1. Kayne, R. 1994. The Antisymmetry of Syntax. MIT Press. Chapter 8 Relatives and Possessives Kayne (1994) Consider the set A of ordered pairs such that for each j, X j asymmetrically c- commands Y j. Let us further take A to be the maximal such set; that is, A contains all pairs of nonterminals such that the first asymmetrically c-commnads the second. Then the central proposal I would like to make is the following (for a given phrase marker P, with T the set of terminals and A as just given): (Kayne 1994, 5-6) (3) LCA (Linear Correspondence Axiom) d(A) is a linear ordering of T. (d: the nonterminal-to-terminal dominance relation)

2. Chapter 2 Deriving X-Bar Theory (1) K J L j M N m P p - the pairs that constitute the set A:,,, (i.e., the pairs of nonterminal nodes such that the first asymmetrically c-commands the second) d(A):,, These three ordered pairs do constitute a linear ordering of the set {j, m, p}, given that (1)transitivity holds (2)Antisymmetry is respected (3)The ordering is total

3. English vs. Amharic (38=10) the [[ NP picture][that [Bill saw [e]]]] (39)the [[ NP picture][that IP]] moving IP to Spec,DP yields (40) (40) IP j [the [[ NP picture][that [e] j ]]] (41) IP j [the [[ NP picture][C 0 [e] j ]]] with both D 0 & C 0 empty = (42) IP j [D 0 [[ NP picture][C 0 [e] j ]]] p. 93 N-first vs. N-final languages:

4. 8.1 Postnominal Possessives in English In (1) the phrase of John’s is not plausibly a complement of picture. (1)I have two pictures of John’s. Cf. two pictures of Mary of John’s (2) D 0 [John [’s [two pictures]]] e.g., in Hungarian, possessors are prenominal; can be in nom Case; preceded by the definite article; (*)the John’s two pictures cf. indefinite D 0 that can precede the possessor phrase  the possessor must move into the specifier of the D 0, picks up dative Case, and then out of the DP entirely D is a valid Case-licensing position (however, indef. D 0 is not a Case licenser) [D 0 [John [’s [two pictures]]] (3) [two pictures] i [[ D of][John [’s [e] i ]]] Quantifiers, every, many, some, any, a(n): generable within the NP/QP below ’s. Def. det. the; uniquely a D 0, not to be generable below ’s (4) *?I found the pictures of John’s/his. (5) *?I found the two pictures of John’s/his.

5. 8.2 Relative Clauses in English Why (6) is fully grammatical? (6) I found [the [(two) pictures of John’s/his [that you lent me]]] constituent head complement Cf. (5) *?I found [the [two pictures of John’s/his]]. pictures ≠ head ≠ constituent head *complement (7) [ CP (two) pictures of John’s] i [that […me [e] i ]] Raising analysis (Vergnaud, 1974) head not a complement of N (8) John bought the picture of himself that Bill saw. nor right-adjoined to N (9) The [picture of himself [that [Bill saw [e]]]] (right adjunction is banned) complement of D (the): [ DP D 0 CP] = (10) The [[ NP picture][that [Bill saw [e]]]]

6. In Romanian, the head noun of the NP in Spec,CP will raise out of CP and left-adjoin to D 0 (11) Cartea pe care am citit-o book-the pe which I-have read it In French and Italian the structures in (13), (15), and (17) are ungrammatical because once cui, qui, and la femme de qui have filled Spec,CP, there is no ‘room’ for persona, personne or homme. There is extra ‘room’ in (14), (16), and (18) because the preposition in those examples provides it by making its specifier position available. (13) *la persona cui Bill ha visto (Italian) the person who Bill has seen (14) la persona con cui Bill ha parlato the person with whom Bill has spoken (15) * la personne qui Bill ha vue (= 13) (French) (16) la personne avec qui Bill ha parlé (=14) (17) *l’homme la femme de qui tu as insulteé the man the wife of who you have insulted (18) l’homme avec la femme de qui tu t’es disputé the man with the wife of who you REFL-is argued ‘the man with whose wife you argued’

7. Relative pronouns originate as determiners that are split off from their associated NP by movement of the latter (19) the [C 0 [he broke it with which hammer]] wh-movement of the PP to Spec,CP  (20) the [with which hammer [C 0 [he broke it [e]]] The NP hammer then raises to Spec,PP (probably via Spec,which)  (21) the [ CP [ PP hammer i [with which [e] i ]] [C 0 … (16)  (22) la [C 0 [Bill a parlé avec qui personne]] wh-movement of the PP to Spec,CP  (23) la [avec qui personne [C 0 [… The NP hammer then raises to Spec,PP (probably via Spec,which)  (24) la [ CP [ PP personne i [avec qui [e] i ]] [C 0 … When the constituent moved to Spec,CP is headed by a preposition, those NPs can raise to the specifier of that preposition.

8. (25) * la [qui personne [C 0 … Cf. (24) la [ CP [ PP personne i [avec qui [e] i ]] [C 0 … In English, they use the specifier position of the wh-determiner itself as a landing site: (why they differ from F & I? no solution yet) (28) the [which picture [C 0 …  (29) the [ CP [ DP picture i [which [e] i ]] [C 0 … (30) the [ CP [[ DP man i [who [e] i ]]’s wife [C 0 … who the man’s wife is moved to Spec,CP, man is moved to the inner Spec,DP (D = who) Longer movement only in English: (31) The book the author of which I know personally (32) ??the man the possibility of you marrying whom became a reality only yesterday (33)*the man the possibility of [who(m) marrying you] became a reality only yesterday  Subject island movement violation A well-formed ‘headed’ restrictive relative clause structure requires that personne reach in the overt syntax a position governed by D 0.

9. English rel-cl: N-initial only as a result of leftward mvnt. cf. many other langes. have rel-cl structures in which the noun follows the rel-cl. p. 93 Specifier positions are always on the left: Since specifiers are an instance of adjunction and since adjunction is always left-adjunction Therefore, the final position of N in rel-structures in langes like Japanese differs from the N-initial one of langes like English. (37) a. N-final relatives lack relative pronouns. b. N-final relatives never display a complementizer that is identical to the normal complementizer of sentential complementation. N-initial & N-final relatives are essentially identical: then, N-final relatives must involve the same ‘D 0 CP’ structure that N-initial relatives have. 8.3 N-Final Relative Clauses

10. p. 93 N-final languages: lack any equivalent of English the; the D 0 will not be visible. Amharic language (Gragg, 1972): provides a significant clue to the syntax of D 0 in N-final relatives: if starting from ‘D 0 CP’, the entire CP moved to Spec,DP, then we would not expect N to follow D. Relative clause that precedes the definite article in Amharic is not CP; The pre-D relative clause must be a projection smaller than CP; let us call it IP English vs. Amharic (38=10) the [[ NP picture][that [Bill saw [e]]]] (39)the [[ NP picture][that IP]] moving IP to Spec,DP yields (40) IP j [the [[ NP picture][that [e] j ]]] (41) IP j [the [[ NP picture][C 0 [e] j ]]] with both D 0 & C 0 empty = (42) IP j [D 0 [[ NP picture][C 0 [e] j ]]] IP remains in situ, so that C 0 can be overt and Can be identical to the normal sentential complementizer. 8.3 N-first vs. N-Final Relative Clauses

11. N-final relativization does not admit the presence of a relative pronoun. (42=44) IP j [D 0 [ CP [ NP picture][C 0 [e] j ]]] cf. Cole (1987) Starting from a ‘D CP’, the NP picture is moved to Spec,CP and then the IP complement of the empty C 0 is moved to Spec,DP. That IP contains the trace of picture. (45) [ IP … [e] j …[D 0 [ CP [ NP picture] j [C 0 … a copy of the moved constituent (Chomsky 1993) (46) [ IP … [picture] j …[D 0 [ CP [ NP picture] j [C 0 … (47) [ IP … [picture] j …[D 0 [ CP [ NP e] j [C 0 … The English structure: (48) D 0 [ CP [ NP picture] j [C 0 [ IP … [e] j … (49) *D 0 [ CP [ NP e] j [C 0 [ IP … [picture] j … (50) A given chain link c k can license PF deletion of another link c 1 of the same chain only if does not c-command c k. 8.3 wh-relatives in English

12. (51) *?We were admiring the sweater of his. (52) We were admiring the sweater of his that was lying on the sofa. (53) (?)We were admiring the (one) sweater of his given to him by his wife. (reduced rel. IP, ≠CP) (54)*the sweater of John’s that/which given to him by … Vergnaud (1974, p. 173ff.): the ‘reduced’ relatives exhibit the behavior of ordinary relatives. When rel.es are ‘stacked’, the 2 nd one can neither have a zero complementizer nor be’reduced’. (55) I just read the book that you told me about *(that) your son gave me last year. (56) I just read the book that you told me about *?(that was) given to me by your son last year. (57) the [ CP [ NP book] i [C 0 [ IP [e] i sent to me (58) *the [ CP [book i which [e] i ] j [C 0 [ IP [e] j sent to me (59) *the book John fond of (60) the [ CP [C 0 [ IP [book] i [I 0 [ XP [e] i sent to me (61) the [ CP XP j [C 0 [ IP [book] [I 0 [e] j (62)*the sent to me book (63) the recently sent book 8.4 Reduced Relatives and Adjectives