Week 14b. Bonus section: Articulating the tree

1
CAS LX 522Syntax I

• Week 14b. Bonus section Articulating the tree

2
Using the microscope
CP

• We started off with a relatively simple
  structure, with a CP, an IP, a VP.

DPj
C?
what
IiC
IP
did
DPk
I?
Pat
I
VP
ti
DP
V?
tk
V
DP
tj
eat
3
Using the microscope
CP

• As we looked closer, we had reason to think that
  the VP was more complicated, involving a
  little v.

DPj
C?
what
IiC
IP
did
DPk
I?
Pat
I
vP
ti
v?
DP
tk
Vmv
VP
eat
DP
V
tm
tj
4
Using the microscope
CP

• But for many purposes, we dont need to focus on
  the minute details of the VP. In those
  situations, youll find that people still write
  VPs like this, with the understanding that the vP
  is there.

DPj
C?
what
IiC
IP
did
DPk
I?
Pat
I
VP
ti
DP
V?
tk
V
DP
tj
eat
5
Using the microscope
CP

• What were going to do now is put IP under the
  microscope, where well find it is more
  complicated. For most purposes, we can continue
  to think about it as IP, but this is a preview
  of where syntax can go from here.

DPj
C?
what
IiC
IP
did
DPk
I?
Pat
I
VP
ti
DP
V?
tk
V
DP
tj
eat
6
Lets go back to French

• Jean mange souvent des pommes.Jean eats
  often of.the applesJean often eat
  apples.Jean souvent mange des pommes.
• Recall that this was one of our early examples
  showing verb-movement to I. French and English
  differ in whether they move finite main verbs to
  I.

IP
DPj
I?
Jean
VP
ViI
mange
tj
V?
V?
AdvP
souvent
ti
PP
des pommes
7
French negation

• This happens with respect to negation toothe
  finite verb move to the left of negative pas
• Jean ne mange pas des pommes.Jean NE eat
  NEG of.the applesJ doesnt eat apples.Jean
  pas ne mange des pommes.
• But fortunately or unfortunately, things are more
  complex that this

IP
I?
DPk
NegP
NegVijI
ne mange
Neg?
pas
tj
VP
tk
V?
ti
PP
8
French and a problem

• Finite verbs (main verbs and auxiliaries) in
  French precede adverbs and precede negative
  pasthey must move to I.
• Now lets look at infinitives, first the
  auxiliaries
• Nêtre pas invité, cest triste.NE beinf NEG
  invited, its sadNot to be invited is sad.
• Ne pas être invité, cest triste. NE NEG beinf
  invited, its sadNot to be invited is sad.
• Nonfinite auxiliaries can either move past pas
  (to I) or not, it appears to be optional.

9
French and a problem

• Fin aux V Adv, V neg Moves to I.
• Fin verb V Adv, V neg Moves to I.
• Fin aux (V) Adv (V), (V) neg (V) (Opt.) Moves
  to I.
• Nonfinite main verbsand adverbs
• Souvent paraître triste pendant son voyage de
  noce, cest rare.Often appearinf sad during
  ones honeymoon, its rareTo often look sad
  during ones honeymoon is rare.
• Paraître souvent triste pendant son voyage de
  noce, cest rare. Appearinf often sad during
  ones honeymoon, its rareTo often look sad
  during ones honeymoon is rare.
• Nonfinite main verbs can either move past adverbs
  or not optional like with auxiliaries.

10
French and a problem

• Fin aux V Adv, V neg Moves to I.
• Fin verb V Adv, V neg Moves to I.
• Fin aux (V) Adv (V), (V) neg (V) (Opt.) Moves
  to I.
• Fin verb (V) Adv (V),
• Nonfinite main verbsand negation
• Ne pas sembler heureux est une condition pour
  écrire des romans.NE NEG seeminf happy is a
  prerequisite for writeinf of.the novelsNot to
  seem happy is a prerequisite for writing novels.
• Ne sembler pas heureux est une condition pour
  écrire des romans. NE seeminf NEG happy is a
  prerequisite for writeinf of.the novels Not to
  seem happy is a prerequisite for writing novels.
• Nonfinite main verbs can not move past negation.

11
French and a problem

• Fin aux/verb V Adv, V neg Moves to I.
• Fin aux (V) Adv (V), (V) neg (V)(Opt.) Moves
  to I.
• Fin verb (V) Adv (V), neg VMoves over adv not
  neg??
• So we have the whole patternand we didnt
  predict it. Where could the verb be moving? A
  head cant adjoin to an XP, it has to be moving
  to a head. (Must remain X-bar compliant)

IP
I?
NegP
I
Neg?
pas
Neg
VP
ne
V?
DPk
V?
AdvP
souvent
PP
V
12
French and a problem

• Fin aux/verb V Adv, V neg Moves to I.
• Fin aux (V) Adv (V), (V) neg (V)(Opt.) Moves
  to I.
• Fin verb (V) Adv (V), neg VMoves over adv not
  neg??
• We need there to be a head here in the tree for
  the verb to move to
• That means we need to insert a whole phrase
  (heads always head something)

IP
I?
NegP
I
Neg?
pas
Neg
VP
ne
V?
DPk
V?
AdvP
souvent
PP
V
13
A new FP
IP
I?
NegP
I

• Fin aux/verb V Adv, V neg Moves to (F, then
  to) I.
• Fin aux (V) Adv (V), (V) neg (V)(Opt.) Moves
  to (F, then to) I.
• Fin verb (V) Adv (V), neg V(Opt.) Moves to F
• Now we have a place for nonfinite main verbs to
  move, past adverbs but under negation. They can
  move to F.

Neg?
pas
Neg
FP
ne
F?
VP
F
V?
DPk
V?
AdvP
souvent
PP
V
14
What is FP?

• Vous avez pris les pommes.you have taken the
  apples 3MSG 3FPLYou took
  the apples.
• Vous les avez prises.you them have taken
  3PL 3FPLYou took them (3fpl).
• Quelles pommes avez-vous prises?Which apples
  have you taken 3FPL
  3FPLWhich apples did you
  take?

• Vous avez pris la pomme.you have taken the
  apple 3MSG 3FSGYou took
  the apple.
• Vous lavez prise.you it have taken
  3SG 3FSGYou took it (3fsg).
• Quelle pomme avez-vous prise?Which apple
  have you taken 3FSG
  3FSGWhich apple did you take?

15
A new FP
CP

• As the verb and the object make their way up the
  tree, assuming the object moves to SpecFP, there
  is a point where the verb and object are in a
  Spec-head configuration.
• This is how the verb would check its object
  agreement features.
• Based on this, FP is generally called AgrOP.
  Object agreement phrase.

C?
DPi
IP
C
I?
DPk
I
FP
ti?
F?
VP
F
V?
tk
ti
V
16
AgrOP
CP

• AgrOP, Object agreement phrase.
• As the verb moves up to I, it has to stop off in
  AgrOP (the Head Movement Constraint requires it),
  forming successively more complex heads.
• V
• AgrOV
• IAgrOV
• But why does the object have to move to SpecAgrOP?

C?
DPi
IP
C
I?
DPk
I
AgrOP
AgrO?
ti?
AgrO
VP
V?
tk
ti
V
17
AgrOP
CP

• Why does the object have to move to SpecAgrOP?
• What makes DPs move? We know the subject moves.
  Partly for the EPP, but partly to get Case.
• The subject gets Case in SpecIP, so we know Case
  can be assigned to a specifier.
• What if we revise our notion of how objects get
  Case and say that they too get Case in a
  specifier, of AgrOP? Then it would have to move.
• Plus, its pleasingly symmetrical

C?
DPi
IP
C
I?
DPk
I
AgrOP
AgrO?
ti?
AgrO
VP
V?
tk
ti
V
18
ECM

• AgrOP can solve a serious problem we had in
  English too
• Heres the current way we analyzed ECM sentences,
  where me gets Case from want because me is in the
  government radius of want.
• The thing is, the embedded subject actually acts
  like its in the matrix clause somewhere.

IP
DPi
I?
Bill
I
VP
ti
V?
IP
V
wants
I?
DPk
1sg
I
VP
to
tk
V
leave
19
ECM v. BT

• Mary wants her to leave.
• Bill considers himself to be a genius.
• Before we said that the binding domain for
  anaphors and pronouns was a clause (say, IP).
• Her and himself above act like they are in the
  higher clause with the matrix subject.
• Our options are basically to
• complicate the definition of binding domain in
  Binding Theory
• suppose the object has really moved out of the
  embedded clause.

20
ECM
IP
DPi
I?
Bill
AgrOP
I

• If
• There is an AgrOP and
• Normal objects generally go there and
• ECM subjects act like objects
• Then
• We can suppose that ECM subjects move there.

DPk
AgrO?
1sg
AgrO
VP
ti
V?
IP
V
wants
I?
tk?
I
VP
to
tk
V
leave
21
ECM
IP
DPi
I?
Bill
AgrOP
I

• Great! Except
• But this isnt the surface word order.
• Bill me wants to leave.
• Where is BT checked? When is it important that
  pronouns be free and anaphors be bound?

DPk
AgrO?
1sg
AgrO
VP
ti
V?
IP
V
wants
I?
tk?
I
VP
to
tk
V
leave
22
ECM
IP
DPi
I?
Bill

• Whats special about ECM subjects?
• Case!
• All accusative objects move to SpecAgrOP
  (covertly in English if they dont need to move
  on) to check Case. They appear with a Case, but
  it needs to be verified by AgrO at LF.
• This is the standard interpretation of AgrOP.
• Also another example of covert movement between
  Spellout and LF.

AgrOP
I
DPk
AgrO?
1sg
AgrO
VP
ti
V?
IP
V
wants
I?
tk?
I
VP
to
V?
tk
V
leave
23
A moment of silence for Case under government

• Lets take stock here for a second.
• French told us
• There needs to be an FP between NegP and VP.
• Objects that move past FP have to stop there
  (inducing object agreement)so FP is AgrOP.
• Why do they have to stop in AgrOP?
• They need Case. So AgrOP is whats responsible
  for accusative Case.
• But V used to be responsible for that!
• Yet now we have a more symmetrical solution Case
  is always assigned in the specifier of a
  functional projection. (just about, anyway)
• And we have no more need for the government
  radius in Case assignment now that ECM is taken
  care of too.
• Plus, we have evidence from binding theory that
  objects do seem to move by LF to someplace
  outside the clause in ECM constructions.

24
A moment of silence for Case under government

• This is a step forward.
• We have a simpler theory (Case is assigned in
  only one way, we dont need the strange-looking
  construct of government radius).
• We have an account for why ECM subjects act like
  theyre in the higher clause by LF.
• Moreover, we have yet another reason to think
  that there is an LF level.
• So what does it mean for a verb to assign
  accusative case?
• Sadly, this is one place where we pay for the
  elegance elsewhereverb that assigns accusative
  case is now another name for verb that has an
  AgrOP above it.
• In Syntax II, well see a potential solution to
  even this apparent inelegance, but for now we
  just assume that transitive verbs are those with
  an AgrOP above them.

25
An AgrO you can see?

• Recall from earlier this semester that Irish is
  VSO, but yet seems to be SVO underlyingly
• Phóg Máire an lucharachán.kissed Mary the
  leprechaunMary kissed the leprechaun.
• Tá Máire ag-pógáil an lucharachán.Is Mary
  ing-kiss the leprechaunMary is kissing the
  leprechaun.
• If an auxiliary occupies the verb slot at the
  beginning of the sentence, the main verb appears
  between the subject and verb. Otherwise, the verb
  moves to first position.

26
Northern Irish

• So, basically everything points to Irish being a
  head-initial language except
• Ba mhaith liom Seán an abairt aL
  scríobhC good with.1S S.ACC the
  sentence.ACC PRT writeI want S to write the
  sentence.S writing the sentence is good with us
  (lit.)
• (cf. also I want him to meet me)
• Ba mhaith liom Seán fanachtC good with.1S
  S.ACC waitI want S to wait.

27
Morphology on French verbs

• Past, varying persons je mange-ai-s eat tu
  mange-ai-s il mange-ai-t
• Fut, varying persons je mange-er-ai eat tu
  mange-er-as il mange-er-a
• Tense morphology is inside and separate from
  subject agreement morphology.
• Kind of looks like after tense, another,
  subject-agreeing morpheme is attached

28
AgrSP?
C?
C
AgrSP

• AgrOP, Object agreement phrase.
• AgrSP, Subject agreement phrase?
• Pleasingly symmetrical!
• Complex heads
• V
• AgrOV
• TAgrOV
• AgrSTAgrOV

AgrS?
DPk
AgrS
TP
T?
ti?
T
AgrOP
AgrO?
DPk
AgrO
VP
V?
tk
ti
V
29
Split-INFL
C?
C
AgrSP

• The assumption of this structure is sometimes
  referred to as the Split-INFL hypothesis the
  INFLectional nodes have been split into subject
  agreement, tense, and object agreement.

AgrS?
AgrS
TP
T?
T
AgrOP
AgrO?
AgrO
VP
V?
DP
V
DP
30
The EPP NOM
C?
C
AgrSP
AgrS?

• We said before the T needs a specifier, thats
  the essential content of the EPP. Plus, we said
  before that this is where NOM is assigned.
• Now there is AgrSP as well.
• AgrOP is responsible for ACC.
• In a symmetrical world, seems like AgrSP should
  be responsible for NOM.
• So, now that (kind of mysterious) double
  motivation for moving to SpecIP has been
  clarified The subject has to move to both SpecTP
  and SpecAgrSP, but each movement happens for a
  different reason. T for EPP, AgrSP for NOM.

AgrS
TP
T?
T
AgrOP
AgrO?
AgrO
VP
V?
DP
V
DP
31
Adopting the Split-INFL hypothesis

• Lots of good syntax has been done both adopting
  the Split-INFL hypothesis (trees contain AgrSP,
  TP, AgrOP) or not (trees contain only IP).
• For many things, it doesnt matter which you
  chooseanalyses can be directly translated into a
  Split-INFL tree or vice-versa.
• Where it doesnt matter, it doesnt matter, but
  sometimes it matters.

32
Adopting the Split-INFL hypothesis

• The general program is that every dissociable
  piece of the structure should get its own place
  in the lexicon, its own functional head
• Subject agreement is basically common across
  verbs, an independent piece.
• Tense too is an independent piece.
• And object agreement
• And plural marking and progressive
  -ing,aspectual -en,
• In Syntax II, well spend a lot of the semester
  looking at places in the tree where functional
  projections need to be added.

33
Split-INFL
CP
C?

• In recent literature, almost everything you read
  will make this assumption, that
  cross-linguistically, the clause is minimally
  constructed of these projections, generally in
  this order
• CP
• AgrSP
• TP
• AgrOP
• VP

C
AgrSP
AgrS?
AgrS
TP
T?
T
AgrOP
AgrO?
AgrO
VP
34
Split-INFL
CP
C?
C
AgrSP

• Another line of thought (described by Radford in
  ch. 9) puts them in a different order (with AgrOP
  between vP and VP), but the same idea
• CP
• AgrSP
• TP
• vP
• AgrOP
• VP
• There are various empirical and theoretical
  advantages and disadvantages to this order they
  jurys still out.

AgrS?
AgrS
TP
T?
T
vP
v?
v
AgrOP
AgrO?
AgrO
VP
35
?

• ? ?
• ?
• ? ?
• ? ?
• ?
• ?