Conference PaperPDF Available

Stripping Structures with Negation in Persian

Authors:

Abstract

In this paper, I introduce two novel Stripping constructions from Persian that occur with negation. I refer to these constructions as Polarity Stripping and Negative Stripping. I argue that they involve clausal coordination, and that in the second coordinate, the entire clause, except for a constituent, is elided under identity with corresponding parts of the first coordinate. This type of construction involves TP ellipsis, which is licensed by the Pol head that carries an [E] feature (Merchant 2001). I also study another structure, which I refer to as Pseudo-stripping. I argue that even though Pseudo-stripping looks like Polarity Stripping and Negative Stripping constructions, it does not involve ellipsis despite what has been claimed in the literature for English (Kolokonte 2008). I propose that Pseudo-stripping is mono-clausal and is derived via movement.
1
Stripping Structures with Negation in Persian
1
Vahideh Rasekhi
University of California, Los Angeles
Abstract: In this paper, I introduce two novel Stripping constructions from
Persian that occur with negation. I refer to these constructions as Polarity
Stripping and Negative Stripping. I argue that they involve clausal
coordination, and that in the second coordinate, the entire clause, except for
a constituent, is elided under identity with corresponding parts of the first
coordinate. This type of construction involves TP ellipsis, which is licensed
by the Pol head that carries an [E] feature (Merchant 2001). I also study
another structure, which I refer to as Pseudo-stripping. I argue that even
though Pseudo-stripping looks like Polarity Stripping and Negative
Stripping constructions, it does not involve ellipsis despite what has been
claimed in the literature for English (Kolokonte 2008). I propose that
Pseudo-stripping is mono-clausal and is derived via movement.
Key words: Ellipsis, Stripping, Negative Stripping, Pseudo-stripping,
Persian
1. Introduction
The goal of this paper is to provide a syntactic analysis for Stripping
structures with negation in Persian, which have not been studied before.
Stripping refers to an operation in which the entire clause except for one
constituent is elided under identity with corresponding parts of the
preceding clause (Hankamer and Sag 1976), as in (1).
(1) Alan likes to play volleyball, but not Sandy.
(Hankamer and Sag 1976: ex. 44)
1
Thanks to John Bailyn, Simin Karimi, Dan Finer, Thomas Graf, the
audience at NACIL 1, and the reviewers for their feedback! I am also
thankful to the Roshan Cultural Heritage Institute and especially its Chair
and President, Dr. Elahé Omidyar Mir-Djalali for making this research
possible.
2
In the second conjunct in (1), the entire clause is elided except for one
element Sandy, which is referred to as remnant, and the negative element
not
2
.
Let us consider Stripping constructions in Persian (2). In both (2a)
and (2b), the remnant Ayda, which precedes the negative marker na, carries
a high pitch accent.
3
In addition, it contrasts with its corresponding element
Araz in the preceding clause. The only difference between these structures,
on the surface, is the presence of the coordinator vali but in (2a). I refer to
the structure in (2a) as Polarity Stripping (PolS) and to the structure in (2b)
as Negative Stripping (NegS).
(2) a. ARAZ ketāb kharid, vali AYDA na (PolS)
Araz book bought.3SG but Ayda NEG
‘Araz bought books, but Ayda did not (buy books).’
b. ARAZ ketāb kharid, AYDA na (NegS)
Araz book bought.3SG Ayda NEG
‘Araz bought books, Ayda did not (buy books).’
Persian also allows structures such as (3), which I refer to as Pseudo-
Stripping (PseS). In this construction, unlike PolS and NegS, the negative
marker carries a high pitch accent.
Araz book bought.3SG NEG Ayda
2
The other element can also be an affirmative element or an adverb. In this
paper, we focus only on structures that include a negative marker. For more
discussion on this type of stripping, also see Rasekhi (2018) and Rasekhi
(2019c). For discussion on different types of stripping in Persian refer to
Rasekhi (2018), Rasekhi (2019a), and Rasekhi (2019b).
3
The remnant in these constructions can be a subject, an object, an adverb,
or an adjective. However, in this paper, for reasons of space, the presented
data only includes subject remnant.
3
On the surface, the difference between the structures in (2) and (3) is in the
order of Ayda and the negative marker na. However, I show that they are
different structures and cannot be accounted for in the same way. I argue
that PolS and NegS are true cases of Stripping, which involve ellipsis. On
the other hand, based on the evidence from Persian, I argue that Pseudo-
stripping does not involve ellipsis despite what has been claimed for English
(Kolokonte 2008).
Adopting Rizzi’s (1997) cartographic approach, I propose that the
remnant in PolS and NegS moves to the Spec of TopP and FocP,
respectively. In addition, I propose that the negative marker na originates
on the Spec of PolP, and the Pol head, which carries an [E] feature
(Merchant 2001), licenses the deletion of its complement, TP, at the PF
level.
On the other hand, I propose that PseS is derived via movement
rather than ellipsis. I propose that the negative marker in this structure is a
constituent negation, and that the NEG XP
4
constituent, in the underlying
structure, adjoins to XP’s corresponding element. However, since it carries
a contrastive focus feature, it has to move to a focus position. I propose that
it undergoes rightward movement and adjoins to FocP in the TP level.
Structure of this paper is as follows: In section 2, I discuss the
differences among PolS, NegS, and PseS in terms of their context of
occurrence and interpretation. In section 3, I provide evidence that the XP
in PolS is a topicalized element while the XP in NegS and PseS is a
focalized element. In section 4, I discuss the nature of the negative marker
in these constructions. In section 5, I provide syntactic analyses for the
constructions under discussion. Section 6 provides a conclusion.
4
XP refers to any element (e.g. DP, PP, AdvP, AdjP) that can follow the
negative marker.
4
2. Context of Occurrence and Interpretation
In this section, I show that PolS, NegS, and PseS constructions cannot occur
in the same contexts since they have different interpretations. To illustrate
their differences, suppose that two friends A and B are talking to each other
about the shopping they did with their friends. Without any further
discourse, speaker A can utter the sentence in (4a), which is an instance of
PolS. However, as the ungrammaticality of the sentences in (4b) and (4c)
shows, NegS and PseS are not possible with a contextual antecedent.
(4) Context: [A and B are talking about the shopping they did with
their friends]
a.
ARAZ dirooz ketāb kharid, vali AYDA na (PolS)
Araz yesterday book bought.3SG but Ayda NEG
‘Araz bought books yesterday, but Ayda did not (buy books).’
b.
#ARAZ dirooz ketāb kharid, AYDA na (NegS)
Araz yesterday book bought.3SG Ayda NEG
c.
#ARAZ dirooz ketāb kharid, NA Ayda (PseS)
Araz yesterday book bought.3SG NEG Ayda
The grammaticality of (4a) shows that PolS can occur in an out of the blue
context, and that it does not require a linguistic antecedent. Therefore, we
can say that in this structure, the speaker is introducing new information to
the discourse by saying that Araz bought books but Ayda did not buy books.
On the other hand, the ungrammaticality of (4b) and (4c) suggests that NegS
and PseS cannot occur in an out of the blue context, which is due to the fact
that these constructions have an obligatory corrective interpretation. We
expect these constructions to be acceptable with a linguistic antecedent.
This is borne out as shown in (5) and (6).
5
In (5a), speaker A makes an assertion that Araz and Ayda bought books.
However, speaker B corrects speaker A by uttering the sentence in (5b),
which is an instance of NegS. The fact that this structure is acceptable with
a linguistic antecedent, as shown in (5b), but not with a contextual
antecedent, as in (4b), shows that it can only occur in contexts in which a
proposition has been made.
(5) a. dirooz Araz va Ayda ketāb kharid-an
yesterday Araz and Ayda book bought-3PL
‘Yesterday, Araz and Ayda bought books.’
b. ARAZ dirooz ketāb kharid, AYDA na (NegS)
Araz yesterday book bought.3SG Ayda NEG
‘(No, you are wrong), ARAZ bought books yesterday, AYDA
did not (buy books).
In (5b), the emphasis is on the predicate and whether Ayda and Araz bought
books. We know that we have two alternatives in this regard; Ayda and Araz
have either bought books or they have not bought books. Therefore, we can
say that in NegS, as in (5b), speaker B corrects speaker As statement by
showing which alternative is true and which one is false. In this context,
Araz bought books is true while Ayda bought books is false.
PseS, similar to NegS, is acceptable when it is provided with a
linguistic antecedent (6b).
(6) a. dirooz Ayda ketāb kharid
yesterday Ayda book bought.3SG
‘Yesterday, Ayda bought books.’
b. ARAZ ketāb kharid, NA Ayda (PseS)
Araz book bought.3SG NEG Ayda
‘(No, you are wrong), ARAZ bought a book, NOT Ayda.’
6
In (6a), speaker A makes an assertion that Ayda bought books. However,
speaker B corrects speaker A’s assertion by saying (no, you are wrong) it
was Araz who bought a book, not Ayda. In (6b), the emphasis is on the
person who bought books; whether it was Araz or Ayda who bought books.
We can say that in this context, we have two alternatives with regard to who
bought books; either Araz bought books or Ayda bought books. We see that
speaker B in (6b) corrects speaker A by replacing Ayda in (6a) with Araz.
Thus, we can say that in Pseudo-stripping, we have correction by
substitution.
Summary
A summary of the characteristics of PolS, NegS, and PseS is presented in
(7).
(7) Context of Occurrence and Interpretation PolS NegS PseS
a. Acceptable in an out of the blue context Yes No No
b. Has a corrective interpretation No Yes Yes
c. Has correction by showing which alternative NA Yes No
is true and which one is false
d. Has correction by substituting one alternative NA No Yes
by another
3. Information Structure
In this section, I discuss how PolS, NegS, and PseS structures are
constrained by information structure. I argue that the XP in PolS has the
characteristics of a topic while the XP in NegS and PseS has the
characteristics of focus. I provide two pieces of evidence for this claim.
One of the differences between topic and focus is that we can have
more than one topic in a sentence while we can only have one focus per
7
sentence.
5
With this in mind, let us have a look at the structures in ((8)-(10)).
(9) *AYDA ketāb-a-ro kharid, ARAZ (NegS)
Ayda book-DEF-ACC bought.3SG Araz
majalle-ha-ro na
magazine- DEF-ACC NEG
In PolS (8), we have two elements before the negative marker; the subject
Araz and the direct object magazine, and the sentence is grammatical.
However, as the ungrammaticality of the sentences in (9) and (10) shows, it
is not possible to have two elements in NegS and PseS. These examples
show that PolS is compatible with topicalized elements while NegS and
PseS are compatible with focalized elements.
The second piece of evidence that shows NegS and PseS are
compatible with focalized elements while PolS is not comes from their
compatibility with focus adverbs such as only. Let us consider the examples
5
Persian allows two elements bearing a contrastive focus feature in the
same sentence only if one of them has an inherent focus feature (Karimi
2005: 133).
(i) KIMEA māh-e gozashte faqat se- film did-e
Kimea month-EZ previous only three-part film saw-3SG
‘It was Kimea who saw only three movies last month.’ (Everyone else
has seen more movies)
(8) AYDA ketāb-a-ro kharid, vali ARAZ (PolS)
Ayda book-DEF-ACC bought.3SG but Araz
majalle-ha-ro na
magazine-DEF-ACC NEG
‘Ayda bought the book but Ayda didn’t (buy) the magazine.’
(10) *AYDA ketāb-a-ro kharid, NA Araz (PseS)
Ayda book-DEF-ACC bought.3SG NEG Araz
majalle-ha-ro
magazine- DEF-ACC
8
in (11).
A: ki- ketāb kharid-an?
who-PL book bought-3PL
‘Who bought books?’
B: *hame ketāb kharid-an, vali FAQAT Ayda na (PolS)
everyone book bought-3PL but only Ayda NEG
Intended: ‘Everyone bought books, but only Ayda did not.’
(Adapted from López and Winkler 2000)
The ungrammaticality of the sentence in (11b) shows that the DP Ayda in
PolS is not compatible with the focus adverb faqat ‘only’. In the same
context, the conversation can continue as in (12). In (12a), the speaker
makes an assertion that only Araz bought books. We see that NegS (12b)
and PseS (12b') structures are acceptable in this context, which shows that
the DP Araz in these structures is compatible with the focus adverb ‘only’.
Now that we have established we have a topicalized element in PolS but a
focalized element in NegS and PseS structures, we need to discuss the
(11) [Context: speaker A and speaker B are talking about their friends
who bought books]
(12) a. pas faqat Araz ketāb kharid
so only Araz book bought.3SG
‘So, only Araz bought books.’
b. HAME ketāb kharid-an, FAQAT Araz na (NegS)
everyone book bought.3PL only Araz NEG
‘(No, you are wrong) everyone bought books and not just Araz.’
b'. HAME ketāb kharid-an, NA FAQAT Araz (PseS)
everyone book bought.3PL NEG only Araz
‘(No, you are wrong), everyone bought books, not just Araz.’
(Adapted from López and Winkler 2000)
9
nature of the negative marker in these structures.
4. The Nature of Negative Marker
In PolS (13a) and NegS (13b), when the sentence is continued, we have two
negative markers in the second clause: one is after the remnant Araz and the
other is affixed to the verb. However, in PseS (13c), we have only one
negative marker that precedes Araz and it is not possible to continue the
sentence after Araz.
b. AYDA ketāb kharid, ARAZ na, ketāb na-kharid
Ayda book bought.3SG Araz NEG book NEG-bought.3SG
‘Ayda bought a book, Araz did not buy a book.’
c. AYDA ketāb kharid, NA Araz (*ketāb na-kharid)
Ayda book bought.3SG NEG Araz book NEG-bought.3SG
‘Ayda bought a book, not Araz.’
The data in (13) indicates that: i) PolS and NegS are bi-clausal while PseS
is mon-clausal, ii) in PolS and NegS, we have two negative markers in the
underlying structure while in PseS, we have only one negative marker. The
question that now needs to be addressed is: What is the nature of the
negation in these constructions?
In languages like English, in which the sentential negation ‘not’ and
constituent negation ‘no’ are different, only the sentential negation is
possible in Stripping, as shown in (14a), and phrasal negation is not possible
(14b).
(14) a. John bought a book, not Mary.
(13) a. AYDA ketāb kharid, vali ARAZ na, ketāb na-kharid
Ayda book bought.3SG but Araz NEG book NEG-bought.3SG
‘Ayda bought a book, but Araz did not buy a book.’
10
b. *John bought a book, no Mary.
However, in Persian, the sentential negation (15) and phrasal negation (16)
are homophonous. Thus, it is not clear whether the negative marker in the
constructions in (13) is sentential or constituent negation.
A: na, na-did-am
NEG NEG-saw-1SG
‘No, I did not see her.’
(16) a. na man chini sohbat mikon-am, na u
NEG I Chinese harf do-1SG NEG they
‘Neither I nor they speak Chinese.’
b. man ketāb mikhoon-am, na majale
I book read-1SG NEG magazine
‘I am reading a book, not a magazine.’ (Kwak 2010: 624)
I propose that the negative marker in PseS is constituent negation. This
claim is based on Klimas (1964) either-and neither conjoining test and
adverbs, which I discuss in the next two subsections. Then, I propose that
the negative marker in PolS and NegS functions as a focusing adverb.
4.1. Either-and Neither Conjoining Test
In this section, I use Klima’s (1964) either-and neither-conjoining test to
determine the nature of negative marker in PolS, NegS, and PseS. As shown
in (17a) and (18a), the structures with sentential negation are grammatical
with either-and neither-conjunction. However, the structures with
constituent negation are not compatible with this type of conjunction, as
shown in (17b) and (18b).
(15) Q: Maryam-ro did-i?
Maryam-ACC saw-2SG
‘Did you see Maryam?’
11
(17) a. Sentential negation: Mary isn’t a happy person and John isn’t
either.
b. Constituent negation:*Mary is a not happy person and John isn’t
either.
(18) a. Sentential negation: Mary isn’t a happy person and neither is John.
b. Constituent negation: *Mary is a not happy person and neither is
John.
Let us apply this test to our structures under discussion to determine whether
the negative marker na in these structures is sentential negation or
constituent negation. As the examples below show, we see that PolS (19)
and NegS (20) constructions are acceptable with either-conjoining while
PseS (21) is not. This shows that the negative marker in PseS is constituent
negation.
(19) AYDA khoshhāl-e , vali ARAZ na, MARYAM ham na (PolS)
Ayda happy-3SG but Araz NEG Maryam also NEG
‘Ayda is happy but Araz is not (happy), Maryam is neither.’
(20) AYDA khoshhāl-e, ARAZ na, MARYAM ham na (NegS)
Ayda happy-3SG Araz NEG Maryam also NEG
‘Ayda is happy, Araz is not (happy), Maryam is neither.’
(21) *AYDA khoshhāl-e, NA Araz, NA ham Maryam (PseS)
Ayda happy-3SG NEG Araz NEG also Maryam
4.2. Adverbs
The second piece of evidence that shows the negative marker in PseS but
not in PolS and NegS is constituent negation comes from examples as in
((22)-(24)).
(22) Ayda hamishe ketāb mikhar-e, vali Araz hamishe na (PolS)
Ayda always book buy-3SG but Araz always NEG
Ayda always buys books, but Araz does not always (buy books).
12
(23) Ayda hamishe ketāb mikhar-e, Araz hamishe na (NegS)
Ayda always book buy-3SG Araz always NEG
‘Ayda always buys books, Araz does not always (buy books).’
(24) *Ayda hamishe ketāb mikhar-e, na hamishe Araz (PseS)
Ayda always book buy-3SG NEG always Araz
As illustrated in these examples, in PolS (22) and NegS (23), an adverb can
occur between Araz and the negative marker. However, this is not possible
in PseS (24), which shows that the negative marker in this structure is
constituent negation. Now that we know the negative marker in PseS
involves constituent negation, in the next section, I show that the negative
marker in PolS and NegS functions as a focusing adverb.
4.3. Focusing Adverb
In this section, I propose that the negative marker in PolS and NegS but not
in PseS belongs to a class of adverbs such as never called focusing adverb
(Rooth 1985, 1996). This proposal is supported by examples such as ((25)-
(27)).
(25) AYDA hamishe ketāb mi-khar-e, vali ARAZ hichvaqt
Ayda always book DUR-buy-3SG but Araz never
(ketāb ne-mi-khar-e)
book NEG-DUR-buy-3SG
‘Ayda always buys books, but Araz never (buys books).’
(26) AYDA hamishe ketāb mi-khar-e, ARAZ hichvaqt
Ayda always book DUR-buy-3SG Araz never
(ketāb ne-mi-khar-e)
book NEG-DUR-buy-3SG
‘AYDA always buys books, ARAZ never (buys books).’
(27) *AYDA hamishe ketāb mi-khar-e, HICHVAQT Araz
Ayda always book DUR-buy-3SG never Araz
In PolS (25) and NegS (26), it is possible to replace the negative marker
13
with the adverb never, while in PseS (27), this is not possible. These
examples show that the negative marker in PolS and NegS functions as a
focusing adverb.
5. Analysis
5.1. Analysis of Polarity Stripping and Negative Stripping
As discussed in section 3, we know that the remnants in PolS and NegS
function as topicalized and focalized elements, respectively. I propose that
the remnant in these structures carries contrastive topic and contrastive
focus features; therefore, it moves to the Spec of TopP and to the Spec of
FocP, respectively, as schematically illustrated in (28).
Regarding the negative marker, as discussed in section 3.5. I propose that
the negative marker in these structures is a focusing adverb. In Persian, the
position of sentential negation is assumed to be in the CP level, as the phrase
structure in (29) illustrates.
(29) [CP [TopP [FocP [NegP [TP [vP PredP]]]]]] (Karimi 2005: 147)
(28) a. Structure of Polarity Stripping b. Structure of Negative Stripping
14
Sentential negation, NegP, selects TP as its complement. However,
negation is morphologically realized on the verb through Agree relation
between Neg, which bears an interpretable negation feature, and v, which
bears an uninterpretable negation feature (Taleghani 2008).
Since the negative marker in PolS and NegS precedes the sentential
negation, which is suffixed on the verb, it has to be in a position higher than
NegP. In addition, since the remnant in these constructions moves to the
Spec of TopP and FocP, respectively, we know that the negative marker is
higher than NegP but lower than FocP. Let us call this position PolP. If our
analysis is on the right track, the phrase structure in (29) should be revised
to include PolP.
(30) [CP [TopP [FocP [PolP [NegP [TP [vP PredP]]]]]]]
I propose that the negative marker na in PolS and NegS, which acts as a
focusing adverb, originates in the Spec of PolP. I also propose that the Pol
head carries an [E] feature (Merchant 2001) that licenses the deletion of its
complement, TP, which is identical to the TP of the antecedent clause and
therefore it becomes redundant, at the PF level.
The structure which I propose for PolS is given in (31). In this
structure, the remnant Ayda in the second clause moves to the Spec of TopP
and the [E] feature on the Pol head licenses the deletion of TP at the PF
level.
15
Similarly, the structure of NegS is schematically represented in (32). Even
though there is not an overt coordinator in this structure, I assume that we
have a null conjunction in this structure. As shown in (32), the remnant Ayda
in the second clause moves to the Spec of FocP, before TP deletion takes
place.
(31) Structure of Polarity Stripping
(32) Structure of Negative Stripping
16
5.2. Analysis of Pseudo-stripping
We already know that PseS is mono-clausal. In addition, as discussed in
sections 3 and 4, we know that the XP in this structure functions as a
focalized element, and the negative marker is constituent negation.
In this section, I propose that PseS in Persian is derived via
movement. This is different from what has been proposed for English;
Kolokonte (2008) has proposed that PseS involves TP ellipsis.
I propose that the underlying structure of PseS is as shown in (33).
The na Araz constituent originates adjacent to its corresponding element
Ayda. However, since this word order is not possible in Persian, the
constituent na Araz undergoes obligatory rightward movement, as shown in
(34).
(33) *AYDA, NA Araz, ketāb kharid (PseS)
Ayda NEG Araz book bought.3SG
Lit: Ayda, not Araz, bought books.
(34) AYDA ti ketāb kharid, [NA Araz]i
Ayda book bought.3SG NEG Araz
A piece of evidence for the movement comes from case marking, as shown
in (35). The DP majale magazineand its corresponding element ketāb
book have -ro marking. If na majale were base-generated, we would not
expect majale to have -ro marking.
(35) KETAB-RO khoond-am, NA majala-ro
book-ACC read-1SG NEG magazine-ACC
Lit: ‘The book I read, not the magazine.’
Even though the structure in (33) does not sound natural in Persian, the
proposal that the NEG XP originates adjacent to the XP’s corresponding
element and arrives at its surface position via movement is supported by the
data from German (36) and Spanish (37). As illustrated in the following
17
examples, the NEG XP can appear at the sentence-final position or in the
middle of the sentence.
(36) German
a. Anna spielt Klavier, nicht Maria
Anna plays piano not Maria
‘Anna plays piano but not Maria.’
b. Anna, nicht Maria, spielt Klavier
Anna not Maria plays piano
‘Anna, not Maria, plays piano.’ (Graf Thomas p.c.)
(37) Spanish
a. Anna taco el piano, no María
Anna plays the piano not Maria
‘Anna plays piano, not Maria.’
The questions that need to be addressed are: How does the movement in
(34) work? Where does the NEG XP move to? I propose that the na Ayda
constituent undergoes right movement
6
and adjoins to the FocP in the TP
level
7
, as schematically illustrated in (38). The DP moves to FocP to satisfy
the uninterruptable strong focus feature [uConf*] that is on the Focus head.
6
To derive PseS via left-ward movement, one has to propose that the direct
object book and the verb bought in (i) move out of the vP. This
movement out of vP is not well-motivated.
(i) [FocP ARAZi [TP ketābj kharidk NA Ayda [vP ti tj tk]]]
7
See Kahnemuyipour (2001) and Rasekhi (2018) for the evidence on the
existence of FocP in the TP level.
b. Anna, no María, taco el piano
Anna not Maria plays the piano
‘Anna, not Maria, plays piano.’ (Elías-Ulloa José p.c.)
18
(38) Structure of Pseudo-stripping
The na Ayda constituent adjoins to the DP Araz. However, since the DP
Araz carries a contrastive focus feature, it moves to the right and adjoins to
FocP.
Kolokonte (2008) proposed that Pseudo-stripping in English
involves clausal coordination. Based on her analysis, the structure of the
relevant parts of the English example in (39) would be as illustrated in (40).
(39) John bought books, not Mary.
19
(40) Structure of Pseudo-stripping in English (Kolokonte 2008)
In (40), the DP Mary moves to the Spec of F1P in the left periphery, before
TP deletion takes place. In addition, the negative marker originates in the
Spec of PolP. There are three main issues with Kolokonte’s analysis. First,
her proposal of two focus positions in the CP domain is not motivated.
8
Second, we know that information focus is not subject to overt movement
to FocP (Kiss 1998).
9
Third, Kolokonte proposes that Pseudo-stripping
8
In her approach, F1P hosts an element with a contrastive focus feature
while F2P hosts an element with an information focus feature.
9
She acknowledges that it is not possible for information focus to be
preposed, as illustrated in (i).
(i) Q: What did you order?
A1: I ordered pizza.
A2: *PIZZA I ordered. (Kolokonte 2008: 127)
However, following Baltazani’s (1999) analysis of focus constructions in
Greek, Kolokonte proposes that first the focused element moves to the left,
and then the TP remnant moves to the left. Based on this proposal, the
sentence in (iA1) is derived as illustrated in (ii).
(ii) Step 1: Focus movement Pizzai [I ordered ti]
Step 2: TP-remnant movement [I ordered ti ]j pizzai tj
20
involves a clausal coordination; however, since it is not possible to continue
the sentence after Mary, as shown in (41), the TP undergoes an obligatory
deletion.
(41) *John bought books, not Mary bought books.
If Pseudo-stripping involves clausal coordination, we would expect to be
able to coordinate it with another clause. However, this is not possible as
the ungrammaticality of the sentence in (42) shows.
(42) *John bought books, not Mary, neither Jack.
6. Conclusion
In this paper, I studied two Stripping constructions with negation: Polarity
Stripping and Negative Stripping. I showed that these constructions involve
clausal coordination. Prior to ellipsis, a constituent that contrasts with its
corresponding element in the preceding clause moves out of TP, which is
specified for deletion. The remnant in Polarity Stripping and Negative
Stripping moves to the Spec of TopP and FocP, respectively. I proposed that
the negative marker in these constructions, which functions as a focusing
adverb, originates in Pol head. In the second coordinate, the Pol head, which
carries an [E] feature licenses the deletion of its complement, TP, at the PF
level.
I also studied another structure, Pseudo-stripping, that looks like
Polarity Stripping and Negative Stripping. Based on the evidence from
Step 2 renders the same word order as in (iA1). The focus movement is
obscured by the subsequent movement of the remnant TP to a position
higher than FocP, presumably TopP. Kolokonte bases her proposal on
Greek data but she does not provide any English data. It is not clear whether
focus fronting in Greek can be extended to English.
21
Persian, I argued that this structure does not involve ellipsis despite what
has been claimed for English (Kolokonte 2008). I proposed that Pseudo-
stripping is derived via right-ward movement. I provided evidence that the
negative marker in this structure is constituent negation and proposed that
the NEG XP constituent originates adjacent to XP’s corresponding element.
Since it carries a contrastive focus feature, it moves to the right and adjoins
to FocP, above vP, in the TP level.
The implications of the analyses put forth above are as follows: a)
the ellipsis site and the antecedent clause must have parallel information
structure status, that is, both the remnant and its correlate must be either
focalized or topicalized elements, b) structures with XP NEG word order
are bi-clausal while the structures with NEG XP word order are mono-
clausal, c) in structures with XP NEG word order, the remnant can be either
a focalized or topicalized element while in structures with NEG XP word
order, the remnant must be a focalized elements.
References
Hankamer, J. and Sag, I. (1976). Deep and surface anaphora. Linguistic
Inquiry 7: 391-426.
Kahnemuyipour, Arsalan. 2001. On wh-questions in Persian. Canadian
Journal of Linguistics 46: 41-61.
Karimi, Simin. 2005. A minimalist approach to scrambling: Evidence from
Persian. Berlin: Mouton de Gruyter.
Klima, Edward. 1964. Negation in English. The structure of language:
readings in the Philosophy of language, 246-323, ed. by Fodor Jerry
and Katz Jerrold, Englewood Cliffs (New Jersey): Prentice Hall.
Kolokonte, Marina. 2008. Bare argument ellipsis & information structure.
Ph.D. dissertation, Newcastle University, United Kingdom.
López, Luis. and Winkler, Susanne. 2000. Focus and Topic in VP-
Anaphora constructions. Linguistics 38:623-664.
Kwak, Saera. 2010. Negation in Persian. Iranian Studies 43:621-636.
Merchant, Jason. 2001. The syntax of silence: Sluicing, islands, and the
theory of ellipsis. Oxford: Oxford University Press.
Rasekhi, Vahideh. 2018. Ellipsis and information structure: Evidence from
22
Persian, Ph.D. dissertation. Stony Brook University.
Rasekhi, Vahideh. 2019a. Two types of why in syntax: Evidence from Why-
stripping in Persian. Poster presented at Sluicing and Ellipsis at 50.
Chicago, Illinois: University of Chicago.
Rasekhi, Vahideh. 2019b. On the syntax of Why-stripping in Persian. Paper
presented at the 2nd North American Conference in Iranian
Linguistics (NACIL2). Tucson, Arizona: University of Arizona.
Rasekhi, Vahideh. 2019c. Negative Stripping and Pseudo-stripping in
Persian. Paper presented at the 52nd Meeting of the Chicago
Linguistic Society. Chicago, Illinois: University of Chicago.
Rizzi, Luigi. 1997. On the fine structure of the left periphery. Elements of
grammar, international handbook of linguistics, 281-337, ed. by
Liliane. Haegeman, Kluwer: Dordrecht.
Rooth, Mats. 1985. Association with Focus, Ph.D. dissertation. University
of Massachusetts, Amherst.
Rooth, Mats. 1996. Focus. The handbook of contemporary semantic theory,
271-297, ed. by Shalom Lappin. Oxford: Blackwell Publishing.
Taleghani, Azita. 2008. The modality, aspect, and negation in Persian.
Philadelphia: John Benjamins.
... 1 An example of stripping is given in (i). For more details on the varieties of stripping in Persian, see Rasekhi (2020). ...
Article
This paper empirically tests the embedding constraints on gapping in Persian. It has been suggested that gapping differs from other kinds of ellipsis in banning embedding. However, the first counter-examples in the literature come from Persian. Following up on previous experiments on embedded gapping in several languages, we report the results of two acceptability judgment tasks. Our results show that, while embedded gapping is overall acceptable in Persian, speakers’ acceptability judgements also vary depending on the semantic type of the embedding predicate, as well as the presence/absence of the complementizer. Data from Persian highlight that, despite the cross-linguistic variation observed with respect to the acceptability of embedded gapping, a general semantic constraint is at work across languages: non-factive verbs embed more easily than factive ones; inside factive verbs, semi-factive (cognitive) predicates embed more easily than true factive (emotive) ones. Moreover, whereas previous theoretical literature indicates no systematic preference for the absence or the presence of the complementizer in Persian, these new experimental data suggest a preference for complementizer drop. To account for the gradience observed in our experimental data, we propose an approach of gapping based on acceptability rather than grammaticality.
... Refer toRasekhi (2018Rasekhi ( , 2019 for the discussion on the syntax of stripping constructions in Persian. ...
Thesis
Full-text available
In this dissertation, adopting Rizzi’s (1997) Split-CP hypothesis, I provide a unified account for various elliptical structures in Persian on the basis of the interaction between information structure and ellipsis licensing feature bundles. I argue that ellipsis is constrained by information structure, in particular topic and focus. Only the elements that carry contrastive topic or contrastive focus features can survive ellipsis while the other elements are elided under identity with their corresponding elements in the antecedent clause. I discuss Merchant’s (2001) theory of ellipsis, in which the elements that survive ellipsis have a focus feature, and show that his approach cannot account for the Persian data. I argue that it is contrast that saves elements from being elided. To account for elliptical constructions in Persian, I propose the Contrast Condition on Ellipsis, which has three conditions: i) Ellipsis Condition: ellipsis is possible only in contexts in which an element in a sentence involving ellipsis contrasts with its corresponding element in the antecedent clause, ii) Identity Condition on the Remnant: the element that survives ellipsis must be in a contrastive relationship with its correlate and they must have identical information structure, and iii) Identity Condition on the Elided Elements: an element can be elided only if it is identical to its corresponding element in the antecedent clause. Regarding licensing ellipsis, I propose that ellipsis in Persian is licensed by the heads that have a focus feature, i.e. Foc(us) head, Pol(arity) head and Int(errogative) head. The proposed licensing heads are the result of the restrictions on the feature bundles that are possible in Persian. The [E] feature (Merchant 2001), which licenses the deletion of its complement at the PF level, can bundle with i) a strong uninterpretable contrastive focus [ConF] feature, [E, uConF*], ii) a strong uninterpretable [wh] feature, [E, uwh*], and iii) an uninterpretable [Pol] feature, [E, uPol].The [uConF*] and [uwh*] features trigger the movement of the elements with interpretable [wh] and [ConF] features out of the phrase that is specified for deletion. On the other hand, the [uPol] feature is satisfied by having a polarity marker in the Spec of PolP. The polarity marker can be affirmative or negative, depending on the polarity of the preceding clause. Based on the proposed theory, I study various elliptical structures in Persian. In chapters 2 and 3, I provide an in depth analysis of Verb-stranding and Stripping constructions with negation, respectively. In chapter 4, I provide an overview of Why-stripping, What-stripping, Cherā-stripping, Gapping, Sluicing, Fragment Answers, and Noun Phrase Ellipsis.
Article
This article explores wh-questions in Persian with the main purpose of determining whether Persian has syntactic wh-movement. This question, though superficially trivial, proves to be rather intriguing for Persian, due to the extensive use of scrambling in this language. 1 In this article, I will attempt to address this question while at the same time abstracting away from all the different word order possibilities introduced by scrambling. One solution is to find a method of type-determination for wh-questions that does not rely on the surface word order. For example, Cheng (1991, 1997) proposes that the typological distinctions among languages with respect to the formation of wh-questions can be attributed to the availability of question particles as well as to the properties of wh-words. In this article, I briefly look at Cheng (1991, 1997) and discuss some problems that arise if one attempts to apply her typology to Persian wh-questions. I then look at a relatively wide range of data and propose an analysis of Persian wh-questions. According to this proposal, Persian is neither a language with syntactic wh-movement (i.e., movement of a wh-phrase to [Spec, CP]) nor a wh-in-situ language; rather it should be classified with languages in which wh-phrases have been argued to undergo focus movement. I argue that in Persian this focus position, which is also the locus of contrastively focussed elements, is directly above vP. I will explore this proposal in more detail and will evaluate it against similar proposals made by other linguists for other languages. In particular, Bošković (1997a, 1997b, 1998a, 1998b, 2000a, 2000b) studies wh-questions in Serbo-Croatian and proposes a similar kind of focus movement (what he calls non-wh-movement) for the wh-elements. He attributes certain properties to this kind of movement. I will discuss these properties in some detail and determine whether Persian wh-questions exhibit the same properties.
Article
This paper explores the information structure of VP-anaphora constructions in English, Spanish, and German. The goal is to provide an initial account of the syntactic and prosodic licensing mechanisms of VP-anaphora and to suggest some theoretical consequences for the syntactic representation of the diVerent types of focus and topic. Recent studies (e.g. Rooth 1992b; Tomioka 1995; Fox 1998) have proposed that the characteristic feature of VP-anaphora constructions is that they bear contrastive focus on the remain- ing subject. We argue against this claim by providing evidence from prosodic and syntactic investigations of the remaining elements. Our hypothesis is that presentational focus in VP-anaphora is realized on the negative/aYr- mative expression and that the remaining constituent is free to take various discourse functions depending on the context and the type of construction in which the VP-anaphora occurs. Our main empirical argument against the above claim comes from Spanish, where the remaining constituent can never be interpreted as a contrastive focus.
Article
This paper classifies Persian negative elements and analyzes the negative structures based on Minimalism and Distributed Morphology. Negative Polarity Items (NPIs) in Persian which have [−neg] check its feature against [+neg] of negative marker for convergence. Negative Concord Items (NCIs), which are inherently negative, have [+neg]. Thus, in the NC Sentences, we have two [+neg]—one in negative marker, and the other in NCI. It can be misread that a NC structure, like Double Negation, delivers an affirmative reading. To prevent the misreading, we assumed Neg-absorption and Neg-criterion. However, that is impossible in Double Negation with two negative markers placed in different NegP, so the [+neg] in the sentence remain till Logical Form (LF) and they cancel each other and, as a result, the sentence becomes affirmative. Based on Minimalism and Distributed Morphology, the unified account on the structure of negation is examined.
Elements of grammar, international handbook of linguistics
  • Luigi Rizzi
Rizzi, Luigi. 1997. On the fine structure of the left periphery. Elements of grammar, international handbook of linguistics, 281-337, ed. by Liliane. Haegeman, Kluwer: Dordrecht.