CAS LX 523 Syntax II Spring 2001 Paul Hagstrom (1) CP 3 IP 3 VP

January 16, 2001 Week 1: Introduction, split-Infl, and argumentation The clause structure of the olden days (say, 1986): CP: force (declarative, interrogative) IP: functional morphology (tense, agreement) VP: lexical/thematic elements

But this doesnt seem to be quite good enough to describe/explain the dataprogressively, they have been split up into further functional projections. More modern clauses tend to look more like this (at least):

(2)

ForceP CP 3 TopP* 3 FocP 3 TopP* 3 FinP 3 AgrSP IP 3 TP 3 NegP 3 AgrOP 3 vP 3 VP

VP

Syntactic argumentation: Present the puzzle (recalcitrant data) Present a hypothesis to account for the data. Consider what else your hypothesis predicts. Check to see if the predictions are met. if so, celebrate them; if not, speculate as to why not (or revise your hypothesis).

The road to (2) began in 1989 Pollock, Jean-Yves (1989). Verb movement, universal grammar, and the structure of IP. Linguistic Inquiry 20(3):365424. Basic tasks: Present data showing that there needs to be another position in the clause structure between negation and adverbs (named: AgrP). Explain the special behavior of have and be both in French and in English. (The paper contains a couple of other points concerning NegP as well)

1.

French: V moves to I; English: have/be move to I (I lowers to other verbs). French and English share a common D-structure. Adverbs and not cant move to the right.

Assume:

(3)

IP 3 subject I 3 I (Neg) 3 not/pas VP 3 (adverb) VP # ... V ... a. b. a. b. c. d. * John likes not Mary. Jean (n) aime pas Marie. * John kisses often Mary. Jean embrasse souvent Marie. John often kisses Mary. * Jean souvent embrasse Marie.

(4) (5)

French negative object rien nothing/anythingmust move, marks left edge of VP. (6) VP 3 rieni VP # ... V ti

(7)

a.

Pierre na rien mang. Pierre ne has nothing eaten Pierre hasnt eaten anything. * Pierre na mang rien. Pierre ne mange rien. Pierre ne eats nothing Pierre doesnt eat anything. * Pierre ne rien mange.

b. c.

d.

beaucoup lots moves optionally. (8) a. b. c. d. Pierre a lu beaucoup de livres. Pierre has read lots of books. Pierre a beaucoup lu de livres. Pierre lit beaucoup de livres. * Pierre beaucoup lit de livres.

Conclusion: All (finite) verbs move to I in French; English finite have and be do too. 2. (Some) infinitives (can) move in French. Finite and non-finite sentences are the same except for [finite] on I. not and nepas stand in the same structural position in tensed clauses and infinitives. Ne pas tre heureux est une condition pour crire des romans. Ne to not be happy is a prerequisite for writing novels. Ntre pas heureux est une condition pour crire des romans. Ne to be not happy... Ne pas avoir eu denfance heureuse est une condition pour crire des romans. Ne not to have had a happy childhood is a prerequisite for writing novels. Navoir pas eu denfance heureuse est une condition pour crire des romans. Ne to have not had a happy childhood...

Assume:

(9)

a. b. c.

d.

Conclude: Verb movement to I is optional for infinitives in French.

But wait (10) a. b. c. d. So: Ne pas sembler heureux est une condition pour crire des romans. Ne not to seem happy is a prerequisite for writing novels. * Ne sembler pas heureux est une condition pour crire des romans. (Ne to sem not happy...) Ne pas regarder la tlvision consolide lesprit critique. Ne not to watch television strengthens ones independence. * Ne regarder pas la tlvision consolide lesprit critique. (Ne to watch not television...)

Only infinitives of have (avoir) and be (tre) can (optionally) move in French. Lexical infinitives cannot move in French.

Quelle concidence! Both English and French have restrictions on moving lexical verbs, but not on moving have/avoir and be/tre. (11) a. b. ? c. d. (?) a. b. d. e. Not to be happy is a prerequisite for writing novels. To be not happy is a prerequisite for writing novels. Not to have had a happy childhood is a prerequisite for writing novels. To have not had a happy childhood is a prerequisite for writing novels.

(12)

Not to seem happy is a prerequisite for writing novels. * To seem not happy is a prerequisite for writing novels. Not to get arrested under such circumstances is a miracle. * To get not arrested under such circumstances is a miracle.

3.

Verb movement to I isnt really always verb movement to I.

Verb movement for French infinitives is never obligatory. Predicts that we can see adverb verb NP sequences (never allowed with finite verbs). Et voil. (13) a. Souvent paratre triste pendant son voyage de noce, cest rare. often to.look sad during ones honeymoon thats rare. To often look sad during ones honeymood is rare. Preseque oublier son nom, a narrive pas frquemment. almost to.forget ones name that ne happens not frequently. To almost forget ones name doesnt happy frequently.

b.

(14)

* Pierre presque oublie son nom. Pierre almost forgets his name.

Ok, now infinitives can move, but only avoir and tre, right? Predicts that lexical infinitives preceding adverbs should be ungrammatical But (15) a. b. So: But: (16) Paratre souvent triste pendant son voyage de noce, cest rare. To look often sad... Oublier presque son nom, a narrive pas frquemment. To forget almost ones name...

Any old infinitive verb seems to be able to move past adverbs. Remember from before, only avoir and tre can get past negation. a. b. * Ne paratre pas triste pendant son voyage de noce, cest normal. Ne to look not sad... * Noublier pas son nom, ce nest pas un explot. Ne to forget not ones name isnt worth writing home about.

Summary of the puzzle. French: V[+fin] pas adverb _ V[fin]* pas adverb _ pas V[fin] adverb _ pas adverb V[fin]

English:

V[+fin]i* not adverb _ not adverb V[+fin] not V[fin]* adverb _ not adverb V[fin]

*Only have/avoir and be/tre. Lexical verbs (verbs other that have/avoir and be/tre). French: All finite verbs raise past pas. All nonfinite verbs can raise past adverbs. Only nonfinite avoir/tre can raise past pas. Only finite have/be can raise past not. Only nonfinite have/be can raise past adverbs. (see below)

English:

4.

The Split-INFL hypothesis: Introducing AgrP.

Hypothesis: There is a position between negation and the adverbs at the left edge of VP where French verbs move to. This is short verb movement. Not all the way to I. French: (17) (18) (19) a. b. a. b. Short verb movement is not lexically restricted (to tre and avoir) any old nonfinite verb can undergo short verb movement. To often look sad during ones honeymoon is rare. To almost forget ones name doesnt happen frequently. * To look often sad during ones honeymoon is rare. * To forget almost ones name doesnt happen frequently.

a. I believe John to often be sarcastic. b. I believe John to often sound sarcastic. c. (?) I believe John to be often sarcastic. d. * I believe John to sound often sarcastic. English short verb movement is lexically restricted to have and be.

Conclude:

So, we have a correlation: English: Short verb movement of infinitivesonly have and be. V I of finite verbsonly have and be. Short verb movement of infinitivesany old verb. V I of finite verbsany old verb. V I is lexically restricted iff short verb movement is.

French:

We can say this like this:

This would be predicted if V I contained an instance of short verb movement: That is, V I is not a one-step jump, its short movement to a middle position, and then a movement from there to I.

(20)

IP 3 NP I 3 I 3 (pas/not) AgrP 3 Agr VP short movement 3 (adv) V

*HMC

5.

Verb movement and -theorywhats special about have and be?

What differentiates have and be from other verbs? Pollocks suggestion: other verbs have -roles to assign; have and be do not. Whats the difference between V Agr/T and affix hopping? One adjoins V to Agr/T, the other adjoins Agr/T to V: (21) a. V Ru V Agri b. Agr rU Vi Agr

Suppose V needs to assign -roles. What might be the problem in (21b)? Pollocks hypothesis: French Agr is rich enough to transmit Vs -role. In French, (21b) can assign a -role. French Agr is transparent English Agr is not rich enough to transmit Vs -role. In English, (21b) cannot assign a -role. English Agr is opaque So, if V has a -role to assign, and Agr is opaque (English), V cant move to Agr. (if V doesnt have a -role to assign, V can happily move to Agr: have, be) And, if Agr is transparent (French), no problem, move away.

What about the French restriction on nonfinite lexical verb movement to I? Pollocks hypothesis here: [finite] tense is (perhaps universally) opaque. Summary: There is a head position between negation and the VP (where adverbs are). Pollock proposes this is Agr (hence, AgrP). IP is more perspicuously renamed TP (TenseP) now, since it just has tense features. Agr can differ in opacitytransparent Agr (rich inflection) allows -transmission. [finite] T is always opaque. Quick mention of Pollocks proposal about NegP: (22) NegP 3 pas Neg 3 Neg ... ne NegP NegP 3 3 not Neg or maybe Neg 3 3 Neg ... Neg ... not

ne is a clitic (like French pronominal clitics le, la) and has to move to (lean on) T.

Belletti, Adriana (1990).

Generalized verb movement. Torino: Rosenberg & Sellier

Hold on there, Tex. Sure, we need to split up INFL. But why that way? (23) a. b. c. (24) a. b. c. (25) a. b. NP je tu il NP je tu il V parl parl parl speak V parl parl parl speak T ai ai ai
PAST

Agr s s t person+number Agr ai as a person+unmber

French

T er er er
FUT

Legg-eva-no they read (read-imperfect-3pl). Parl-er- I will speak (speak-fut-1sg).

Italian

(26)

a. b.

...werk-t-en (they) work-past-3pl ...werk-t-e (she) work-past-3sg In general, tense is closer to the verb root than subject agreement.

Dutch

Conclude:

If the finite verb moves to one and then the other, it would look like this: (27) F1P 1 F1 3 F1 F2P 1 F2 3 V+F2 VP 1 V 3 t ... (28) F1P 1 F1 3 [V+F2]+F1 F2P 1 F 2 3 t VP 1 V 3 t ...

So, in [V+F2]+F1, F2 is closer to the verb and is the head that was moved to first. F2 is the lower head. F2 is closer to the verb. The tense affix is closer to the verb. F2 = T. TP is below AgrP. Mirror Principle. Morphological derivations (suffixation, prefixation, etc.) directly reflect syntactic derivations... (Baker 1985, 1988) So, we should really have clauses that look like this: (29) CP 3 AgrP 3 TP 3 VP

Chomsky, Noam (1989).

Some notes on economy of derivation and representation. Reprinted in (1995) The Minimalist Program, MIT Press.

Youre both right. Pollocks AgrP is actually object agreement. Bellettis AgrP is subject agreement. (30) AgrSP 3 subject AgrS 3 AgrS TP 3 T (NegP) 3 Neg AgrOP 3 AgrO VP rU (adverb) VP

Kayne (1989). Facets of past participle agreement in Romance. In P. Beninc (ed.), Dialect variation and the theory of grammar. Dordrecht: Foris. (31) a. b. Combien de tables Paul a repeintes? Combien de tables Paul a repeint? How many tables did Paul repaint?

Idea: On the way up combien de tables (feminine, plural) can (or not) stop in SpecAgrOP. (32) ... AgrOP 3 (ti) AgrO 3 AgrO VP # repeint- ti a. b. Paul a repeint les tables. * Paul a repeintes les tables.

(33)