This squib argues against Merchant’s (2008) analysis of voice mismatches in VP-deletion and pseudogapping by pointing out that it makes a false prediction when extended to sentences in which more than one VP-deletion operation applies.

1 VP-Deletion and Pseudogapping

Merchant (2008) argues that while the target of pseudogapping is VoiceP, whose head encodes the active/passive voices, VP-deletion targets VP, the complement of the Voice head.1 The targets of (i) VP-deletion and (ii) pseudogapping are schematically represented as follows:

(1)

  • a.

    graphic

  • b.

    graphic

Subj is occupied either by an external argument (YPEA) in active sentences or by an internal argument (ZPIA) in passive sentences. XPFoc in (1b) is the position for pseudogapped remnants (see Jayaseelan 1990, Lasnik 2001, Takahashi 2004).

Given that deletion is subject to morphosyntactic identity conditions as well as semantic ones (see, e.g., Sag 1976, Fiengo and May 1994, Johnson 2001, 2004, Merchant 2001), the analysis depicted in (1) predicts an asymmetry between VP-deletion and pseudogapping with respect to voice mismatches.

It is widely known that VP-deletion allows voice mismatches, that is, mismatches between the voice of the antecedent and that of the deletion site, as shown in (2).2 The deletion sites are in angle brackets.

(2)

  • a.

    Passive antecedent, active ellipsis

    The system can be used by anyone who wants to 〈use it〉.

  • b.

    Active antecedent, passive ellipsis

    The janitor must remove the trash whenever it is apparent that it should be 〈removed〉.

    (Merchant 2008:169)

In (2a), the voice of the antecedent is passive, but that of the deletion site is active. The resulting sentence is acceptable. As shown in (2b), an active antecedent also allows the deletion of a passive VP.

On the other hand, as Merchant (2008) first pointed out, pseudogapping apparently does not permit voice mismatches, even though it seems to target the same portion of structure. Relevant examples are given in (3).

(3)

  • a.

    Passive antecedent, active ellipsis

    *Roses were brought by some, and others did lilies 〈bring〉.

  • b.

    Active antecedent, passive ellipsis

    *Some brought roses, and lilies were by others 〈brought〉.

    (Merchant 2008:170)

The examples in (3) show that pseudogapping requires that the voice of the antecedent correspond to that of the deletion site.

Given Merchant’s analysis illustrated in (1), the asymmetry is accounted for under the copy theory of movement; the VP-deletion sentence in (2b), for example, has the following structure:

(4)

graphic

The VP in (4) is similar enough to the antecedent in its interpretation. The VP also satisfies the morphosyntactic identity conditions, especially the identity of voice morphology with its antecedent, in a trivial manner, since the deletion site, VP, excludes the Voice head that has the potential to impede identity in the first place. Thus, the example in (2b) is acceptable.

On the other hand, the pseudogapped sentence in (3b), for example, will be assigned the structure in (5). Semantic identity holds between the deletion site, VoiceP, in (5) and its antecedent in (6), since pseudogapped remnants and their contrastive correlates are regarded as the same owing to Focus Closure and '-type shifting (see Merchant 2001 for details).

(5)

graphic

(6)

graphic

However, as indicated in boldface, morphosyntactic identity is not licensed by the voice mismatch between the passive morphology of the deletion site in (5) and the active morphology of its antecedent in (6). As a result, the sentence in (3b) is unacceptable.

To summarize, Merchant’s analysis attributes the asymmetry between VP-deletion and pseudogapping with respect to voice mismatches to the difference in the size of the structures they target.

2 More Data

So far, we have looked only at simple sentences. If we expand our scope to complex sentences where the deleted VP embeds a voice mismatch context, we will observe that VP-deletion is still insensitive to the voice mismatch, contrary to what Merchant’s analysis predicts. Consider the following example:

(7)

  • A: When John had to praise a student, he didn’t want to 〈praise a student〉.

  • B: ?When John had to be scolded by a dean, he didn’t 〈want to be 〈scolded by a dean〉〉, either.3

As indicated with angle brackets, the main clause in (7B) undergoes a series of VP-deletion operations: deletions of the VP headed by scolded and the one headed by want. The type of example in (7) is called sloppy VP-ellipsis because the deleted VP embedded in another deleted VP can be assigned a sloppy interpretation with respect to the antecedent VP (for discussion, see Hardt 1999, Schwarz 2000, Tomioka 2008).

It should be noted that morphological identity is not met here (but the example is acceptable). To see this more clearly, let us consider the structures assigned to the target and antecedent VPs headed by want. They are given in (8) and (9), respectively.

(8)

graphic

(9)

graphic

As indicated in boldface, the target VP contains the passive Voice head stranded by the deletion of the embedded VP scolded by a dean while the active Voice head is left intact by the deletion of the embedded VP praise a student in the antecedent VP. Thus, a voice mismatch should occur here as in the case of pseudogapping. Therefore, Merchant’s analysis would wrongly predict that the example in (7) should be unacceptable.

Now, we have seen that VP-deletion is still insensitive to the identity of voice morphology even when the offending Voice head is embedded in the target VP. One might object, however, that what appears to be an otherwise problematic VP-deletion in (7) is, in fact, null complement anaphora (see, e.g., Hankamer and Sag 1976, Grimshaw 1979, Napoli 1983). Since null complement anaphora is deep anaphora (see Hankamer and Sag 1976), it tolerates voice mismatches. In (10), the null complement is enclosed in curly brackets.

(10)

  • The oats had to be taken down to the bin, so Sandy volunteered {to take the oats down to the bin}.

  • (Zohrab 1986:434)

Then, B’s utterance in (7) would also be predicted to permit the voice mismatch in question. However, this account does not hold for the verb want, used in (7), the example in question. Consider the following examples:

(11)

  • a.

    John tried to take out the garbage, and I think that Bill also tried {to take out the garbage}.

  • b.

    *John wants to leave, but I don’t think that Bill wants {to leave}.

    (Jacobson 1990:439)

As the contrast in (11) shows, the verb want, unlike try, does not permit null complement anaphora. Therefore, the example in (7) is still problematic for Merchant’s analysis.

It could also be objected that B’s utterance in (7) involves IP-deletion followed by VP-deletion. It would then have the following informal representation:

(12)

  • B: When John had [IP to be scolded by a dean], he didn’t

  • [VP 〈want [IP 〈to be scolded by a dean〉]〉], either.

The deleted IP could then probe an appropriate antecedent in the when-clause in (12). The IP-deletion could not only satisfy the semantic and morphosyntactic identity conditions but also eliminate the offending voice mismatch context. Thus, we would correctly predict that B’s utterance in (12) is acceptable. However, this alternative analysis also fails. It is well-known that IP-deletion is limited to wh-questions in English, resulting in what Ross (1969) calls sluicing. Consider the following examples:

(13)

  • a. Somebody just left—guess who 〈 just left〉.

  • (Ross 1969:252)

  • b. Bob knows how to crane his neck, but I don’t know how 〈to crane his neck〉.

  • (Ross 1969:271)

  • c. *Jim says that UConn will win the NCAA, but I don’t believe that 〈UConn will win the NCAA〉.

  • (Takahashi 1994:274)

As shown in (13a–b), sluicing can apply to finite and nonfinite wh-questions. On the other hand, when the target constitutes a declarative clause as in (13c), IP-deletion yields an unacceptable result. Now let us return to the account given in (12). Since B’s utterance in (12) is a nonfinite declarative clause, IP-deletion is not a viable option.4

An anonymous reviewer raises yet another objection. The reviewer correctly points out that the bracketing in (7B)/(12) ignores the option of beP-deletion. This option, if available, could also eliminate the offending voice mismatch context, as shown in (14).

(14)

  • B: When John had to [beP be scolded by a dean], he didn’t

  • [VP 〈want to [beP 〈be scolded by a dean〉]〉], either.

In this respect, note that the nature of beP-deletion is still unclear. First, beP-deletion is semantically conditioned when embedded within infinitival complements to lexical verbs as in the case at hand (see Levin 1985 for discussion). Second, and more important, beP-deletion, even if available, seems to be insensitive to voice mismatches in the first place, as shown in the following examples:

(15)

  • a.

    ‘‘Do I have to talk to these people?’’ ‘‘They will want to 〈be talked to by you〉,’’ Scarborough said.

  • b.

    We tried to move him to a nicer place, but he didn’t want to 〈be moved to nicer place〉.5

    (The Corpus of Contemporary American English)

Since beP layers above VoiceP, Merchant’s analysis would wrongly predict that the deletion of beP is unacceptable in (15). To avoid the problem, we could posit an ad hoc be-deletion. If it applied after VPdeletion and a Voice head is stranded between the two deletion sites, the examples in (15) would be accounted for. However, this analysis cannot be extended to the example in (14), since deletion of the embedding VP includes the stranded Voice head. Thus, both beP- and be-deletion still pose a problem for Merchant’s analysis. Therefore, I conclude that beP-deletion should also be excluded from available options.

To summarize, Merchant’s syntactic analysis of voice mismatches cannot explain the voice insensitivity of sloppy VP-ellipsis in (7).

3 Discussion

Now the following question arises: What degrades pseudogapping examples involving voice mismatches? First, let us note that VP-deletion does not always permit voice mismatches (see, e.g., Kehler 2002, Kertz 2008, 2010, Tanaka 2011). VP-deletion is disallowed in (16), for example.

(16)

  • *This problem was looked into by John, and (similarly) Bob did, too.

  • (Tanaka 2011:478)

Second, Tanaka (2011) points out that pseudogapping also allows voice mismatches in some cases (see Miller 1991, Coppock 2001, Merchant 2008).

(17)

  • ?My problem will be looked into by Tom, but he won’t into yours 〈look〉.

  • (Tanaka 2011:477)

Given these two facts, Kertz (2008, 2010) points out that focus marking, rather than voice mismatch per se, affects the acceptability of VP-deletion and pseudogapping.6 When the subject is focused and becomes a contrastive topic, as in (3) and (16), its correlate cannot be demoted via passive formation but must occupy the same subject position to form a well-formed contrastive topic discourse. Thus, subject focus disallows voice mismatches in VP-deletion and pseudogapping. On the other hand, auxiliary focus, typically signaled with a subject pronoun, does not have such a parallelism requirement and permits voice mismatches, as shown in (2), (17), and possibly also (15).7

Kertz’s findings can be readily extended to the sloppy VP-deletion cases we are discussing. The example in (7), repeated in (18), involves auxiliary focus.

(18)

  • A: When John had to praise a student, he didn’t want to 〈praise a student〉.

  • B: ?When John had to be scolded by a dean, he didn’t 〈want to be 〈scolded by a dean〉〉, either.

Thus, the example in (18B) is acceptable. On the other hand, sloppy VP-deletion does not permit voice mismatches when subject focus is involved through obligatory control, as in (19).8

(19)

  • A: John had to praise a student and Billi wanted PROi to 〈praise a student〉.

  • B: *John had to be scolded by a dean, and Billi did 〈want PROi to be 〈scolded by a dean〉〉.

Thus, Kertz’s analysis not only correctly rules in the example in (18) but also accounts for the contrast between (18) and (19).9

4 Conclusion

In this squib, I have shown that a VP-deletion operation can target the VP that embeds an offending Voice head stranded by another VPdeletion operation. In other words, VP-deletion is still insensitive to the identity of voice morphology between a deletion site and its antecedent even when it matters under Merchant’s (2008) syntactic analysis. I have argued here that voice mismatch phenomena in VP-deletion and pseudogapping are best accounted for by Kertz’s (2008, 2010) analysis based on focus marking.

Notes

I am very grateful to Yoshiaki Kaneko, Etsuro Shima, and two anonymous LI reviewers for invaluable comments and suggestions. I would also like to thank Akihiko Arano, Cornelia Daniela Lupsa, Masahiko Takahashi, and Kensuke Takita for their helpful comments. All remaining errors are my own.

1 For expository purposes, I use Voice, instead of v, to encode the morphosyntactic distinction between active and passive (see, e.g., Kratzer 1996). Moreover, I will assume with Merchant (2001) that a feature [E] on the head licenses deletion of its complement under identity conditions.

3 I have checked the example with ten informants. Seven of them found it acceptable, and thus I marked it ?.

4 An anonymous reviewer points out that stripping would be another IP-deletion operation (see Depiante 2000). However, it is widely held that stripping cannot occur inside an embedded clause, as shown in (i).

(i)

  • *John read El Quijote but I think not Mary 〈read El Quijote〉.

  • (Depiante 2000:104)

Thus, stripping should not be a viable option, either (see also Culicover and Jackendoff 2005).

5 The deleted part might be an unaccusative vP instead. It then suggests that VP-deletion also permits what Merchant (2012) calls subject/nonsubject alternations.

6 I would like to thank an anonymous reviewer for bringing Kertz 2008, 2010 to my attention. I also owe the reviewer much credit for the analysis in this section. See Kertz 2010 for further details.

7 See Tanaka 2011, which employs Kehler’s (2002) resemblance and cause-effect relations to account for the facts. However, I should note that the example in (7)/(18) forms a resemblance relation and thus would wrongly be predicted to be unacceptable.

8 The same reviewer mentioned in footnote 6 kindly offers the example in (19), but there is a caveat: in the reviewer’s original example, too is added after B’s utterance. I omit it to single out subject focus as a sole factor. I have checked the example with four of the seven informants who found the example in (7)/(18) acceptable. All of them found (19) unacceptable.

9 An anonymous reviewer kindly offers the following examples:

(i)

  • a.

    *An ID was required to be shown by everyone even though Bill wasn’t 〈required to 〈show an ID〉〉.

  • b.

    *An ID was required to be shown by everyone except for Bill, who wasn’t 〈required to 〈show an ID〉〉.

They are simpler than the example in (7)/(18) in that no resolution of the sloppy VP-ellipsis is necessary. Moreover, we can exclude the IP/beP-deletion options discarded in section 2 as unavailable options to eliminate a voice mismatch context. I have checked (ia–b) with five informants who accept symmetric passive. Three found them unacceptable. Of the remaining two, one was unsure about his judgment, and the other found (ib) acceptable. Thus, I marked (ia–b) *. At first sight, the result appears to discredit the example in (7)/(18). On the contrary, the result is expected under an alternative analysis based on the subject/auxiliary focus distinction just given: in fact, both (ia) and (ib) are subject focus sentences.

References

Coppock
,
Elizabeth
.
2001
.
Gapping: In defense of deletion
. In
CLS 37. Vol. 1, The Main Session
, ed. by
Mary
Andronis
,
Christopher
Ball
,
Heidi
Elston
, and
Sylvain
Neuvel
,
133
148
.
Chicago
:
University of Chicago, Chicago Linguistic Society
.
Culicover
,
Peter W
., and
Ray
Jackendoff
.
2005
.
Simpler syntax
.
Oxford
:
Oxford University Press
.
Dalrymple
,
Mary
,
Stuart M.
Shieber
, and
Fernando C. N.
Pereira
.
1991
.
Ellipsis and higher-order unification
.
Linguistics and Philosophy
14
:
399
452
.
Davies
,
Mark
.
2008–
.
The corpus of contemporary American English: 450 million words, 1990–present
. .
Depiante
,
Marcela
.
2000
.
The syntax of deep and surface anaphora: A study of null complement anaphora and stripping/bare argument ellipsis
.
Doctoral dissertation, University of Connecticut, Storrs
.
Fiengo
,
Robert
, and
Robert
May
.
1994
.
Indices and identity
.
Cambridge, MA
:
MIT Press
.
Grimshaw
,
Jane
.
1979
.
Complement selection and the lexicon
.
Linguistic Inquiry
10
:
279
326
.
Hankamer
,
Jorge
, and
Ivan
Sag
.
1976
.
Deep and surface anaphora
.
Linguistic Inquiry
7
:
391
428
.
Hardt
,
Daniel
.
1993
.
Verb phrase ellipsis: Form, meaning, and processing
.
Doctoral dissertation, University of Pennsylvania, Philadelphia
.
Hardt
,
Daniel
.
1999
.
Dynamic interpretation of verb phrase ellipsis
.
Linguistics and Philosophy
22
:
185
219
.
Jacobson
,
Pauline
.
1990
.
Raising as functional composition
.
Linguistics and Philosophy
13
:
423
475
.
Jayaseelan
,
K
. A.
1990
.
Incomplete VP deletion and gapping
.
Linguistic Analysis
20
:
64
81
.
Johnson
,
Kyle
.
2001
.
What VP ellipsis can do, and what it can’t, but not why
. In
The handbook of contemporary syntactic theory
, ed. by
Mark
Baltin
and
Chris
Collins
,
439
479
.
Malden, MA
:
Blackwell
.
Johnson
,
Kyle
.
2004
.
How to be quiet
. In
CLS 39-1. Vol. 1, The Main Session
, ed. by
John
Cihlar
,
Amy
Franklin
,
Dave
Kaiser
, and
Irene
Kimbara
,
1
20
.
Chicago
:
University of Chicago, Chicago Linguistic Society
.
Kehler
,
Andrew
.
2002
.
Coherence, reference, and the theory of grammar
.
Stanford, CA
:
CSLI Publications
.
Kertz
,
Laura
.
2008
.
Focus structure and acceptability in verb phrase ellipsis
. In
WCCFL 27: Proceedings of the 27th West Coast Conference on Formal Linguistics
, ed. by
Natasha
Abner
and
Jason
Bishop
,
283
291
.
Somerville, MA
:
Cascadilla Proceedings Project
.
Kertz
,
Laura
.
2010
.
Ellipsis reconsidered
.
Doctoral dissertation, University of California, San Diego
.
Kratzer
,
Angelika
.
1996
.
Severing the external argument from its verb
. In
Phrase structure and the lexicon
, ed. by
Johan
Rooryck
and
Laurie
Zaring
,
109
137
.
Dordrecht
:
Kluwer
.
Lasnik
,
Howard
.
2001
.
When can you save a structure by destroying it?
In
NELS 31
, ed. by
Minjoo
Kim
and
Uri
Strauss
,
301
320
.
Amherst
:
University of Massachusetts, Graduate Linguistic Student Association
.
Levin
,
Nancy
.
1985
.
Main-verb ellipsis in spoken English
.
New York
:
Garland
.
Merchant
,
Jason
.
2001
.
The syntax of silence
.
Oxford
:
Oxford University Press
.
Merchant
,
Jason
.
2008
.
An asymmetry in voice mismatches in VP-ellipsis and pseudogapping
.
Linguistic Inquiry
39
:
169
179
.
Merchant
,
Jason
.
2012
.
Voice and ellipsis
.
Ms., University of Chicago, Chicago, IL
.
Miller
,
Philip
.
1991
.
Clitics and constituents in phrase structure grammar
.
Doctoral dissertation, Utrecht University
.
Napoli
,
Donna Jo
.
1983
.
Missing complement sentences in English: A base analysis of null complement anaphora
.
Linguistic Analysis
12
:
1
28
.
Ross
,
John Robert
.
1969
.
Guess who?
In
Papers from the 5th Regional Meeting of the Chicago Linguistic Society
, ed. by
Robert
Binnick
,
Alice
Davison
,
Georgia M.
Green
, and
Jerry L.
Morgan
,
252
286
.
Chicago
:
University of Chicago, Chicago Linguistic Society
.
Sag
,
Ivan
.
1976
.
Deletion and Logical Form. Doctoral dissertation, MIT, Cambridge, MA
.
Published, New York
:
Garland (1979)
.
Schwarz
,
Bernhard
.
2000
.
Topics in ellipsis
.
Doctoral dissertation, University of Massachusetts, Amherst
.
Takahashi
,
Daiko
.
1994
.
Sluicing in Japanese
.
Journal of East Asian Linguistics
3
:
265
300
.
Takahashi
,
Shoichi
.
2004
.
Pseudogapping and cyclic linearization
. In
NELS 34
, ed. by
Keir
Moulton
and
Matthew
Wolf
,
571
585
.
Amherst
:
University of Massachusetts, Graduate Linguistic Student Association
.
Tanaka
,
Hidekazu
.
2011
.
Voice mismatch and syntactic identity
.
Linguistic Inquiry
42
:
470
490
.
Tomioka
,
Satoshi
.
2008
.
A step-by-step guide to ellipsis resolution
. In
Topics in ellipsis
, ed. by
Kyle
Johnson
,
210
228
.
Cambridge
:
Cambridge University Press
.
Zohrab
,
P. D
.
1986
.
Verb-phrase anaphora: Linguistics or cognitive science?
Studies in Language
10
:
425
442
.