CAS LX 522 Previously, in LX522…

Syntax I • Last time, we looked at the TP

phenomenon of head-movement. SS
DP T′
• Recall, for example, French,
which moves V up to T as shown T VP
here.
Week 6. NP/DP movement • At DS, the verb heads the VP, Vi T V′
and Case and by SS, the verb has moved to mange [PRES]
head-adjoin to T. AdvP V′
• This was proposed in order to ti PP
account for word order facts.

TP

Previously, in LX522… It is likely… T′ DS

T VP
• Let’s think back to the case of [pres]
• Today, we’re going to look at TP
SS It is likely that Mary left from a V′
another kind of movement, the
movement of DPs. DP T′ couple of weeks ago.
• Likely has one θ-role to assign V
be
AdjP
T VP
• In many respects, the idea is (Proposition) which it assigns Adj′
similar—a DP will originate in Vi T V′ to its complement, the
mange [PRES] embedded CP. Adj CP
one place in the DS and will likely θ
appear in a different place in the AdvP V′
• Consider leave in the C′
SS. ti PP embedded clause. Leave also
has one θ-role to assign, which C
that
TP
θ
it assigns to Mary. Mary left

TP TP

It is likely… T′ DS It is likely… DP T′ SS
• Notice that both θ-roles are T
[pres]
VP • Moving to SS… D′ Vi +T VP
is
assigned to things that are in V′ • Because the EPP requires D V′
the same clause as the predicate SpecTP to be filled, it
that assigns the θ-role. V
be
AdjP Expletive Insertion applies, ti AdjP

Adj′ inserting it into SpecTP, Adj′

• This is a general property of θ- resulting in this SS
role assignment:
Adj
likely θ
CP representation. Adj
likely
CP

• A θ-role must be assigned C′ C′

locally (within the same • This is the story of
C TP C TP
clause). that θ
It is likely that Mary left. that
Mary left Mary left

1
 It is likely… It is likely…
• Now, consider:
– Mary is likely [to leave]. – Mary is likely [to leave]
θ
• We already know a lot about this sentence; we
know that likely has one θ-role to assign, which it • Concerning θ-roles, it’s clear from the
assigns to the embedded clause, we know that meaning that leave really does assign its θ-
leave has one θ-role to assign, which it assigns to role to Mary and not likely (Mary is
Mary. leaving—she’s isn’t in any way likely).
• There are two problems here: • This is definitely not local—Mary is not in
– The embedded clause has no subject (*EPP)
the same clause as leave.
– The θ-role assigned to Mary seems to be assigned
outside of its clause.

It is likely… It is likely…
– Mary is likely [to leave] – Mary is likely [to leave]
• For θ-role assignment to be local, Mary has to
• And with respect to the EPP, we see that be in the same clause. θ-role assignment takes
although the main clause TP has something place at DS, after which movement rules (like
in its specifier (Mary), the embedded clause head-movement from last time) apply. We can
solve both problems at once by supposing that
seems to have nothing.
Mary moves from the embedded subject position
at DS to the main clause subject position at SS.
• How can we reconcile this? – DS: — is likely [Mary to leave]
– SS: Maryi is likely [ ti to leave]

TP TP

T′ DS DP i T′ SS
It is likely… It is likely… Mary
T VP Vj+T VP
• That is, we start out [pres] V′ • That is, we start out is V′
with Mary in the with Mary in the
embedded clause, in V AdjP embedded clause, in tj AdjP
be Adj′ Adj′
the specifier of TP, the specifier of TP,
receiving its θ-role Adj CP receiving its θ-role Adj CP
likely C′ likely C′
locally. locally.
C TP • Then Mary moves up C TP
Ø Ø
to SpecTP in the
DP T′ ti T′
Mary main clause by SS.
T VP T VP
to to
θ leave leave

2
 TP TP

DP i T′ SS DP i T′ SS
It is likely… Mary It is likely… Mary
Vj+T VP Vj+T VP
• Notice that this is V′ • This type of is V′
satisfies the EPP in movement is called
both clauses. The tj AdjP DP-movement. tj AdjP
Adj′ Adj′
main clause has Mary • This specific instance
in SpecTP. The Adj CP of DP-movement, Adj CP
likely C′ likely C′
embedded clause has where we move a
the trace in SpecTP. C TP subject from an C TP
Ø Ø
embedded clause to a
ti T′ ti T′
higher clause is
T VP generally called T VP
to subject raising. to
leave leave

TP

DP i T′ SS
It is likely… Mary Passive
Vj+T VP
• Historical idiosyncrasy: is V′
Because a lot of terminology • Now, recall the passive. The passive form of a verb
was established before the tj AdjP seems to directly affect the theta grid of a verb;
DP had been “discovered,” Adj′ consider:
people often still, out of
Adj CP – Bill ate the sandwich.
habit, refer to this kind of likely C′
movement as NP-movement – The sandwich was eaten.
rather than DP-movement. C TP • Eat has two θ-roles to assign. By putting it in the
Ø
These are not different passive, we seem to have transitive (two θ-role) verb
ti T′
things: People who say NP- into an intransitive (one θ-role) verb.
movement generally mean
T VP
DP-movement. to
leave

Passive Passive
• Since optional thematic relations do not get included in
– Bill ate the sandwich. the θ-grid, what we conclude about the passive is that it
• Here, Bill is the Agent (gets the θ-role changes the θ-grid of the verb by removing the external θ-
including Agent) and the sandwich is the role.
Theme (gets the θ-role including Theme).
– The sandwich was eaten (by Bill). eat Agent Theme
• In the passive, the roles are the same but i j
now the Theme is the subject and the Agent
is in an optional by-phrase (a PP). eat+en Agent Theme
i j

3
 Passive Active
• Now, what does the structure of a passive • Let’s start with the DS tree for the
active sentence, Bill ate the TP
sentence look like? sandwich. DS
• There are two possibilities we could entertain. • Here, the (internal) Theme θ-role DP T′
– The Theme in the passive becomes an external θ- is assigned to the object DP and
the (external) Agent θ-role is Bill T VP
role (as opposed to in the active, where the Theme [past]
assigned to the subject DP.
gets an internal θ-role). V′
– The Theme in both cases gets an internal θ-role,
• Now, suppose that for the passive θ V θ DP
but in the passive, it moves to the subject position. we simply eliminate the external θ- eat
• Let’s pursue the second option first… role… the
sandwich

TP
Passive TP
Passive DS – The sandwich was eaten. DS
• (The passive also requires the T′ • Now, what needs to happen? T′
addition of the auxiliary verb be, – SpecTP must be filled (EPP).
T VP T VP
but this is not relevant to the point [past] – The word order needs to be altered [past]
at hand) V′ from was eaten the sandwich to the V′
• We have changed the main verb to sandwich was eaten.
the passive form, thereby V VP V VP
be be
removing the external θ-role, • It should be clear where this is
leaving us with this DS for V′ V′
going—here, we posit another
– The sandwich was eaten. V θ DP instance of DP-movement, like with V θ DP
eaten raising. In the passive, the object eaten
• Now, what needs to happen? the moves to SpecTP satisfying the the
sandwich EPP. sandwich

Passive TP
Passive TP
– The sandwich was eaten. SS – The sandwich was eaten by Bill. SS
DP i T′ DP i T′
• So, to review, the idea is that • As for the optionally expressed
the active and the passive the Vj+T VP Agent in the by-phrase, we take the Vj+T VP
have very similar DS sandwich was this to be like any optionally sandwich was
representations, except that V′ expressed adjoined phrase, a PP V′
the passive has had its adjoined to V′.
external θ-role removed and tj VP • As expected, the by-phrase can be tj VP
thus no subject is generated re-ordered with respect to other
in SpecTP (as required by V′ V′
adjuncts.
the Theta Criterion). Then – The sandwich was eaten…
V ti V′ PP
the object moves into eaten – …by Bill under the tree at noon.
SpecTP, satisfying the EPP – …under the tree by Bill at noon. V ti by Bill
at SS. eaten
– …at noon under the tree by Bill.

4
 Passive Not the passive
• Let’s return for a moment to the two • The first option hypothesizes that the passive
possibilities we could have entertained… form of the verb removes the external θ-role and
– The Theme in the passive becomes an external θ- promotes the internal θ-role to an external θ-
role (as opposed to in the active, where the Theme role: eat Agent Theme eat+en Agent Theme
gets an internal θ-role).
– The Theme in both cases gets an internal θ-role, i j i j
but in the passive, it moves to the subject position.
• Under this view, then, the Theme is not moved
• We have worked out what the second option
into SpecTP but rather just starts out there.
looks like, let’s take a second to see why the
first option wouldn’t have worked.

Not the passive Not the passive

• Consider this active sentence. • The predicted result is:
– Wilma considers [Fred to be foolish]. – *[Fred to be foolish] was considered.
• And suppose we want to make a passive. We • …which is not what we want. Rather, what
eliminate the external θ-role from considers we want is:
– Fred was considered [to be foolish].
(meaning the role assigned to Wilma above).
Then we make the internal θ-role (assigned to • But notice, Fred was never assigned a θ-role
by considered (Fred’s θ-role comes from
the embedded proposition) external. What
foolish) so we couldn’t have changed the θ-
should the result be? role Fred got to be external.

Passive Nagging questions

• Things have been working out well so far, but
– Fredi is considered [ ti to be foolish]
there are a couple of things that are still
• However, the account of the passive that we unexplained…
developed before, where the object moves – If in the passive, movement of the object into
into SpecTP has no trouble explaining this. subject position is done in order to satisfy the EPP,
This is basically a case of subject raising, why couldn’t we instead insert it in SpecTP like
we do in it rains or it is likely that…?
the EPP needs to be satisfied and is satisfied
– Similarly, for raising, what is wrong with *It is
by moving Fred into the main clause’s likely John to leave?
SpecTP. • The answer to this will be Case—which we
turn to now.

5
 Case English pronouns and case
• As has been mentioned before, many languages
mark the grammatical relations of their DPs • In English, although we generally don’t
with case markers. mark the grammatical relations with case…
– Korean: ka/i = subject, (l)ul = object – The president met the students.
Chelswu-ka Sunhi-lul manna-ss-ta – The students met the president.
Chelswu-nom Sunhi-acc met-past-decl • …we do mark the grammatical relations of
‘Chelswu met Sunhi.’
the pronouns with case…
– Japanese: ga = subject, o = object
– He met her.
Akira ga ringo o tabeta
Akira nom apple acc ate
– She met him.
‘Akira ate an apple.’

English pronouns and case In the spirit of global unity…

• Given that
• A pronoun in subject position of a finite clause
– some languages show case marking on all nouns
has nominative (subject) case: (not just pronouns)
– I left; he left; she left; we left; they left. – in English we see case marking on at least some
• A pronoun in object position has accusative nouns (the pronouns)
(object) case: – We’re striving to create a syntactic system that
– J met me; J met him; J met her; explains all languages
J met us; J met them. • We will suppose that all English nouns get case
too, it’s just that you can’t see it on anything but
the pronouns.

In the spirit of global unity… In the spirit of global unity…

• This is in a sense an extension of the idea • That is, there is an abstract tense/agreement
that even though you can’t see a present suffix which is always present and which
tense marker on walk in you walk and I walk, can be morphologically realized in a couple
the fact that we do see it on he walks and of different ways.
the fact that we see past tense markers on I
walked and you walked, we simply assume • Returning to Case, we suppose that there is
that there is always a tense/agreement suffix, an abstract Case marker on all nouns, but
but that sometimes it is pronounced as -ed, that it is morphologically realized as Ø in
sometimes as -s, and sometimes as Ø. English except on the pronouns.

6
 Case Case vs. θ-roles
• Case is tied to syntactic position; a subject (that
is, the DP in SpecTP) gets one Case • It is important to notice that Case is not
(nominative), the object (sister of a transitive V) correlated with θ-roles.
gets a different Case (accusative). – I met him (at the airport).
– He was met by me (at the airport).
• We formalize this idea that all nouns have
abstract Case by making it a requirement—all • In both sentences, the Theme is the
nouns in a grammatical sentence must show same—him. But in the first sentence, him is
their syntactic position. marked with accusative Case, and in the
second sentence he is marked with nominative
Case.

Case vs. θ-roles Case Theory

• It is important to notice that Case is not • Case Filter (SS)
correlated with θ-roles. All DPs must have Case
– I met him (at the airport).
– He was met by me (at the airport). • Case is available (roughly)
– To the specifier of a finite T (nominative)
• Case has to do with where the DP ends up at – To the sister of a V or a P (accusative, oblique)
SS, and θ-roles have to do with where the DP
starts out at DS.

Case Theory Privileged positions

• The idea is that there are a few privileged • In particular, there are certain elements of the
positions in the syntactic structure in which structure which are Case-assigners. These are
Case is available—if a DP starts out in a things which can provide Case to a DP.
position where no Case is available, it must – Finite T is a Case assigner, it provides
move to a position where it can get Case (or nominative Case.
face ungrammaticality). – Transitive verbs are Case assigners, they provide
accusative Case.
– Prepositions are Case assigners, they provide
oblique Case.

7
 Licensing Accreditation revoked
• In order to get Case from a Case-assigner, • The thing which makes Case Theory run is
the DP has to be close to the Case-assigner the fact that under certain situations T or V
– (we’ll postpone discussion of what exactly it
cannot assign Case.
means to be “close” for a while).
• Some places which are close enough to get • For T, only finite T is a Case-assigner— a
case are SpecTP (close to T) or sister to V nonfinite T does not assign Case.
(close to V). • For V, only transitive verbs assign Case—
intransitive verbs and passive verbs do not
assign Case.

TP TP

T′ DS DP T′ SS
Back to raising… Back to raising… Mary i

T VP Vj+T VP Finite T
• Let’s go back to Mary is [pres] V′ • When the DP Mary is V′
likely to leave. Recall moves up to the main can assign
that this is the DS. clause SpecTP, it gets Case
V AdjP tj AdjP
be Adj′ close enough to the Adj′
• In the embedded clause, finite T to receive Case
Adj CP (thus satisfying the Case Adj CP
Mary is in SpecTP, but likely C′ likely C′
nonfinite T cannot filter).
assign Case. C TP C TP
Ø Ø
• So, this movement does
Nonfinite T DP T′ two things: It satisfies ti T′
• Unless the DP Mary cannot assign
Mary
moves, the Case Filter T VP
the EPP and it satisfies T VP
Case
will be violated at SS. to the Case Filter. to
θ leave leave

TP TP

Back to raising… * DP T′ SS Back to raising… DP T′ SS

D′ Vj+T VP Mary violates D′ Vj+T VP She gets
• Notice that this explains is V′
the Case Filter • When the embedded is V′
Case from T
why… D clause is finite… D
it tj AdjP it tj AdjP
– *It is likely Mary to leave Adj′ – It is likely that she left. Adj′
• …is ungrammatical, • …everything is fine
Adj CP Adj CP
though: Even though the likely C′ because she gets likely C′
sentence satisfies the EPP, (nominative) Case
it violates the Case Filter C TP from the embedded C TP
Ø that
(Mary doesn’t get Case). DP T′
finite T. DP T′
Nonfinite T Mary Finite T she
cannot assign T VP assigns nom. T VP
to -ed
Case Case
leave leave

8
 Back to passives… Back to passives…
• We had a similar question about what was • What we can say here is that the addition of
wrong with: the passive morpheme -en to a transitive
– *It was eaten the sandwich
verb not only removes its external θ-role,
but also revokes its ability to assign Case.
• …where it appears that even though the
EPP could be satisfied by inserting the
• Burzio’s Generalization
expletive it, the sentence is still
A verb which does not assign an external θ-
ungrammatical. role cannot assign accusative Case.

Bill gets
Active again… Active again… Case from T

• Let’s review the DS tree for the • At SS, Bill gets (nominative) Case
active sentence, Bill ate the TP from the finite T, and the sandwich TP
sandwich. DS gets (accusative) Case from the V. SS
DP T′ DP T′
• Here, eat assigns two θ-roles, the Finite T
Bill Bill
internal θ-role (Theme) to the DP T
[past]
VP
assigns nom.
ti VP
the sandwich, and the external θ- Case
V′ V′
role (Agent) to the DP Bill. V assigns
θ V θ DP acc. Case V+T i DP
eat ate
• Since it assigns an external θ-role, The sandwich
eat is also a Case-assigner. the gets Case the
sandwich from V sandwich

Passive again… TP
Passive again… TP
Finite T
assigns nom.
– The sandwich was eaten. DS The sandwich Case
T′ DP i T′
• Now, let’s look at the passive gets Case
sentence. from T
T VP the Vj+T VP SS
[past] sandwich was
• By moving the DP the
• The external θ-role was removed V′ sandwich to SpecTP we V′
from eaten and thus V can no satisfy both the Case Filter
V VP tj VP
longer assign Case. be and the EPP.
V′ • Simply satisfying the EPP by V′
• Unless the DP the sandwich moves inserting it into SpecTP
to a place where it can get Case, it V θ DP wouldn’t solve the problem of V ti
eaten eaten
will violate the Case Filter at SS. getting Case for the sandwich;
the hence the ungrammaticality of
sandwich *It was eaten the sandwich.

9
 Flavors of intransitives… Unaccusatives
• Let’s think for a moment about intransitive verbs. These
are verbs have a theta grid with a single θ-role to assign. • And it turns out that, yes, such verbs do exist.
Like walk, say. For example:
• Walk: Agent. • Fall: Theme.
• Now, think about the passive of a transitive verb; this is a
verb with only a single internal θ-role. • Fall is an “inherently passive” verb, an
unaccusative verb. It has only one θ-role to
• Eat: Agent Theme
assign, and that θ-role is an internal θ-role.
• Eaten: Theme
Because it has no external θ-role, by Burzio’s
• Taken together, it might occur to us to wonder whether
there might be intransitive verbs that inherently (like
Generalization, it also cannot assign
eaten) have only a single internal θ-role to assign… accusative Case.

Unaccusatives vs. unergatives Unaccusatives vs. unergatives

• There are many reasons to think that verbs like fall have
only an internal argument. • The point is really that we can distinguish
• First, the subject is really a Theme as far as thematic two types of single-argument (intransitive)
relations go, it is affected, not an agent. Themes are verbs in terms of their theta grid with
always objects. respect to whether they have an external θ-
• Another suggestive piece of evidence comes from role to assign or not. Their (highly
Romance languages like French, where passives and unintuitive) names, for the record, are:
verbs like fall act similarly, and differently from other
(truly agentive) intransitive verbs.
– Jean est tombé. ‘John fell.’ (past unaccusative) • Unaccusatives: Have one, internal θ-role.
– Le frômage a été mangé. ‘The cheese was eaten.’ (passive) • Unergatives: Have one, external θ-role.
– Jean a marché. ‘John walked.’ (past unergative)

Bill fell Revisiting VSO order in Irish

TP TP
SS • Recall these examples from last time (Irish):
DS – An bhfaca tú an madra?
T′ DP i T′
– Q See you the dog
T Bill tj VP – ‘Did you see the dog?’
VP
[past]
V′ Finite T can V′ – Duirt mé gur phóg Máire an lucharachán.
assign Case – Said I that kissed Mary the leprechaun
V θ DP V+Tj ti
– ‘I said that Mary kissed the leprechaun.’
fall fell
Unaccusative • VSO order was supposed to be derived by verb
Bill V cannot movement, but since an and gur are in C, it must
assign Case not be movement to C but rather to T.

10
 A VP-internal subject? A VP-internal subject?
• We ended up with a CP SS • First, since all DPs need Case, it CP SS
representation like this one, must be possible for the subject
C′ C′
where the subject was in SpecVP to get Case in SpecVP in Irish.
rather than in SpecTP. C TP • Second, since SpecTP is empty C TP

• That is, the subject appears to be T′ at SS, it must be that the EPP is T′
VP-internal in Irish. not active in Irish.
T+V i VP T+V i VP
• If this is right, there are a couple • We need to conclude that these
of things that must be true in DP V′ are dimensions along which DP V′

Irish under our current approach. ti … languages can vary. ti …

A VP-internal subject? A VP-internal subject?

• Parameter: EPP • How does the subject get Case in
CP SS SpecVP? CP SS
– On: SpecTP must be filled
(English) C′ C′
• Recall that we said before that a DP
– Off: no restriction on SpecTP TP has to be close to its Case-assigner. TP
C C
(Irish) – Being in SpecTP was close enough to T,
T′ being sister of V was close enough to V. T′
– (Note for later: we will want to • But this configuration also appears to
T+V i VP T+V i VP
revise this in light of future have the DP close to the Case-
DP V′ assigner. If we suppose this is close DP V′
developments, but for the moment
enough for Case assignment,
we are forced to this conclusion) everything is fine.
ti … ti …

Government Government
The radius of
• We will at some point want to • These three environments government
define more precisely what – Sister
counts as close enough for Case- – Specifier
assignment. Right now we have – Specifier of sister
three places which count as close XP XP
• …are together sometimes
enough (to the Case-assigning DP X′ called the positions which DP X′
head X) are governed by the head X.
– Sister X YP X YP
– (For now, we will not go into
– Specifier Y′ a more formal definition, but Y′
DP DP
– Specifier of sister we will look at this later.)
Y … Y …

11
 Government A VP-internal subject?
The radius of
• The idea is then that a Case- government
assigning head X can assign • Back to the question of the VP- CP DS
Case to a DP which is any of internal subject.
C′
these positions.
XP C TP
• Since the guiding intuition of our
• Case-assignment can only DP X′ approach has been that T′
take place between a Case- languages are fundamentally
X YP T VP
assigner and a DP within the alike, it is a bit jarring to think
radius of government. DP Y′ that English and Irish could DP V′

Y … differ in such a deep way as this. V …

A VP-internal subject? A VP-internal subject?

• However, there is some evidence to
support the idea that in English the CP – All the students will leave.
DS TP DS
subject originates in SpecVP too, – The students will all leave.
contrary to what we’ve been C′ T′
– *The students will leave all.
assuming—and moves to SpecTP.
C TP • The idea is that all the students is a T VP
• One of the least complex arguments
for this concerns the “floating unit at DS, which we can write as a will
T′ QP V′
quantifier” all. “QP” (Quantifier Phrase) headed
– All the students will leave. T VP by all. Q′ V
– The students will all leave. • Then, at this point, one of two leave
DP V′ Q DP
– *The students will leave all. things can happen—either the QP all the students
• Where can all be found? V … moves to SpecTP or the DP does.

A VP-internal subject? A VP-internal subject?

– All the students will leave. – All the students will leave.
TP SS1 TP SS2
– The students will all leave. – The students will all leave.
QP i T′ DP i T′
– *The students will leave all. – *The students will leave all. the students
• If the QP moves, we get the Q′ T VP • If the QP moves, we get the T VP
first sentence above. will first sentence above. will
Q DP ti V′ QP V′
all the students • If just the DP moves, we get the
V second sentence above. Q′ V
leave leave
• Yet neither option could
Q ti
produce the third sentence… all

12
 A VP-internal subject? A VP-internal subject?
– All the students will leave. • There are several other, more
TP SS2 complex (but perhaps even more
– The students will all leave.
DP i T′ convincing) arguments for the VP-
– *The students will leave all. the students internal Subject Hypothesis as well, TP
• Notice that this gives a T VP
reasonably natural way to will but let us take this as good enough DS
QP V′ evidence to adopt it. T′
explain where all can be, but it
is not available unless we Q′ V T VP
believe that the subject leave • VP-Internal Subject Hypothesis
Q ti DP V′
originates at DS somewhere The subject originates in the
all
below the position of will. V …
specifier of VP at DS.

θ-role assignment θ-role assignment

• If we suppose that the subject originates in SpecVP, • A predicate can only assign its θ-roles within the
then we can also strengthen our view of where θ-roles maximal projection of that predicate.
can be assigned. – A V can only assign its θ-roles within the VP.
• Earlier, we’d supposed that θ-roles can only be
assigned within the same clause. • Adopting this requires a (very) slight tweak in
• Now, we can in fact go further: what we consider to be an external θ-role. We
can no longer consider it to be a θ-role assigned
• A predicate can only assign its θ-roles within the external to the VP, since there are no longer any
maximal projection of that predicate. such θ-roles. Instead, we say that the external θ-
– A V can only assign its θ-roles within the VP.
role is the θ-role assigned to SpecVP.

Small clauses Small clauses

• Armed with the VP-internal subject • I find Bill intolerable.
hypothesis, we are also now in a position to • I consider Bill incompetent.
understand another type of sentence which • I want Bill off this ship. (Immediately!)
we have not thus far considered.

• These have a pretty similar meaning as

• I find Bill intolerable. sentences with to be inserted after Bill, but
• I consider Bill incompetent. yet there’s no to and no be… there’s no
• I want Bill off this ship. (Immediately!) evidence of a TP or a VP in Bill intolerable.

13
 Small clauses Small clauses
• A common way to look at these • Even in a small clause, all DPs
sentences is as containing small TP DS need to get Case. TP SS
clauses—a little proposition • In this sentence I gets
T′ DP i T′
headed not by a verb but by nominative Case from the finite I
another kind of predicate, like an T VP main clause T. Where does Bill T VP
adjective. [pres] get Case? tj
• Just like the subject of a regular DP V′ ti V′
clause, the subject of a small I
clause is in its specifier. V AP V+Tj AP
find find
• But unlike in a regular clause, it DP A′ DP A′
stays there, so we can see it in Bill Bill
the specifier of the predicate. A A
intolerable intolerable

Small clauses Small clauses

• Even in a small clause, all DPs • How do we know that?
need to get Case. TP SS TP SS
– Bill finds me intolerable.
• In this sentence I gets DP i T′ DP i T′
nominative Case from the finite I Bill
• Notice that the case of the
main clause T. Where does Bill T VP T VP
pronoun which is the subject of
get Case? tj tj
V′
the small clause is accusative—it V′
• Answer: The same place Bill ti ti
is the type of Case assigned by a
gets Case in I find Bill to be transitive verb (and not the type
V+Tj AP V+Tj AP
intolerable—from the transitive find of Case assigned by finite T). find
verb find, allowed because Bill is DP A′ – *Bill finds I intolerable. DP A′
in the its radius of government. Bill me
A A
intolerable intolerable

Genitive Case TP
Genitive Case TP
• Consider • In general, Case-assigners
DP T′ don’t get to assign two Cases, DP T′
– The president’s brother left.
• Every DP needs to get Case. DP D′ T so it can’t be T—plus, the DP D′ T
possessor DP is not in the
• The entire DP the president’s D′ D NP government radius of T. D′ D NP
brother gets Case like any ’s ’s
other DP—in this case it gets D NP N′ • This leaves us one choice… D NP N′
the the
nominative Case from the N′ N N′ N
finite T. brother brother
N N
• But where does the president president president
get its Case?

14
 Genitive Case TP
Let’s regroup
• In general, Case-assigners
don’t get to assign two Cases, DP T′ • Last time, we saw that we needed to differentiate
so it can’t be T—plus, the two different levels of structure (DS and SS) and
DP D′ T allow for movement of parts of the structure in
possessor DP is not in the order to get the word order facts of English and of
government radius of T. D′ D NP
’s other languages. X-bar theory alone wouldn’t
• This leaves us one choice… D NP N′ allow us to describe the facts.
• The case that possessors the • Last time, we saw examples of head-movement,
N′ N moving the head of an X-bar structure up the tree
receive is called genitive Case brother
to the next head up. For example, V-to-T, T-to-C,
and it is assigned by the N
and N-to-D movement.
possessive D ’s. president

Let’s regroup Let’s regroup

• We saw the role that Case plays, summarized here…
• This time, we saw that we also need to • Case Filter: All DPs must have Case at SS.
allow for movement of DPs as well. For • Finite T assigns nominative Case.
example,
• Transitive V assigns accusative Case.
• Raising: Billi is likely [ ti to win the race]. • P assigns oblique Case.
• Passive: [The sandwich]i was eaten ti . • A case-assigner can only assign Case to a DP within its
• Unaccusatives: Billi fell ti . “radius of government”:
– Its specifier
• Ordinary subjects: Billi will ti leave.
– Its sister
– The specifier of its sister.

Let’s regroup The Y model

• We have now explored a large part of the top section of
• We also concluded that the subject does not originate the “Y model” introduced to you a few weeks ago. Still
in SpecTP at DS, but rather in SpecVP and moves to to come are wh-movement and then our explorations of
SpecTP. This allowed us to say that: the “LF branch” and (question formation).
• A predicate can only assign its θ-roles within the
maximal projection of that predicate.
θ Theory
– A V can only assign its θ-roles within the VP. Overt movement, DS Subcategorization
Expletive insertion X-bar theory
• Finally, we looked at nonverbal predicates which also Case theory, EPP SS
seem to be able to head “small clauses”, as in I find Bill Covert movement
intolerable and which also have their subject in their Phonology/
specifier. Morphology PF LF
Binding theory

15