Extending X-bar Theory

DPs, TPs, and CPs

 The Puzzle of Determiners
 Specifier Rule XP(YP) X’
– requires the specifier to be phrasal
– *That the book (however cf. Those two books)
 Only example of a specifier we’ve seen.
The DP proposal
DP Abney 1987

D’

D NP

N’

N
 The DP hypothesis
 Explains why D isn’t a phrase (it is a head
of its own phrase!)
 Notice we now have NO examples of
specifiers!!
 Evidence???????
 ’s Genitives
 The man’s coat
 Not a suffix:
– [The man standing over there]’s coat
– [The dancer from New York]’s shoes
 ’s attaches to phrases.
 ’s Genitives
– The man’s coat ’s genitive
– The coat of the man free genitive
 ’s is in complementary distribution with
determiners:
– [The man standing over there]’s coat
– *The man standing over there’s the coat
 Complementary distribution means: two
items are examples of the same thing!
 ’s Genitives
 ’s is a determiner
NP
?

NP D N’
’s

the man standing over there N

hat

If ’s is a determiner, where does the possessor

go? (Remember the possessor modifies hat).
 ’s Genitives
 Problem solved by DP hypothesis
DP1

DP2 D’

the man standing over there D NP

’s
N
hat
notice this is in the specifier of DP1.
Is this the subject of the NP?
 Two other rules that don’t fit
X-bar theory
 SNP (T) VP
 S’  (C) S
 Problems:
– Category Specific
– No intermediate structure
– What are the heads, complements, adjuncts?
 The S Rule
SNP (T) VP

 What is the head?

– NP? not a head; it’s a phrase!
– VP? not a head; it’s a phrase!
– T? This is a head, but it’s optional!
 HMMM! Let’s think about headedness...
 Heads
 Give their category to the phrase
– [NPThe big linguistN from Calgary]
 Contribute other features to their phrase
• Linguist [+animate]
• [The linguist from Calgary] is pregnant [+animate]
• Fridge [-animate]
• [The fridge from Calgary] is pregnant
[-animate]
 The predicate “is pregnant” selects for an
animate subject.
 Heads of Clauses
 What are the relevant features of clauses?
– Tense/Finiteness!!
 Some examples
• I think [that Bill should leave]
• *I think [Bill to leave]
• ?I asked [that Bill leave]
• I asked [Bill to leave]
 The main verb is said to select for certain types of
embedded clause, based on finiteness.
 The head of clauses
 Tense provides the features selected for, so
perhaps T is the head of the sentence:
TP = S

DP T’
subject T VP
 HOLD ON!!!!
 We’ve only seen T in clauses with
auxiliaries!!
What about sentences without auxiliaries??
– John loves peanut butter sandwiches
 If T is optional, how can S=TP?
 Maybe T is obligatory in all sentences!
 T = Auxs, and suffixes
 Observation: auxiliaries and inflectional
suffixes on verbs are in complementary
distribution:
– I will dance
– I danced
– *I will danced
– I can dance
– *I can danced
 Proposal
 Inflectional tense & agreement suffixes are
also instances of T. T is obligatory in all
clauses
TP TP
DP T’ DP T’

John T VP John T VP
will -ed
V’ V’
V V
dance dance
 WAIT A MINUTE!
 The SUFFIX appears before the Verb?
HUH?
 Well the suffixes are in complementary
distribution with the auxiliaries…
 What is the difference between an
inflectional suffix and an aux?
– suffixes must be attached to something
– auxs are free (don’t have to be attached)
 suffixes as T
 Proposal: Inflectional suffixes are generated
under T, but they must be attached to a
verb, so they move by lowering and
attaching to the verb. TP
DP T’

John T VP
-ed
maybe, but it will get us V’
something later (chpt 8) V
dance
 Irregular verb morphology
 John runs (easy case)
 John ran ???? Inflectional suffix.
TP

DP T’

John T VP
Ø[past]
V’
run + Ø[past] = ran V
run
 TP
 T is obligatory, occupied by auxs or
inflectional suffixes (which lower and attach to the
verb.)
 The T head gives the finiteness properties to
the clause.
 TP = S
 The specifier of TP is occupied by the
subject of the clause
 the complement of TP is the VP
 S’(C) TP???
 What is the head of S’? C is the obvious
choice!
CP = S’

C’

C TP

What is the specifier of CP for? We’ll use it in

chapter 11 when we look at wh-movement. It is
where question words like “what” go.
 Is there a CP in every clause?
 We’ve claimed there is an TP in every
clause. Is there a CP in every clause?
 Embedded clauses without an overt
complementizer?
– I said [Louise loved rubber duckies]
 Main clauses
– Louise loved rubber duckies?
 Evidence from Yes/No
questions
– You have seen the rubber ducky.
– Have you seen the rubber ducky?
 Many languages don’t do this. Instead they
have special question complementizers:
– Ar fhag Seán
Q leave John
“Did John leave?”
 These are in complementary distribution with
complementizers
Evidence from Yes/No
questions
CP

C’

C[+Q] TP
Ar
fhag Seán
 Evidence from Yes/No
questions
CP CP
C’ C’
C[+Q] TP C[+Q] TP
Ø Have+Ø
DPsubj T’ DPsubj T’
you you
T VP tT VP
have
The null [+Q] C must be pronounced, so the T head moves to the
position to fill it.
 Evidence for [+Q] Cs in
English
 English has a [+Q] C found in embedded
clauses: (whether)
– I wonder whether Louise likes rubber duckies
 Subject/Aux inversion disallowed (in
complementary distribution) with whether:
• *I wonder whether has Louise owned a rubber ducky.
• I wonder whether Louise has owned a rubber ducky.
 This means that subject/aux inversion is a
diagnostic for the presence of C in English!
 Conclusion of discussion so
far
 Root questions in English contain a
phonologically null [+Q] complementizer.
 T raises to this [+Q] to give it phonological
content.
 Evidence that non-questions
have null C?
 Recall that conjunction only links together
items of the same category. If questions
have a null C (indicated by subject/aux
inversion), then anything they are conjoined
with must ALSO have a C.
– You can lead a horse to water but can you make
him drink?
 Second clause has a null C (indicated by
subject/aux; therefore, first clause must also
have a null C.
 CP

CP Conj CP
but
C’ C’
C TP C TP
Ø[-Q] Ø[+Q]
DP T’ DP T’
you you
T VP T VP
can can
lead a horse make him
to water drink
since there must be a CP in the second clause, for subject/aux
inversion, then there must ALSO be a CP in the first clause.
Therefore all clauses have a CP, even if the C head is null.
 Summary
 D isn’t a specifier -- it is a head. Evidence
from ’s genitives. DP hypothesis
 The head of the sentence is T. The sentence
type is determined by the finiteness of T
 S is replaced by TP
 The subject is the the spec of TP
 All sentences have TP, when T is suffixal it
lowers to the verb
 Summary
 S’ is replaced by CP
 All clauses have a C head. It may be null.
Evidence comes from subject/aux inversion
in yes/no questions.