the adoption of t h i s assumption, we can proceed to show how t h i s a n a l y s i s can account f o r m i s s i n g s u b j e c t s and o b j e c t s i n matrix sentences as w e l l as embedded c l a u s e s . The missing subject property can be e x p l a i n e d i n the f o l l o w i n g manner. Since we are assuming that AGR i s generated under INFL, as mentioned e a r l i e r i n S e c t i o n Three, t h i s blocks PRO from appearing i n s u b j e c t p o s i t i o n at D-s t r u c t u r e s i n c e i t i s a governed p o s i t i o n . Pro, on the other hand can appear in t h i s governed p o s i t i o n . R e c a l l t h a t the f e a t u r e s of pro must be determined by the r i c h agreement on the verb and that AGR governs and a s s i g n s Case to the subject NP to the r i g h t . We w i l l f u r t h e r assume that the i d e n t i f i c a t i o n of pro i s rightward as w e l l . Thus AGR moves to the l e f t of pro i n order to i d e n t i f y i t s f e a t u r e s . Since AGR has to a t t a c h to the verb, the verb moves a d j o i n i n g to AGR. That r i c h agreement ( i e . i d e n t i f i c a t i o n of pro by AGR) i s a necessary c o n d i t i o n f o r i d e n t i f i c a t i o n of m i s s i n g s u b j e c t s i s i l l u s t r a t e d by the p a r t i c i p l e form of the verb. T h i s p a r t i c i p l e form agrees i n number and gender but not i n 38 person. Therefore, AGR i s not r i c h enough to identify a missing subject and the subject must be l e x i c a l as shown in (52)b. (52)a. * pro Sahebuna going (part, pl.m) (52) b. al-alwadu Sahebuna2 the boys going (patr.pl.m) 'The boys are going' That INFL i s adjacent to the verb in surface form supports our assumption that the verb in fact moves to the AGR element. If t h i s i s true, then AGR never moves down adjoining the verb. Therefore, we can assume that in matrix sentences with missing subjects, AGR as well as verb movement apply in order to i d e n t i f y the features of pro. This i s also the case in 'an' clauses, as i l l u s t r a t e d in (53) : (53) arada [_ an [ pro I NFL ya3haba ]] s s want(3sg.m) that go(3sg.m.subj.) 'He wanted to go' AGR i s generated in a position that governs the subject pro. 2 Although there i s no verbal element in t h i s sentence, we are assuming that the base order i s SVO and involves an empty copula. This empty copula functions as i t s overt counter part where AGR and verb preposing must apply to assign case to the subject. The surface word order i s VSO although i t does not show i t . 39 But since we are assuming that the i d e n t i f i c a t i o n of pro is to the right, AGR must move to the l e f t of the subject position in order to identif y i t s features. The verb in turn moves and adjoins AGR. This in fact i s the required word order in an 'an' clause since 'an' i s a mood assigner and i t assigns the Subjunctive mood to the verb adjacent to i t . In 'anna' clause on the other hand, there i s obligatory presence of a c l i t i c when there is a missing subject as in (54) : (54)a. amara [-anna- hu[ spro I NFL yaktubu ] 3 ordered(3sg.m) that-he write(3sg.m. i n d i e ) . 'He ordered him to write.' (54)b. * amara [.anna- Q[^pro I NFL yaktubu 3 3 Since we are assuming that the i d e n t i f i c a t i o n of pro i s right ward, AGR in an 'anna' clause can not ide n t i f y the features of pro. Furthermore since 'anna' is a Case assigner and i t assigns accusative Case to the subject position, according to the Case Realization P r i n c i p l e AGR and verb preposing do not apply since we can not have case c o n f l i c t . The c l i t i c i s then obligatory in order to ide n t i f y the features of pro as well as absorb the Case assigned by 'anna' (54)a. Without the c l i t i c , the derivation i s ungrammatical since the features of pro are not i d e n t i f i e d and the Case that 'anna' assigns is not 40 re a l i z e d (54)b. Further evidence to support the claim that the c l i t i c in fact absorbs Case i s provided from examples where the c l i t i c appears in 'anna' clauses together with verb movement as i n : (55)a. a'refu [ anna- hu[ akala 1-walad -u t-tufahat-a] ] know- I t h a t - i t ate(3sg.m)the boy-nom the apple-acc 'I know that the boy ate the apple' (55)b. * a'refu [ anna-hu [ 1-walad- akala t-tufahat-a ] ] If the c l i t i c appears as i s (55)a., i t absorbs the Case assigned by 'anna'. If verb movement does not apply to assign Case to the subject of the embedded clause, the derivation is ungrammatical as i t i s in v i o l a t i o n of the Case f i l t e r as in (55)b. Therefore verb movement i s obligatory in the embedded clause in order to assign Case to i t s subject as in (55)a. In summary, the pro-drop phenomena and c l i t i c i z a t i o n are related. In fact, both deal with empty pronominal positions f i l l e d with pro. The features of pro must be i d e n t i f i e d under government. However, the i d e n t i f i c a t i o n of pro in subject position d i f f e r s from when pro i s in a complement position. Pro i s permitted in subject position only i f the AGR element i s r i c h enough to ident i f y i t s features. In complement position however, a c l i t i c must appear to ident i f y i t s features. Thus cross l i n g u i s t i c a l l y , 41 there are two options to identi f y pro, either by r i c h agreement or by a c l i t i c . Furthermore, the c l i t i c has an additional function, i t absorbs the Case assigned by the head. This i s well supported by instances when the c l i t i c appears as a ' s p e l l out' of the Case of the 'anna' complementizer. In such a case INFL preposing as well as verb movement must apply to ensure that Case is assigned to the l e x i c a l subject of the embedded clause. V. WH-EXTRACTION. In t h i s f i n a l S e c t i o n , we w i l l d i s c u s s the e x t r a c t i o n of NP's from both matrix and embedded c l a u s e s . A f t e r i l l u s t r a t i n g the problem i n SA, we w i l l show how the d i s t r i b u t i o n of c l i t i c s and empty c a t e g o r i e s f a l l s out from the proposed a n a l y s i s , given the ECP and s e v e r a l supported assumptions. A. THE PROBLEM E x t r a c t i o n of noun phrases from matrix sentences i n SA a p p l i e s f r e e l y , obeying the Subjacency c o n d i t i o n d i s c u s s e d in S e c t i o n I I . To i l l u s t r a t e : (56) man. kataba t. d-dars- a who wrote(3sg.m) the lesson-acc 'Who wrote the lesson?' (57) ma5a kataba 1-walad- u t i i what wroteOsg.m) the boy -nom 'What d i d the boy w r i t e ? ' E x t r a c t i o n of noun phrases from o b j e c t p o s i t i o n , i e . from post v e r b a l p o s i t i o n i n embedded c l a u s e s , a l s o a p p l i e s f r e e l y : (58) 42 43 ma3a. arada 1-mudarres-u [.an [ yaktub-a t-elmi5 - u t.] ] i i s 1 what wanted the teacher-nom that w r i t e - s u b j the student-nom 'What d i d the teacher want the student to w r i t e ? ' (59) mI3a. arada 1-mudarres -u [ anna [ t-elmi3 -a whatiwanted theteacher-nom that --thestudent-acc yaktubu- t j ]] w r i t e - i n d i c . 'What d i d the teacher want the student to w r i t e ? ' E x t r a c t i o n of the subject from embedded c l a u s e s d i f f e r s depending on the complementizer t h a t i n t r o d u c e s the embedded c l a u s e . E x t r a c t i o n of the subject of an 'an' c l a u s e a p p l i e s f r e e l y : (60) man. arada 1-lmudarrese-u [.an [ yaktub-a t. d-dars -a ] ] i s s i who wanted theteacher-nom that w r i t e - s u b j thelesson-acc 'Who d i d the teacher want to w r i t e the l e s s o n ? ' E x t r a c t i o n of the s u b j e c t from an 'anna' c l a u s e , on the other hand, r e q u i r e s an o b l i g a t o r y c l i t i c to appear i n p l a c e of the moved subject NP. I f no c l i t i c appears, the d e r i v a t i o n i s ungrammatical. (61) i l l u s t r a t e s t h i s : (61)a. 44 man arada 1-mudarres - u fanna-hu F yaktub-u who wanted theteacher-nomthat-he write-indic. d-dars-a]] the lesson-acc 'Who did the teacher want to write the lesson?' (61)b. * man arada 1-mudarres-uj^anna j^yaktub-u d-dars-a B. THE SOLUTION To explain this d i s t i n c t behavior of the extraction of the subject of an 'anna' clause, we w i l l rely on the proposed analyses of word order and c l i t i c s as well as the Empty Category P r i n c i p l e (ECP). The ECP requires that every trace must be properly governed. A trace i s either a variable or an NP trace. Roughly speaking, an empty category i s properly governed i f 1) i t i s a complement of a head such as V, N, P, or 2) i t i s coindexed with a l o c a l antecedent. Proper government i s formally defined by Chomsky (1981:250) as: (62) a properly governs/3 i f and only i f a governs 0and [a#AGR] ECP : [a e ] must be properly governed. In GB, an empty category i s a variable i f and only i f i t i s 45 Case marked and bound by an operator in COMP po s i t i o n , i . e . i t i s A-bound. An NP-trace, on the other hand, is bound by an argument, i e . A-bound. Since SA does not have r a i s i n g verbs such as verb the 'seem' which allow the trace to be bound by an argument, our discussion of traces w i l l be limited to variables only. To i l l u s t r a t e : (63) man qara'a t i i 'who read?' (64) al-walad-u qara'a t i i 'The boy read.' The traces in (63) and (64) are A bound by the WH element and the Topic respectively. In both derivations, the trace is properly governed by the verb. Thus, the ECP i s s a t i s f i e d . The ECP is formulated to apply to traces but not to PRO as PRO has features and thus i s not an empty category. This allows derivations such as: (65) I dont't know [ what [ PRO to do t ] ] i i 46 In (65).the trace, t, is properly governed, but PRO is not. Thus the ECP as well as the binding conditions are s a t i s f ied. On the other hand,an English sentence such as the following is ungrammatical: (66) * B i l l was preferred [ for [ t to have seen Tom ] ] i i Since 'for' i s not a proper governor, therfore the trace of the movement is not in a properly governed position and the derivation i s ruled out by the ECP. If we assume that the 'anna' complementizer in SA i s l i k e 'for' in English, i e . i s not a proper governor, then the extraction of the subject of 'anna' clauses leaves a trace which i s not properly governed, a v i o l a t i o n of the ECP. Futhermore, i f we assume that the c l i t i c i s a proper governor in SA (Borer 0 984)) and below, i t must appear when the subject of an 'anna' clause i s extracted in order to properly govern the trace, as i l l u s t r a t e d in (61)a. The assumption that c l i t i c s are proper governors is supported by"evidence of extractions from other - structures. (67) and (68) are examples of the extraction of complements from prepositional phrases of construct state structures. (67)a. 47 man. sallama 1-mudarresu- u 'alay-hi. i i who shook hands the teacher-nom with-him 'Whom did the teacher shake hands with?' (67)b. * man sallama 1-mudarres- ( a l a (68)a. man rama 1-walad-u kitaba-hu i i who threw the boy-nom book-acc-his 'Whose book did the boy throw?' (68)b. * man rama 1-walad-u kitaba (67) and (68), show that prepostions as well as nouns can not function as proper governors. Only the a v a i l a b i l i t y of the c l i t i c makes extraction possible (67)a. and (67)b. It must be the case, then, that the c l i t i c functions to properly govern the trace. Since both preposition and noun stranding result in ungrammaticality (67)b. and (68)b., we conclude that prepostions and nouns are not proper governors as i s also the case in Hebrew. Now consider the following examples: (69) man ra'a t i i who saw(3sg.m.) 'Whom did he see?' 48 (70) ma3a kataba t i i what wrote(3sg.m) 'What did he write?' (69) and (70) do not require a c l i t i c to appear in order to properly govern the trace. From th i s we can assume that only verbs, are proper governors in SA. (Since adjectives do not take bare NP complements, we can generalize this to the feature [+V]). Furthermore, since the appearance of the c l i t i c regularizes the derivation as shown in (61)a., (67) and (68), we w i l l assume that the c l i t i c i s coindexed with the empty position in order to properly govern i t . To capture these facts we w i l l adopt the d e f i n i t i o n of proper government proposed by Borer (1984:71): (71 ) a properly governs 0 i f f a governs 0 and (i) a i s [ +V ] or ( i i ) a i s coindexed with 0. Now, with these assumptions we w i l l i l l u s t r a t e how the proposed analysis accounts for the extraction of noun phrases from matrix and embedded clauses. WH-extraction of objects from matrix sentences follows automatically from the previously stated assumptions. Since extraction of the object would leave the trace properly governed by the verb, the derivation w i l l not v i o l a t e the ECP, e.g. (57), (69), and (70). 49 For extraction of the subject, r e c a l l from Section III that AGR as well as verb preposing apply for the purpose of Case assignment of a l e x i c a l subject. Since variables are Case marked, we have to assume that extraction of the subject from matrix sentences applies from a post verbal position. Movement i s permissible since the trace i s properly governed by the verb as in (56) and (63). Extraction of noun phrases from embedded clauses obeys Subjacency as discussed in Section II, and follows this analysis in the same manner as extraction from matrix sentences. Extraction of the object i s permissible since the trace i s properly governed by the verb of the embedded clause as i s (58) and (59). Extraction of the subject from an 'an' clause applies in the same mannner as extraction of the subject of matrix sentences. Since i t applies from a post-verbal position, for reasons discussed above, the trace does not violate the ECP. It i s governed by the verb of the embedded clause as i l l u s t r a t e d by (60). The extraction of the subject of an 'anna' clause, on the other hand, i s treated d i f f e r e n t l y . Recall that 'anna' i s a Case assigner assigning accusative Case to the subject in i t s clause. This does not trigger AGR and verb preposing. Hence, extraction of the subject from an 'anna' clause must apply from a pre-verbal position. This leaves a trace not properly governed in v i o l a t i o n of ECP. For subject extraction from an 'anna' clause to be grammatical, 50 then a c l i t i c must appear to properly govern the trace l e f t by movement as in (61)a. Otherwise the derivation i s ungrammatical in v i o l a t i o n of the ECP as in (61)b. The proposed analysis also predicts that the c l i t i c must be associated with an empty category. To i l l u s t r a t e , consider the following examples: (72) *man arada 1-mudarres- u [.anna-hu [l-walad- yaktub-u who wanted theteacher-nom that-he theboy write-indic. d-darrsa-a] ] the lesson (73) *man arada 1-mudarres-u £_an £ yaktuba-hu d - d a r r s J J who wanted the teacher-nom that w r i t e - i t the lesson (74) *ma3a arada almudarres-u |\an j" yaktuba-hu alwalad-u JJ what wanted the teacher that w r i t e - i t the boy-nom If the c l i t i c appears attached to a head of a phrase, i t indicates that this head i s associated with an empty category. Therefore, the ungrammatically of (72) and (73) could be explained in the following manner. If the c l i t i c appears as well a l e x i c a l category, the Projection P r i n c i p l e and the 0- c r i t e r i o n w i l l be violated since the empty 51 position, the trace and the l e x i c a l element are assigned one ©-role. Furthermore, since the the c l i t i c absorbs the case assigned by the head, the l e x i c a l element w i l l not be assigned Case in v i o l a t i o n of the Case f i l t e r . (74) i s also ungrammatical since the c l i t i c appears in a properly governed po s i t i o n . In summary, the proposed analysis correctly predicts the subject extraction p o s s i b i l i t i e s from embedded clauses. Extraction of the subject from an 'an' clause does not require a c l i t i c to appear. In fact i f a c l i t i c appears, i t renders the sentence ungrammatical. On the other hand, extraction of the subject from an 'anna' clause requires a c l i t i c to appear in order to properly govern the trace. Thus, this analysis accounts for the following paradigm: (75) Subject extraction from an 'an' clause. (a) *'an' without verb movement. (b) *'an' without verb movement, with a c l i t i c . (c) *'an' with verb movement, with a c l i t i c . (d) 'an' with verb movement. (76) Subject extraction from an 'anna' clause. (a) *'anna' without verb movement. (b) 'anna' without verb movement, with c l i t i c . (c) *'anna' with verb movement. 52 (d) 'anna' with verb movement, with c l i t i c If subject extraction takes place from a pre-verbal position as in (75)a. and (75)b., the variable w i l l not be Case marked. The trace i s also not properly governed and is in v i o l a t i o n of ECP. Furthermore, the c l i t i c in (75)b. can not appear attached to 'an' since 'an' is a mood assigner and must be followed by a verb. If the c l i t i c however, appears attached to the verb i t w i l l be in no position to properly govern the trace. Therefore the derivation is in v i o l a t i o n of the ECP. In ( 7 5 ) c , extraction from a post-verbal postion leaves the trace properly governed. If the c l i t i c also appears in a configuration of proper government, this results in the ungrammaticality of the derivation. In (75)d. on the other hand, the extraction i s from a properly governed p o s i t i o n . The trace which i s l e f t is properly governed, obeying ECP and the derivation i s grammatical. In (76)a, extraction of the subject of the 'anna' clause leaves a trace that i s not properly governed since 'anna' i s not a proper governor. This violates the ECP. If extraction of the subject of the 'anna' clause takes place and the the c l i t i c appears, i t properly governs the trace, and the derivation i s grammatical as in (76)b. Extraction of the subject from an 'anna' clause, with verb movement, leaves a trace properly governed. However, 53 the Case assigned by 'anna' is not absorbed, hence, the derivation is ungrammatical as in (76)c. In (76)d., extraction of the subject, with verb movement, leaves the trace properly governed by the verb. Furthermore, the Case assigned by 'anna' i s absorbed by the c l i t i c . The derivation i s grammatical. CONCLUSION It has been argued in t h i s paper that Standard Arabic is an SVO language rather than a VSO. With th i s underlying word order grammatical r e l a t i o n s are expressed by their constituent structure and Case assignment of the subjects of matrix sentences and embedded clauses i s u n i f i e d . VSO word order according to this analysis is derived by verb movement since government and Case assignment is assumed to apply only to the right, this forces agreement and verb preposing to apply. VSO surface word order is derived by T o p i c a l i z a t i o n . It has also been argued that c l i t i c s always appear with empty categories. They must appear to i d e n t i f y the features of t h i s empty pronominal pro as well as absorb the Case assigned by the head. Furthermore, i t has been argued that c l i t i c s may appear as a r e s u l t of movement in order to properly govern the l e f t trace v i o l a t i n g ECP. Now, the c l i t i c appearing in the extraction of the subject of an 'anna' clause follow naturally from the proposed analysis. Since 'anna' is a Case assigner and i t must assign i t s Case according to the Case Realization P r i n c i p l e , AGR and verb preposing may not apply so as not to have a c o n f l i c t in the assignment of Case to the subject in the embedded clause. Thus, extraction of the subject leaves a trace not properly governed in v i o l a t i o n of ECP. A c l i t i c must then appear in order to properly govern the trace l e f t by movement. The extraction of the subject from an 'an' 54 55 clause, on the other hand, apply from a properly governed position since AGR and verb preposing must apply in order for the subject to be assigned Case. This leaves a trace properly governed by the verb. The c l i t i c does not appear since i t i s a properly governed configuration. In conclusion, the adoption of the theore t i c a l tools of the GB framework f a c i l i t a t e s the analysis of various relevant features of Arabic syntax. This not only explains the d i s t i n c t behavior in the extraction of the subject from embedded clauses in SA in a uni f i e d manner, but t h i s d i s t i n c t i o n i s in fact predicted by the proposed analysis. BIBLIOGRAPHY Aoun, J . (1979b) "On Government, Case Marking and C l i t i c Placement," mimeographed, MIT. Aoun, J . (1980A) "ECP, Move and Subjacency." Linguistic Inquiry 12: 637-644. Aoun, J . (1982) " E x p l e t i v e PRO'S." MIT Working Papers in Linguistics, MIT 4:1-4. Borer, H. (1981) "Comments on the Pro-Drop Phenomena." MIT Working Papers in Linguistics, Volume 2, MIT. Borer, H. (1981) "Parametric V a r i a t i o n i n C l i t i c C o n s t r u c t i o n s , " D i s s e r t a t i o n , MIT. Borer, H. (1981) "On E x t r a c t i o n from C l i t i c Doubled C o n s t r u c t i o n , " NELS 11: 22-37. Borer, H. (1981) "Parametric Syntax", F o r i s ; Dordrecht. Chomsky, N. (1973) "Co n d i t i o n s on Transformations." i n Anderson and P. Kiparsky ( e d s . ) . A F e s t s c h r i s p to M o r r i s H a l l e , H o l t , Reinhart and Winston: New York. Chomsky, N. (1976) "Conditions on Rules of Grammar," Linguistic Analysis 2: 303-351. Chomsky, N. (1972) "On WH-movement," in P.W. C u l i c o v e r , T. Wasow and A. Akmajian (ed s . ) , Formal Syntax, Academic Pr e s s : New York. Chomsky, N. (1981) "Lectures on Government and Binding" F o r i s , Dordrecht. Chomsky, N. (1981) "Some Concepts and Consequences of the Theory of Gover nment and Binding," MIT P r e s s : Cambridge. 56 57 Chomsky, N. and Lasnik (1977) " F i l t e r s and C o n t r o l , " Linguistic Inquiry 8: 425-504. Haywood, J.A./Nahmed H.M. (1970) "A New Arabic Grammar of the Written Language," Lund Humphrires, London. Hurtado, A. (to appear) " C l i t i c Chains" in A. Hurtado (ed.) L i n g u i s t i c Theory and Spanish Syntax. R e i d e l : Dordrecht. Kayne, R.S. (1981) "ECP E x t e n s i o n s . " Linguistic Inquiry, 12: 93-134. Kenstowicz, M. (1984) "The N u l l Subject Parameter i n Modern Ar a b i c D i a l e c t s , " NELS 14:207-219. Koopman H. (1985) "Verb Movement in the Kru Languages," For i s , Dordrecht. Roberge, Y. (1985) "Subject Doubling, Free I n v e r s i o n and N u l l Agreement Languages," CLA Meeting, U n i v e r s i t y de Montreal. R i z z i L. (1978b) " V i o l a t i o n of WH. I s l a n d C o n s t r a i n t in I t a l i a n and the Subjacency C o n d i t i o n , " i n C. Dubisson, D. L i g h t f o o t and Y.C. Morin (ed s . ) , Montreal Working Papers in Linguistics 11 : 155-190. S p o r t i c h e , D. (1981) "Bounding Nodes i n French." The Linguistic Review 1 : 219-246. Sproat, R. (1983) "VSO Languages and Welsh Conf i g u r a t i o n a l i t y , " MIT Working Papers in Linguistics MIT 5:243-276. S t o w e l l (1981) " O r i g i n s of Phrase S t r u c t u r e s " . Unpublished paper. 58 Torrego, E. (1984) "On Inversion in Spanish and Some of Its E f f e c t s . " L i n g u i s t i c Inquiry 15: 103-129. Wright, W. (1971) "A Grammar of the Arabic Language," Cambridge University Press, London. INDEX AN ALTERNATIVE ANALYSIS, 19 AN OVERVIEW OF THE GB FRAMEWORK, 4 BIBLIOGRAPHY, 56 CLITICS, 29 CONCLUSION, 54 INTRODUCTION, 1 PRO-DROP ANALYSIS, 28 PRO-DROP PHENOMENA AND CLITICS., 27 THE GB FRAMEWORK, 4 THE PROBLEM, 42 THE RULE SYSTEM, 5 THE SOLUTION, 44 THE SYSTEM OF PRINCIPLES, 7 VSO ANALYSIS, 18 WH-EXTRACTION., 4 2 WORD-ORDER AND CASE ASSIGNMENT, 18 59 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0096473/manifest

Comment

Related Items