See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/284513708

Gerunds vs. Infinitives in English: Not Meaning but Form

Conference Paper · September 2013

CITATIONS READS

0 340

1 author:

Joseph Emonds
Palacký University Olomouc
81 PUBLICATIONS   2,925 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

English plurals, via Middle English, have their source in Old Norse plurals. View project

All content following this page was uploaded by Joseph Emonds on 24 November 2015.

The user has requested enhancement of the downloaded file.

 From Theory to Practice 2013

Proceedings of the Fifth International Conference

on Anglophone Studies

September 5–6, 2013

Tomas Bata University in Zlín, Czech Republic

Edited by
Roman Trušník
Gregory Jason Bell
Katarína Nemčoková

Univerzita Tomáše Bati ve Zlíně

Zlín 2015
Organizing Committee: Gregory Jason Bell
Hana Čechová
Vladimíra Fonfárová
Katarína Nemčoková
Jana Semotamová
Roman Trušník
Petr Vinklárek

Reviewers: Ema Jelínková

Susan Nagelsen
Ludmila Veselovská
Bartosz Wójcik

First Edition

Arrangement copyright © Roman Trušník, Gregory Jason Bell, Katarína Nemčoková,

2015
Editors’ Note copyright © Roman Trušník, Gregory Jason Bell, Katarína Nemčoková, 2015
Papers copyright © Gregory Jason Bell, Barbora Blažková, Šárka Bubíková, Joseph E.
Emonds, Jiří Flajšar, Pavlína Flajšarová, Vladimíra Fonfárová, Markéta Gregorová,
Kateřina Havranová, Petra Huschová, Karla Kovalová, Ivan Lacko, Roman Ličko,
Thomas Lorman, Dagmar Machová, Marcela Malá, Małgorzata Martynuska,
Małgorzata Paprota, Jozef Pecina, Jana Richterová, Zuzana Starovecká, Dita Trčková,
Roman Trušník, Ludmila Veselovská, Radek Vogel, Michaela Weiss, 2015
All rights revert to the authors upon publication.

Published by Tomas Bata University in Zlín, Czech Republic

ISBN 978-80-7454-450-7
ISSN 1805-9899
 Table of Contents

Editors’ Note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Roman Trušník, Gregory Jason Bell, Katarína Nemčoková

Linguistics
Gerunds vs. Infinitives in English: Not Meaning but Form . . . . . . . . . . . . . . . . . . . . . . . . 13
Joseph E. Emonds

Adjective Hierarchy: Comparing the Order of Adjectives

in the Prenominal Field in English and Czech . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Ludmila Veselovská

(De)verbal Modifiers in Attribute + Noun Collocations and Compounds:

Verbs, Deverbal Nouns or Suffixed Adjectives? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Radek Vogel

Modal Verbs from a Crosslinguistic Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Dagmar Machová

On Possibility Meanings of Could in Written English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

Petra Huschová

Recent Syntactic Research on Event and Result Nominals in English and Czech . . . . 99
Kateřina Havranová

Changing Clause Types in Written English . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

Marcela Malá

The Highlighting Syntactic Structures in English and Their Translation

into Czech in Collected Samples of Literary Texts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Jana Richterová

The “Wrong Direction”? Translating into English as a Foreign Language

in Slovakia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Roman Ličko
Self-marketing in British Online Personal Advertisements . . . . . . . . . . . . . . . . . . . . . . . 141
Dita Trčková

Voices in the Headlines: A Critical Discourse Analysis

of British Web-news Headlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Barbora Blažková

Inclusion and Role Allocation in the Representation of the Social Actors

of the Welfare State in Conservative British Press (2008–2012) . . . . . . . . . . . . . . . . . . . 165
Małgorzata Paprota

Literature and Cultural Studies

Anglo-American Historians and Czechoslovakia, 1918–1989 . . . . . . . . . . . . . . . . . . . . . 179
Thomas Lorman

Ye Mystic Krewe of Historical Revisionists:

The Origins of Tampa’s Gasparilla Parade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Gregory Jason Bell

A Star-Shaped Crossroad: From (Counterfactual) Historiography

to Historiographic Metafiction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
Vladimíra Fonfárová

Antebellum Urban America and Sensational Novels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213

Jozef Pecina

Suburban Humor in the Poetry of Billy Collins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

Jiří Flajšar

How Visual Poetry Can Get: Visual and Textual Aspects

in the Poetry of Grace Nichols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
Pavlína Flajšarová

To Look High, Low, and Beyond: Shifting the Textual Terrain

of Black Feminist Literary Criticism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Karla Kovalová
Multivocality, Identity and Tradition in Michael Dorris’s
A Yellow Raft in Blue Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
Šárka Bubíková

Freedom Stuck in the Throat: Yussef El Guindi’s Back of the Throat through
the Prisms of an Individual Conflict and a Polarized Society . . . . . . . . . . . . . . . . . . . . . 257
Ivan Lacko

Destabilizing Ethnic Stereotypes in American Mainstream TV:

Latino/a Representations in Ugly Betty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 265
Małgorzata Martynuska

Beauty Is the Beast:

Narcissism, Identity and the Ethics of Violence in Invisible Monsters . . . . . . . . . . . . . . 273
Zuzana Starovecká

Becoming Whole Again: Deconstructing Gender Dualities

in Emma Tennant’s The Bad Sister . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
Markéta Gregorová

Prison as a Queer Space in Sarah Waters’s Affinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289

Michaela Weiss

Michael Cunningham in the Czech Republic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297

Roman Trušník
 Editors’ Note
Roman Trušník, Gregory Jason Bell, Katarína Nemčoková

The present volume, the fifth in the Zlín Proceedings in Humanities book series, contains
selected papers from “From Theory to Practice 2013: The Fifth International Conference
on Anglophone Studies,” hosted on September 5–6, 2013, by the Department of Modern
Languages and Literatures, Faculty of Humanities, Tomas Bata University in Zlín, Czech
Republic.
The 2013 conference was quite possibly the best so far, with eighty participants
from eight different countries presenting their findings on a wide array of Anglophone
Studies topics, from Mozart to motorcycles, and vampires to verbs. Although not every
presenter submitted a paper to the proceedings, the editors had plenty to choose
from, yet they were burdened with the responsibility of making difficult choices.
Ultimately, this proceedings contains articles written by scholars from the Czech
Republic, Slovakia, Poland, and the United Kingdom, along with two articles written
by American expatriates representing Czech universities. Due to the number of articles
submitted and reviewed, the editorial process has taken more time than in the previous
years; on the other hand, this enabled many authors to bring their papers up to date
and reflect on the latest developments, thus affirming the conference’s status as one
monitoring the latest research trends in the region.
While we are always happy to see articles from previous volumes cited (and, they do
get cited all around the globe, demonstrating the academic fecundity of the conference
series), we are even more pleased to see a paper presented at our conference and
published in our proceedings expanded into a full-fledged book. Indeed, the roots
of English: The Language of the Vikings (2014) by Joseph Embley Emonds and Jan
Terje Faarlund can be traced back to Emonds’s article “English as a North Germanic
Language: From the Norman Conquest to the Present,” published in the second volume
of the series.
As previously, this book is published in print to be distributed primarily to
libraries and, in order to facilitate the rapid exchange of information among scholars
all over the world, it is simultaneously released in PDF form on the Internet
(http://conference.uaa.utb.cz/tp2013). The volume is divided into a linguistics section
and a literature and cultural studies section, and each section adheres to the appropriate
citation formats defined in the current edition of The Chicago Manual of Style: papers on
linguistics make use of the author-date form, while papers in the literature and cultural
studies section are in the notes-bibliography style.
We are thankful to the participants at the conference, without whose contributions
this book would not have been possible. Thanks are due to our review board, as well
as our student assistant, Dana Jašková. Moreover, we wish to thank the conference
organizers, our two conference keynote speakers, Thomas A. Lorman and Christiane
Nord, the dean of the Faculty of Humanities, and the Zlín Region.
10 From Theory to Practice 2013

Works Cited
Emonds, Joseph Embley. 2011. “English as a North Germanic Language: From the
Norman Conquest to the Present.” In Theories and Practice: Proceedings of the Second
International Conference on English and American Studies, September 7–8, 2010,
edited by Roman Trušník, Katarína Nemčoková, and Gregory Jason Bell, 13–26.
Zlín: Univerzita Tomáše Bati ve Zlíně.
Emonds, Joseph Embley, and Jan Terje Faarlund. 2014. English: The Language of the
Vikings. Olomouc: Palacký University, Olomouc.
 Gerunds vs. Infinitives in English:
Not Meaning but Form
Joseph E. Emonds
Tomas Bata University in Zlín, Faculty of Humanities, Department of Modern Languages and Literatures,
Mostní 5139, 760 01 Zlín, Czech Republic. Email: jeemonds@hotmail.com

Abstract:
verb phrases as syntactic subjects and objects, even though this assumption is falsified by the

expandable noun phrases, but infinitives and finite clauses, termed ‘verbal clauses’ in this study,
actually never occur in noun phrase positions. What is shown here is that (i) verbal clauses that
appear to be ‘objects’ of verbs are in clause-final position, rather than in the position of object
noun phrases, and that (ii) initial verbal clauses that appear to be ‘subjects’ are in a pre-subject

showing that initial verbal clauses, in contrast to lexical noun phrase and gerunds, never occur in
embedded or inverted subject positions. In this way, initial verbal clauses have the same distribution
as other well-known ‘root’ or ‘focus’ constructions, which in pre-subject position are limited to
main clauses.

Keywords: clausal complement; expletive; extraposition; gerund; infinitive; topicalization

1. Some misleading similarities of gerunds and infinitives

Grammars of English point outthat according to the form oftheir verbs, there are three
types of subordinate clause complements: (i) finite subordinate clauses introduced by
that (which can sometimes be zeroed), (ii) infinitives, (the verb has no inflection), and
(iii) gerunds / participles, whose first verb ends in -ing.1
Apparently all three types can serve as either subjects (1a-c) or objects (2a-c), with
various restrictions on the classes of verbs that selectthem.

(1) a. we had to use that airport annoyed us.

Bill knows German well was made obvious to all of us.
b. To find a well-paying job nearby would be a pleasant surprise.
To read so many magazines is a waste oftime.
For the house to be painted would confuse him.
c. Your being able to find a new job nearby would be surprising.

1. With certain main verbs, subordinate clause complements can also take the form o “indirect questions,”
i.e., finite or infinitival clauses introduced by if, whether or a Wh-phrase. For our purposes here, finite
indirect questions are treated with finite complements, and infinitival indirect questions are grouped
with infinitives. Traditional grammar glosses over the impossibility offorming indirect questions with
gerunds (
14 From Theory to Practice 2013

Reading so many magazines is a waste of time.

Mary’s having so many books impressed him.

(2) a. Ann believed (that) Mary was a foreign agent.

Ann will see to it that you have a reservation.
b. Bill would prefer for Mary to stay awhile.
Susan tried to buy a ticket on time.
Barbara decided to buy fewer foreign books.
c. Ann regretted stealing Mary’s book.
Ann will see to your buying a ticket in time.
Bill would prefer buying fewer foreign books.

However, these apparently free alternations are very misleading; there are more
differences than similarities between infinitives and gerunds. For example, one widely
recognized asymmetry is that as objects of prepositions, only gerunds (V+ing) occur
freely as in (3a), in contrast to finite and infinitival clauses, which appear after only a
few temporal prepositions such as until and after, but with others not at all, as in (3b).
For some examples with temporal prepositions, see (19).

(3) a. John just came back from driving his cab.

She blamed it on Bill’s being too strict.
Because of John’s being old, Mary gets a pension.
Your explanation for the table’s being badly scratched sounds suspicious.
b. *John just came back from (to) drive his cab.
*She blamed it on (Bill) to be too strict.
*Because of (John) to be old, Mary gets a pension.
*Your explanation for the table to be badly scratched sounds suspicious.

This restriction is usually attributed to some unpredictable asymmetry in English, with

no further analysis. We will see below that the difference in (3a-b) falls out naturally
from a more general structural hypothesis about gerunds.
Terminology. Since this paper groups together finite and infinitival clauses on the
one hand, and gerunds and participles on the other, we need a term that encompasses
the former types but not the latter. So rather than resort to a new label or to the
clumsy phrase, “non-gerund, non-participial clauses,” the term “verbal clause” will
here throughout include only finite and infinitival clauses and exclude gerunds and
participles. As we proceed, an informal justification for this term will arise from the fact
that gerunds share aspects of noun phrases, and participles share aspects of adjective
phrases, which finite and infinitival clauses don’t share. So in a way, only the latter are
“purely verbal.”
(4) Verbal Clauses. In this study, “Verbal Clauses” refer to finite and infinitival
clauses, and exclude gerunds and participles, i.e., clauses whose head is V-ing.
 Joseph E. Emonds 15

Thus, the term “verbal clause” does not encompass the gerunds in (1c), (2c) and (3a).
This decision has no implication for how clauses should be assigned structure or
interpretation; the term in (4) simply provides an easy way to refer to non-gerund and
non-participle clauses.
A typical approach to differences between infinitives (1b)-(2b) and gerunds (1c)-(2c)
is to say that they are alternatives located in the same structural positions, distinguished
by some delicate and formally unspecified “differences in meaning.” The nature of these
differences is elusive, since the complements in, e.g., Mary forgot to turn down the
heat and Mary forgot turning down the heat differ in factivity, while John started to
work on his paper and John started working on his paper differ only in some “semantic
nuance” which does not involve truth values of the complement clause.2 Similarly, the
difference between I’d like to have / having an expensive car is again only a “nuance.”
Such imprecise semantics-based discussions, no matter how many ways rephrased,
never succeed in uncovering a predictive general distinction between infinitives and
gerunds / participles.

2. Gerunds are Noun Phrases

The semantics-based distinctions of traditional grammar, which some generative
treatments have more or less continued, can be fruitfully replaced by an accurate, simple
and empirically testable structural distinction between gerunds and verbal clauses. It
turns out that the “pre-verbal” position of verbal clauses in (1a-b) is not the same as
the subject position of gerunds in (1c). In the same vein, the “post-verbal” position of
embedded clauses in (2a-b) is not the same as the object position of gerunds in (2c).3
(5) English Gerund Distribution. English gerunds are DPs. They occur in all and
only the positions where noun phrases (“DPs”) can be freely generated.
Ordinarily, there is a single phrasal position before a clause’s first verbal position (which
here as in many sources is termed “I”), which contains either a finite auxiliary or the
infinitival marker to. This single phrase is the subject DP. So by (5) gerunds should
be uniformly acceptable as DP subjects not only in main clauses but also in all types of
embedded clauses. The underlined gerunds (6i-vi) confirm this prediction. (Throughout,
σ stands for Verbal Clauses, and does not refer to gerunds.)

2. The “method” of traditional descriptions is to take an individual verb and comment on whatever
meanings come into the mind of the analyst. Thus with remember to V, the remembering precedes
the action, whereas with remember V+ing, the remembering follows the action. The comments must be
modified for each verb considered (in forget to V, the activity need not happen at all, and with either
use of start, one cannot distinguish the starting and the action). These discrepancies are of no matter
in these analyses, as their goal is to accumulate commentary rather than to express generalizations.
3. Throughout, I use the now common symbol DP for (Determined) Noun Phrase, or Noun Phrase which
has a possible Determiner. Readers used to the symbol NP should simply take the two symbols as
equivalent here.
16 From Theory to Practice 2013

(6) (i) Subjects of clausal complements to Verbs, including indirect questions:

I don’t believe (that) [σ your studying history helps you].
She forgets [CP how expensive [σ visiting the dentist is]].

(ii) Subjects of clausal complements to Adjectives:

John was happy that [σ owning a car didn’t disqualify you].
Nobody is ready for [σ wearing headphones to be legally required].

(iii) Subjects of clausal complements to Nouns:

He protested the decision that [σ his / him being on time counted for nothing].
We are noticing a tendency for [σ smoking tobacco to be criminalized].

(iv) Subjects of comparative clauses:

A day at the beach is more fun than [σ playing golf is].
Going by car doesn’t seem as relaxing as [σ riding a horse used to seem].

(v) Subjects of adverbial clauses:

Although [σ the house(’s) being empty may depress you], it pleases me.
He exercises so rarely that [σ lifting those bricks is bad for his heart].

(vi) Subjects of relative clauses, whatever the function of the relativized phrase:
Pupils for whom [σ diagramming sentences was easy] often became linguists.
Being a citizen is the reason why [σ your having insurance can protect you].
Situations in which [σ writing out a check is necessary] should be avoided.
It was the salesman who [σ my / me buying a car seemed important to].
She likes the kind of man that [σ seeing a few movies a year will satisfy].

The next section will contrast the pattern of gerunds in DP subject positions in (1c) with
the distinct “pre-verbal position” of verbal clauses in (1a-b).

3. Verbal clause subjects: not in subject position

Unlike gerunds, verbal clauses in the pre-verbal position of main clauses are not DPs.
They are rather “topicalized” initial constituents in a clause (CP). As soon as some clause
σ is embedded in another CP, a verbal clause (finite or infinitival) in σ’s initial position
becomes ungrammatical, as shown by the underlined examples (7i-vi); compare these
point for point with (6i-vi).

(7) Six contexts for subject where verbal clauses cannot appear:
(i) Verbal clause subjects in complements to Verbs:
*I don’t believe (that) [σ for you to study history would help you].
*She forgets [CP how expensive [σ to visit the dentist is]].
 Joseph E. Emonds 17

(ii) Verbal clause subjects in complements to Adjectives:

*John was happy that [σ to own a car didn’t disqualify him].
*Nobody is ready for [σ that one exercise daily to be legally required].

(iii) Verbal clause subjects in complements to Nouns:

*He protested the decision that [σ for him to be on time counted for nothing].
*We are noticing a tendency for [σ that we can smoke them to be criminalized].

(iv) Verbal clause subjects in comparative clauses:

*A day at the beach is more fun than [σ to play golf is].
*Going by car doesn’t seem as relaxing as [σ to ride a horse used to seem].

(v) Verbal clause subjects in adverbial clauses:

*Although [σ that the house is empty may depress you], it pleases me.
*Although [σ for the house to be empty may depress you], it pleases me.
*He exercises so rarely that [σ to lift those bricks is bad for his heart].

(vi) Verbal clause subjects in relative clauses, whatever the relative’s function:
*Pupils for whom [σ to diagram sentences was easy] often became linguists.
*Being a citizen is the reason why [σ that you have insurance protects you].
*Situations in which [σ to write out a check is necessary] should be avoided.
*It was the salesman who [σ for me to buy a car seemed important to].
*She likes the kind of man that [σ to see a few movies a year will satisfy].

These diverse paradigms show clearly that verbal clauses (= infinitives and finite
clauses) cannot generally occur in the “subject position,” in contrast to the gerunds in
(6). Rather, these clausal CPs occur initially in the “topicalized” position in only main,
or “root” clauses.4 An issue in more recent literature concerns whether different types
of root clauses have special labels such as TopP and FocP (Rizzi 1997) or whether these
types are to be identified only by their tree geometry (whether they contain a gap or
not; cf. Emonds 2012). This distinction plays no role in the discussion here.
The actual subject position in the main clauses of (1a-b) is a null DP between the
preposed verbal clause CP and the finite or infinitival position I. This topicalized, pre-
subject CP is then co-indexed with an “empty expletive” DP, as in (8). Section 7 returns
to characterising expletives more formally.
(8) [CP That Bill knows German well]j [IP [DP Ø]j [I was] [VP made obvious to all of us]].
[CP For the house to be painted]j [IP [DP Ø]j [I would] [VP confuse him]].

4. The term “topicalized” for pre-subject phrases other than adverbials is due to Ross (1967), who in the
same work gave the first patterns showing that such phrases are largely limited to main clauses. This
restriction is more fully exemplified and explained theoretically in Emonds (1976), where the term “root
clause” is defined to include main clauses and some other unembedded constructions that share certain
of their properties.
18 From Theory to Practice 2013

The following generalization unites the exclusions in (7) with the structure in (8).
(9) Exclusion of verbal clauses in subject position. English verbal clauses
never occur in a DP position.
That is, clause-initial verbal clauses are either excluded entirely or external to IP and
initial in CP. This conclusion is further confirmed by clauses in which the finiteness
constituent I separates the initial positions of CP and IP, rather than following them
both. As (5) predicts, only gerunds are then allowed in the subject position of IP, for
example when a Wh or a negative phrase and an inverted auxiliary I both precede IP:
These preposed (underlined) constituents cannot be followed by a topicalized position,
and so verbal clauses can never follow them:

(10) Why did [IP {Mary’s liking old records / *that Mary liked old records} irritate him]?
When did [IP {arriving so early / *to arrive so early} become a requirement]?
N ever will [IP {being comfortable / *to be comfortable} be a priority in this office].
A disease like that [IP {taking a lot of pills / *to take a lot of pills} won’t cure].

As we will now see, the Wh-phrase is not the crucial factor here. The problem with
the ungrammatical examples in (10) is that finite I inverts only over an DP position, but
never over a topicalized clause σ in pre-subject position.5 Since verbal clause subjects in
declaratives like (1a-b) never invert with an auxiliary, as in (11), it must be that, unlike
gerunds, verbal clauses are not DPs.

(11) *Does [σ that we have to use that airport] annoy you?

*Wouldn’t [σ to find a well-paying job nearby] be a pleasant surprise?
*Isn’t [σ to read so many magazines] a waste of time?
*Would [σ for the house to be painted] confuse him?

As expected by (5), no such restriction applies to inverted gerund subjects (12), because
they are possible in any DP position:

(12) Does us / our having to use that airport annoy you?

Wouldn’t finding a well-paying job nearby be a pleasant surprise?
Isn’t reading so many magazines a waste of time?
Would the house being painted confuse him?

Since initial verbal clauses are “to the left of the subject,” we should determine more
exactly where. In fact, because they are incompatible with either preceding (10) or
following (13) Wh-phrases, they must be in the same surface position as a Wh-phrase:

5. This fronting is traditionally termed “subject-auxiliary inversion,” though in current terms it is more
accurately described as “I to C movement” in the “Barriers framework” of Chomsky (1986).
 Joseph E. Emonds 19

(13) *That Mary liked old records why irritated him?

*To arrive so early when became a requirement?
*That Sue is unqualified how obvious is?
*To hire John how easy was?

Ross (1967) suggests that these verbal clauses are excluded in subject position only
because in (10)-(12) they are “sentence-internal.” However, examples (14a) show rather
that verbal clauses are equally well excluded at the right edge as well as “internally.”

(14) a. *How obvious is that Sue is unqualified?

*How easy was to hire John?
b. How obvious is Sue’s being unqualified?
How easy was hiring John?

As expected by (5), gerunds are immune to this restriction on verbal clauses, as seen in
(14b).
A final paradigm illustrating (9) is that verbal clauses are always ungrammatical as
subjects of other clauses that are already initial. The initial bracketed clauses in (15a)
allow gerunds in this position (15b), as (5) predicts. But (15c) shows that these same
clauses do not allow another verbal clause in their first position, because in general, as
seen in (15d), subordinate clauses exclude pre-subject XPs.

(15) a. [That cigar smoke bothers the teacher] is quite possible.

[For the house repairs to be so expensive] is no surprise.
b. [That smoking cigars bothers the teacher] is quite possible.
[For repainting the house to be so expensive] is no surprise.
c. *[That to smoke cigars bothers the teacher] is quite possible.
*[For that we are repainting the house to be so expensive] is no surprise.
d. *[That this teacher cigar smoke bothers] is quite possible.
*[For how very expensive the house repairs are] is no surprise.

This section has now substantiated the claim that initial verbal clauses occur not as
subject DPs, but only in a main clause initial position. In this way they resemble several
other English “root constructions” studied in the last forty years, such as those in (16):

(16) a. Down the street rolled the baby carriage.

b. Into the street with those boxes!
c. More interesting would be a talk on ceramics.
d. In the hall was hanging a portrait of Lincoln.
e. One more bottle out the window and I’ll call the police.
f. Isn’t Susan wonderful!
20 From Theory to Practice 2013

For the purposes of this paper, I will simply take (17) as a broad and accurate descriptive
generalization:
(17) English verbal clause “subjects.” A non-DP argument of a verb, such as a
verbal clause, can occur sentence-initially only in a main or “root” clause.
Exactly how to account for such root / embedded asymmetries is beyond the scope of
this paper.6
As a counterpart to the distributional generalization for gerunds (5), I now make
an equally general claim about the distribution of English (finite and infinitival) verbal
clauses. The first part of this claim is what has just been established, and the second
part is the topic of the next two sections.
(18) Distribution of English Verbal Clauses. When embedded, verbal clauses can
occur only (i) topicalized in main (=root) clauses, or (ii) extraposed, at the end of
phrases.7

4. Gerund objects vs. verbal clause complements of P

The category P in current formal grammar has a larger extension than in traditional
grammar, and for very good reasons. Different uses of a word which both exhibit
the same pre-modification system (e.g., the intensifier right, measure phrases) and
which also satisfy the same selectional requirements should be in a single category
(for example, transitive and intransitive verbs are both in the same category V). Thus,
traditional prepositions and subordinating conjunctions of time such as until, before,
after and since are all in the category P, and differ only in whether they take DP or IP
complements.

6. There are several approaches in the generative literature to how grammatical theory should exclude
embedded structures as in (17). In the last decade, that of Haegeman (2009, 2011) has been very
influential. She and other authors claim that the fronted phrases themselves block subordinators and
complementizers like when and if from entering trees (“Merging”). This approach involves somewhat
elaborate accounts of various movements into clause-initial positions, which can be blocked by
“intervention effects,” which force a derivation either to terminate as a root clause or alternatively
to “crash.” In my view, this framework has serious problems, and Emonds (2012) is an account of a
different sort.
7. Though we have shown that English verbal clauses do not occur in a DP subject position, one might
try to maintain that when topicalized, as in (1a-b), they somehow “become” DPs, so as to defend the
traditional view that verbal clauses are some kind of noun phrase. This might be plausible if pre-subject
topicalized constituents were always limited to noun phrases, but pre-subject topics can be of any
phrasal category, as shown by the “root constructions” discussed in Emonds (1976, Ch. II):
With John she often talked about politics.
Into the storeroom Mary threw the extra blankets.
Down the street the baby carriage rolled / rolled the baby carriage.
We thought she would win the prize, and win the prize she finally did.
How ill did the boy seem?
Clever though the speaker was, he failed to convince.
So special it all seemed.
 Joseph E. Emonds 21

(19) The girls worked right until {their bedtime / they went to bed}.
The explosion occurred two hours before {the cabinet meeting / the cabinet met}.
They will put the interview after {the cabinet meeting / the cabinet meets}.

Quite generally, Ps, even in this wider sense, select a single phrasal complement (Stowell
1981).8 So a traditionally named preposition is lexically specified as P, +___DP, while
a traditional subordinating conjunction is specified as P, +___IP, i.e., it takes verbal
clauses rather than DPs.
Given the distributional generalization for gerunds (5), it is expected that gerunds
will occur with any and all prepositions (which have the selection feature +___DP). This
is what in fact we have already observed for the typical Ps from, on, for, because of in
(3a), repeated here:

(3) a. John just came back from driving his cab.

She blamed it on Bill(’s) being too strict.
Your explanation for the table(’s) being badly scratched sounds suspicious.
Because of John(’s) being old, Mary gets a pension.

Since these same four Ps lack a selection feature +___IP, they are incompatible with
infinitival verbal clause complements, as we have seen in (3b):

(3) b. *John just came back from (to) drive his cab.
*She blamed it on (Bill) to be too strict.
*Your explanation for the table to be badly scratched sounds suspicious.
*Because of (John) to be old, Mary gets a pension.

If we instead choose Ps with the selection feature +___IP (such as in order, although,
because) but lacking the selection feature +___DP, gerunds are predictably excluded:

(20) a. Our friends ordered schnitzels, in order that they could / to drive all night.
*Our friends ordered schnitzels, in order for a better trip.
*Our friends ordered schnitzels, in order (for) driving all night.
b. We were cold, although we were wearing warm clothing.
*We were cold, although warm clothing.
*We were cold, although dressing warmly.
c. Because John is so old, Mary gets a pension.
*Because John’s age, Mary gets a pension.
*Because John’s being so old, Mary gets a pension.

8. Emonds (1985, Ch. 7) presents arguments that verbal clauses represented by the symbol CP (= C + IP) are
special cases of PPs, whose head is a grammatical P that takes a clausal sister. That is, the subordinate
conjunctions of traditional grammar should be re-analysed as prepositions with clausal complements.
22 From Theory to Practice 2013

In (21), despite, which seems to mean the same as although but differs from it only in
its subcategorization feature, exhibits the expected reversal of acceptability judgments,
with the gerund again patterning as DPs.

(21) *We were cold, despite we were wearing warm clothing.

We were cold, despite warm clothing.
We were cold, despite dressing warmly.

The verbal clauses selected by traditional “subordinating conjunctions” such as in order

and although are required by (18) to be final in their PPs. Of course, since they are sole
sisters to P, it is not word order that distinguishes them from gerunds, but rather the
classes of Ps that select them. Gerunds are predictably selected like other DPs, while
infinitives are selected like finite clauses.
Finally, it follows that verbal clauses, even though they appear as topicalized subjects
of active verbs (22a), are impossible in passive by-phrases, which can contain only DPs
(22b).

(22) a. That some boys were dancing together was amusing John.
For Susan to arrive early would cause embarrassment.
That you spoke out of turn didn’t help the situation.
To suggestion devaluation would anger the bankers.
b. *John was being amused (by) that some boys were dancing together.
*Embarrassment would be caused (by) for Susan to arrive early.
*The situation wasn’t helped (by) that you spoke out of turn.
*The bankers would be angered (by) to suggest devaluation.

We have now seen that, in contrast to gerunds, English verbal (finite and infinitival)
clauses occur neither in subject position nor as objects of Ps such as by, for and with,
whose selection feature is +___DP. The next section will show that verbal clauses do
not actually occur in direct object position either, counter to the impression given by
(2a-b). Only gerunds are in DP object positions.

5. Gerund objects of V vs. phrase-final (“extraposed”) verbal clauses

The syntactic distinctions between verbs that select gerund complements and those that
take verbal clause complements become evident only in more complex structures. Such
structures reveal that the seeming alternation between verbal clause complements and
gerunds in (2a-b) vs. (2c) is as illusory as their seemingly free alteration in (1).
Consider a verb like prefer, which can select a gerund, an infinitive or a finite
complement. The gerund is necessarily a direct object DP, and so can be followed by
another selected PP, which serves to compare the preference for the object DP. Similarly,
the idiom take upon oneself can select a gerund or an infinitive.
 Joseph E. Emonds 23

(23) a. Bill prefers that we ride bicycles / to ride a bicycle / that she ride a bicycle.
b. Bill preferred riding a bicycle [PP to endless hitchhiking].
I will take this responsibility [PP upon myself].
I will take repairing the roof [PP upon myself].

Since verbal clauses cannot be in the DP object position, they are necessarily in VP-final
position (= “extraposition”). Hence, unlike the gerunds in (23), they cannot precede a
selected PP complement.

(24) *Bill prefers that she ride a bicycle [PP to her hitchhiking].
*Bill preferred to ride a bicycle [PP to endless hitchhiking].
*I will take to repair the roof [PP upon myself].

Along the same lines, the verb report can have a finite clause or a gerund as an object,
and also an indirect object. If the object clause is finite, it must follow the indirect object
in extraposition and not be in DP position, whereas, as expected, an object gerund can
precede an indirect object.

(25) John reported having seen the fight to the police.

John reported to the police that he had seen the fight.
*John reported that he had seen the fight to the police.

The analyses of this study thus explain the contrasts in (23)-(25).9

A contrast between gerunds and verbal clauses, such as after verbs like prefer and
report, is otherwise not so widespread, which in itself suggests that the constructions
belong to different categories, i.e., DPs vs. IPs. There are other verbs whose object DPs
cannot be gerunds, even though these objects alternate with verbal clauses. For several
such verbs (tell, promise, teach), the direct and indirect objects can appear in either order.

(26) She told a fairy tale to the children.

She told the children a fairy tale.
You promised a new hat to Mary.
You promised Mary a new hat.
The man taught the importance of books to his sons.
The man taught his sons the importance of books.

However, Distribution of Verbal Clauses (18) predicts that a finite or infinitival verbal
clause complement cannot be a DP internal in a VP, but only at the end of the VP, in
“extraposition”:

9. Gerund objects can appear at the end of the VP by virtue of the “Heavy NP Shift” of Ross (1967), which
allows object DPs containing phrasal constituents such as VP to move to clause-final position.
24 From Theory to Practice 2013

(27) She told the children how to make a kite / that she was ill.
*She told how to make a kite / that she was ill to the children.
You promised Mary you would do the wash / to be quiet.
*You promised that you would do the wash / to be quiet to Mary.
The man taught his sons that books contain much wisdom.
*The man taught that books contain much wisdom to his sons.

Still other verbs (say, expect) can take verbal clauses but exclude gerunds as objects, and
require any indirect object DP to have an overt head P:

(28) They expect that you co-operate.

They expect some cooperation of (from) you.
John said that her family was dysfunctional.
John said something nasty to Mary.

When such verbs take both an indirect object and a verbal clause, the clause cannot be
in object DP position (before the indirect object), but must rather, by (18), be at the end
of the containing VP:

(29) They expect (it) of you that you cooperate.

*They expect that you cooperate of you.
John said to Mary that her family was dysfunctional.
*John said that her family was dysfunctional to Mary.

Finally, adjectival secondary predicates modify a direct object and must follow it inside
a verb phrase, as in (30b). If one tries to replace a direct object with a verbal clause
(i.e., to use an infinitive as subject of the secondary predicate), the result is clearly
ungrammatical (30c):

(30) a. Writing out / To write out their grievances is silly.

Complaining a lot / To complain a lot would be pretty useless.
b. The women consider writing out their grievances silly.
We found complaining a lot pretty useless.
c. *The women consider to write out their grievances silly.
*We found to complain a lot pretty useless.

This contrast is again explained by my claim that gerunds but not verbal clauses are
DPs.
Overall, we have seen that any selected complements of a verb can always
follow a gerund, because a direct object (the gerund) can always precede other
complement phrases. In contrast, verbal clause complements do not precede these same
 Jo se ph E. Emo n d s 25

gerunds are well-behaved DPs in standard positions of subjects and objects, while verbal
clause complements, when post-verbal, must be at the end o VPs.

are excluded, and that is possessive position modifying nouns. In line with our
expectations, the unacceptability of verbal clause infinitives in this DP position is
stronger than that of gerunds.

(31) a. ?Does he know about [DP [DP smoking pot’s] being illegal]?
*We agree about [ DP [DP shovelling snow’s] being fruitless].
b. **Does he know about [DP [VP to smoke pot’s] being illegal]?
**We agree about [ DP [VP to shovel snow’s] being fruitless].

From all that has been said so far, we might expectthe examples (31a) to be acceptable,
butthey are not. However, there is an independent reason for this that leaves
the generalization covering Gerund Distribution (5) fully intact.

possessive position, and one such restriction is that possessive DPs cannot contain any
type of event nominalization, whatever its internal form:

(32) *I don’t know the arrival in Chicago’s time.

*I found outthe cheapestticketto Chicago’s cost.
*Did you meetthat man teaching history’s cousin?

reason, the gerunds in (31a) are excluded, even though by (5) the structures themselves
10

6. Two further confirmations that only Gerunds have the distribution of

DPs

6.1 Coordination with lexical NPs

Generally speaking, categorical identity is a necessary condition for coordination
(Higgins 1973, 191), with the exception of some coordinated Adjective and Prepositional

10. It is of some interestthat more recent grammars prefer to drop the ending ’s on subjects of gerunds,
and this same choice improves the examples in (31a):
Does he know about [DP [DP smoking pot] being illegal]?
?We agree about [ DP [DP shovelling snow] being fruitless].
In fact, in some very recent usagein both Britain and the United States, overt subjects of gerunds
have entirely fallen out of use, though I have not seen documentation. I disregard this development
throughoutthis paper.
26 From Theory to Practice 2013

Phrases, e.g., Mary is clever and without financial worries. The fact that DPs with
lexical N heads freely coordinate with English gerunds in all DP positions confirms
the correctness of English Gerund Distribution (5):

(33) She always liked physical exercise and watching television shows.
The town proposed a tax increase and reviving the translation service.
Outdoor bathrooms and pitching a tent every day would wear us out.
The country aimed at self-sufficiency in fuel and discontinuing food imports.
Someone arranged for a new swimming pool and painting the house.

The same diagnostic shows that English infinitives and finite clauses are not DPs, again
confirming (18):

(34) *She always liked physical exercise and to watch television shows.
*The town proposed a tax increase and to revive the translation service.
*The town proposed a tax increase and that offices be moved to the outskirts.
*Outdoor bathrooms and to pitch a tent every day would wear us out.
*Outdoor bathrooms and that we couldn’t cook ourselves would irritate us.
*The country aimed at self-sufficiency in fuel and to discontinue food imports.
*Someone arranged for the house to be painted and a new swimming pool.

The different categorical status this paper accords to gerunds and infinitives also
accounts automatically for the inability of gerunds and verbal clauses to coordinate:

(35) *She always liked watching television shows and to play volleyball.
*Eating canned foods and to pitch a tent every day would irritate us.
*The town proposed reviving the translation service and that its offices be moved.
*The advisor suggested that we rent our house and selling some stocks.

These robust grammaticality distinctions remain mysterious in accounts that consider

gerunds to differ from verbal clauses only in “semantic nuance” (meaning) rather than
in categorical status (form). In contrast, these differences are explained by the structural
generalizations (5) and (18).
The coordination patterns just discussed indicate that coordinated verbal clauses are
simply not separate DPs at all, while coordination of gerunds can be coordinate DPs
(and hence lead to plural number agreement). A third possibility, given that a gerund is
some kind of (possibly extended) VP, is that coordinate gerunds can also form a single
(extended) VP under a single DP. These possibilities predict precisely three different
patterns of number agreement with coordinated clausal subjects, and these are exactly
what is observed.
 Joseph E. Emonds 27

(36) Coordinate verbal clause subjects, with a single, singular

null expletive DP subject:
[CP To paintthe house and to build a garage]
[IP [DP Ø] is / *are a year’s work.]
[CP
[IP [DP Ø] was / *were widely known.]
[CP
[IP [DP Ø] doesn’t / *don’t surprise me.]

(37) Gerund subjects, forming either coordinate DPs (plural agreement) or

coordinate VPs in a single DP (singular agreement):
[DP Painting the house and building a garage] is / are a year’s work.
[DP Her marrying a priest and her brother staying single]
was / were part of village lore.
[DP

expletive (always with singular agreement) and coordinate gerund subjects, with either

a DP, while no verbal clause can ever be.

6.2 The focus position in cleft sentences as a diagnostic for DP

constituents DP and PP.

(38) It’s the lemon pie that we disliked / talked about.

Was it John that you were speaking to?
Was it John that broke the window?
It was to John that she spoke.
It’s because of the flood / because it was raining thatthey are leaving.
It is with great pleasure that I present our speaker.

As seen in examples (39), this position excludes APs, VPs (including participles
introduced by V-ing) and CPs (with the complementizers that / for).

(39) *It’s very unhappy that Bill is / appeared.

*It was too quickly that she spoke.
*It’s quite dark that he likes his study.
*It is sell some buildings that you should (do).
*It was playing outside that Bill continued.
*It was ask John for money that I heard you.
28 From Theory to Practice 2013

*It was to report on time that we failed.

*It was that the guests left that John drank so much.
*It’s for the kids to be able to walk to school that they moved back.

Gerund Distribution (5) correctly predicts that gerunds appear freely in cleft focus
position:

(40) It was buying a new hat that he enjoyed.

It was John’s knowing the location that surprised her.
It’s driving carelessly that upsets me.
It was explaining your motives that was important.
It might be filing that report that Susan is concentrating on.
Was it her smoking cigars that John was struck by?

An interesting contrast with gerunds is furnished by the present participles that occur
after temporal aspect verbs (begin, start, continue, resume, keep, go on, cease, finish, etc.)
and verbs of perception (see, hear, feel, watch, notice, smell, catch, find, etc.). As shown
in Rosenbaum (1967), the V-ing complements of these verbs fail to act like DPs with
respect to passivization and coordination, and so are participles rather than gerunds.
Thus, unlike the examples (40), these participles are not DPs and thus cannot appear as
the focus in cleft sentences:

(41) *It is drinking beer from the bottle that she keeps.
*Is it painting the house that you’ve finished?
*It was throwing away some letters that John noticed Bill.
*It was stealing my money that she caught him.

The cleft sentence diagnostic for DPs demonstrates equally well that neither finite nor
infinitival Verbal Clauses are DPs, another correct consequence of (18):

(42) *It was to buy a new hat that he wanted.

*It’s for Mary to drive carelessly that upsets Ann.
*It is to always be on time that you should decide.
*It was that you explain your motives that was important.
*It’s that John has arrived too late that Bill realizes.
*Was it that Mary had cashed the check that Bill regretted?

In conclusion, all available constituency tests converge to show that English gerunds
behave as DPs, while English verbal clauses are not DPs.
 Joseph E. Emonds 29

7. Expletive Chains: interpreting clauses as arguments

Section 3 has shown that the structure for an initial verbal clause should be as in (8),
where this clause is co-indexed with (or “binds”) a “null expletive” in the main clause
subject position:

(8) [CP That Bill knows German well]j [IP [DP Ø]j [I was] [VP made obvious to all of us]].
[CP For the house to be painted]j [IP [DP Ø]j [I would] [VP confuse him]].

To motivate this structure, I discuss the nature of “expletives” in more general terms.
In a number of constructions, some XP in a clause which is not a subject (or an object)
can be interpreted as a “substitute” for that subject (or object). Thus, I propose here that
the bracketed CPs in (8) above and the underlined CPs in (43) below are neither DPs
nor structural subjects; the subjects are rather the null DPs inside IP, which are then
co-indexed with these CPs:

(43) [CP That the boys were dancing together]i [IP [DP Ø]i was amusing John].
[CP For John to arrive now]i [IP [DP Ø]i would cause embarrassment].
[CP That the children are always late]i [IP [DP Ø]i shows the necessity of discipline].
[CP That you spoke out of turn]i [IP [DP Ø]i didn’t help the situation].
[CP To suggest devaluation]i [IP [DP Ø]i would anger the bankers].

Thus, the underlined verbal clauses in (43) are interpreted with the semantic roles that
these predicates ordinarily assign to their subjects. For example, the underlined clause
in the last example has the same role as e.g. devaluation in Devaluation would anger
the bankers. An analysis with a null expletive is unavailable in traditional grammar,
because it never uses or even imagines “empty phrasal categories.” (Hence traditional
treatments are at a loss to explain the many differences between English infinitival and
gerund subjects).
(44) Expletive Constructions. A semantically null subject or object DP is an
“expletive.” An XP other than a subject or object co-indexed with an expletive is
called its associate.
(45) Interpretation of Associates. An associate is interpreted as having the
semantic role that a predicate can assign in the position of its expletive.
The only expletives in traditional grammar are overt but meaningless pronominals such
as it and there. Such expletives always bind (= are co-indexed with) lower phrases XP
in the same clause, which are, as in (44), called their associates. The overt expletive for
associate predicate nominals is there (46a), while the overt expletive for PP associates
(46b-c) is it.11

11. Some authors such as Andrews (1971) and Kallulli (1999) have argued that predicate nominals may
be only NPs, because they lack some properties of full DPs such as certain quantifiers. Perhaps a
30 From Theory to Practice 2013

(46) a. There was a warmer place in the hallway.

There will be many new students who lack rooms.
There has never occurred such a cold summer.
Today, there are a lot of people homeless.
b. It might be warmer in the hallway.
It is more comfortable by the fireplace.
c. The idea that it might be warmer in the hallway seemed silly.
Let’s go into the library because it is more comfortable by the fireplace.

Overt expletives are thus hierarchically distinct from the earlier null expletives in (8)
and (43), which are lower in a tree than their associates. But both types are alike in
being in subject position and assigning the semantic role of the subject of a verb to a
co-indexed associate XP.
In main clauses, the underlined associate PPs in (46b) can be sentence-initial
(“topicalized”) and can bind a null expletive in subject position, just like topicalized
verbal clauses in (8) and (43):

(47) [PP In the hallway]i [IP [DP Ø]j might be warmer].

[PP By the fireplace]i [IP [DP Ø]j would be a better location].

Just like topicalized clauses, pre-subject PP associates are excluded in dependent

clauses:

(48) *The idea that in the hallway might be warmer seemed silly.
*Let’s go into the library because by the fireplace is more comfortable.

This restriction on topicalized PPs is exactly the same restriction on pre-subject CP

associates that was observed in Section 2 with many examples in (7). Cf. note 8.
The following descriptive generalization seems adequate for describing the patterns
of expletives and associates.
(49) Associate-Expletive Binding (English):
(i) If an expletive asymmetrically binds (= c-commands) its associate, it must be
overt.
(ii) If an associate (e.g., topicalized) binds an expletive, the expletive is null.
In these terms, the “extraposed” clausal subjects underlined in (50) are the associates of
the subject expletive it, which binds them.

more general distinction is between case-marked arguments (subjects and objects) and all other XP
arguments, i.e., those which are not assigned case.
 Joseph E. Emonds 31

(50) It was amusing John that the boys were dancing together.
It would cause embarrassment for John to arrive now.
It shows the necessity of discipline that the children are always late.
It didn’t help the situation that you spoke out of turn.
It would anger the bankers to suggest devaluation.
It is obvious to all of us that Bill knows German well.
It is a waste oftime to read so many magazines.

also appear sentence-initially, as in examples (8) and (15a-b). In both positions, these
verbal clauses are linked with expletive subjects, and so have interpretations as subjects.
But unlike the ungrammatical topicalized clauses inside embedded sentences in (7i-vi),
extraposed clauses are well-formed in any type of embedded clause:

(51) I don’t believe (that) [σ it would help you for you to study history].
She forgets [CP how expensive [σ it is to go to the dentist]].
Nobody is ready for [σ itto be legally required that one exercise daily].
He protested the decision that [σ it counted for nothing for him to be on time].
A day atthe beach is more fun than [σ it is to play gol].
Although [σ it may depress you that the house is empty], it pleases me.
Situations in which [σ it is necessary to write out a check] should be avoided.
It was the salesman who [σ it seemed importantto for me to buy a car].

It also appears that empty expletive DPs can be generated in the object position and then
moved by passivization to the subject position, where an overt expletive it is inserted
to bind the extraposed verbal complement interpreted as a surface subject:

(52) It was said by John that we had betrayed him.

It was suggested to us that he was correct.
It was explained to her by several people how it should be done.

We will see more examples of null object expletives and verbal complement associates
in Section 8.

adjuncts, generated as rightmost daughters oflarger phrases.

(53) Interpretation of verbal clauses:
a. If verbal clauses are rightmost sisters of selecting head V, they are interpreted
not DPs.)12

12. With some verbs, fewer than is usually assumed, extraposed object clauses can be “doubled” by the
idiomatic expletive it in direct object position: She likes it a lot / doesn’t believe it at all that her brother
32 From Theory to Practice 2013

b. If verbal clauses are rightmost adjuncts of phrases, they are interpreted like
other adjuncts, such as PPs, etc.
c. If verbal clauses are co-indexed associates in any position, they are subject to
Interpetation of Associates (45).
A final detail is that as sisters to verbs, clausal complements, though in principle “heavy”
constituents that can be rightmost, should be able to precede adverbial phrase adjuncts
of a selecting verb. And indeed this seems to be the case:

(54) They plan to announce that we are capable of reaching Mars today.
It means nothing to speak of simultaneity in Einstein’s framework.
It isn’t required that the players be tall in this school.
It pleased me that they played those records very much.
It doesn’t frighten me to watch horror movies anymore.
It isn’t necessary to be smart on this campus.

The important generalization to be retained from this study is that unlike gerunds,
English verbal clauses are never structurally DPs. Though they can be interpreted as
subjects, as in (1a-b), they are in this use in topicalized pre-subject position. Therefore,
they can be neither the focus constituent in clefts like DPs, nor coordinate with DPs, nor
agree like DPs. Their interpretation as subjects, which has confused traditional analyses,
is due to the fact that they bind phonetically null expletive subjects of following
predicates.
English verbal clauses can similarly also be interpreted as direct objects, for the good
reason that both CPs and DPs can be selected as sisters to often the same verbs, such as
forget, hate, like, prefer, remember, take upon oneself, try, etc. But when verbal clauses are
selected by such verbs, they do not show the properties of DPs, unlike gerunds, which
do. For example, such verbal clauses do not passivize (Rosenbaum 1967; Stowell 1981).
When we realize that only root / main clauses tolerate pre-subject positioning of
verbal clauses, we are led to ask: are these constituents transformationally fronted
from either the position of the expletive or from an underlying extraposed position
of an associate? Such possibilities were in fact proposed in classical transformational
grammar, in which topicalized constituents are both selected and interpreted in clause-
internal argument positions.
But in fact there is no motivation for such movements. Since Chomsky’s (1981)
proposal of a Principle of Full Interpretation, constituents can be generated in their
surface positions, provided independently motivated principles can account for both
their positions and interpretation. In the present case, the principle of Interpretation of
Associates (45) accomplishes this.

walks to work. Rosenbaum (1967) took these expletives to mean that these extraposed clauses are moved
from an underlying direct object position, but there is no independent evidence for this account of
doubling.
 Joseph E. Emonds 33

In fact, there are verb classes that can take two arguments that are verbal clauses,
one a subject and one a complement. In such situations, the subject-interpreted verbal
clause can appear nowhere except at the beginning of the sentence:

(56) a. That John has blood on his hands proves (that) Mary is innocent.
*It proves (that) Mary is innocent that John has blood on his hands.
b. That John is late persuades me that the train was delayed.
*It persuades me that the train was delayed that John is late.
c. To see this movie is to relive the past.
*It is to relive the past to see this movie.

On the basis of this data, one must conclude that the sentence-initial verbal clauses of
(1a-b) are generated in (not moved to) this position, and linked to null expletives in the
subject position by Associate-Expletive Binding, thereby satisfying Full Interpretation.
Of course, as stated earlier, these pre-sentential verbal clauses are allowed only
in root constructions, as shown at length in Section 3. But nothing excludes certain
generative mechanisms from having effect only in root clauses. There are many such
constructions in the literature on root clauses, and the devices proposed for them in
Emonds (2012) work equally well whether they result from movement (e.g., topicalized
DP arguments) or from generation in situ.

8. Verbal Clauses limited to phrase-final (extraposed) position

This section will discuss several classes of English clausal complements which are not
DPs, but are rather in VP-final position (53a), after any direct or indirect objects of verbs.
Many of them are verbal clauses that do not even alternate with DPs. Here is a table of
such clausal complements:
(57) English Verbal Clause Complements (neither DPs nor associates of expletive
DPs)
Selecting verb type intransitive V transitive V: ___DP intransitive V with
+ verbal clause underlying empty
subjects
clause = present a. temporal aspect: d. perception verbs: g. participles “in
participles be, start, keep (on), see, hear, feel, find extraposition” (rare):
resume, finish, stop watch, notice, catch be fun / nasty / hellish
infinitives with b. decide, fail, e. cause, force, make, h. subjects raise to
empty subjects hesitate, manage, try oblige, persuade, tell main clauses: appear,
be likely, happen,
seem, turn out
finite clauses c. manner of f. answer, convince, i. expletive subjects:
speaking: growl, inform, persuade, appear, be likely,
mutter, quip, shout, remind, tell happen, seem, turn
whisper out

All these classes of complements have in common that they occur only where a V
selects them as a sister, i.e., at the end of VP. If the selecting verb is intransitive, these
34 From Theory to Pract ice 2013

complements can neither be passivized (showing they are not DPs) nor topicalized, nor

in (57a-b):

(58) a. Complement in selected position in VP:

My friend keeps staying out late.
Her husband finally finished writing that book.
A teacher rarely manages to satisfy a class.
Some residents tried to fix their plumbing.
b. Passivization fails:
*Staying out late is kept by my friend.
*Writing that book was finally finished by her husband.
*To satisfy a class is rarely managed by a teacher.
*To fix the plumbing was tried by some residents.
c. Topicalization fails:
*Staying out late my friend keeps.
*Writing that book her husband finally finished.
*To satisfy a class a teacher rarely manages.
*To fix their plumbing some residents tried.

*It’s staying out late that my friend keeps.

*It was writing that book that her husband finally finished.
*It’s to satisfy a class that a teacher rarely manages.
*It was to fix their plumbing that some residents tried.

of speaking” verbs. Zwicky (1971) gives examples and properties of this class of
complements.13

(59) Morris whispered that night was falling.

Some kid shrieked that Manchester had scored two goals.
One guest growled to the waiter that an hour was too long.
John quipped that she could pass without trying.

Zwicky’s descriptive generalization is thatthese clausal complements never act like

DPs. Rather, like other complements in Table (57), these clausal complements are final
in the VP and so always follow an indirect object PP if there is one.

13. In discussing these verbs, Zwicky adds that they are “understood communicatively.” Infinitival
complements of manner of speaking verbs are interpreted as embedded imperatives: A guest whispered
to Harry to move over; A bystander shouted to shoot the assailant
in Emonds (2000, Ch. 9).
 Joseph E. Emonds 35

(60) Clausal complements of manner of speaking verbs cannot undergo passive:

*That night was falling was whispered by Morris.
*That Manchester had scored two goals was shrieked by some kid.
*That an hour was too long was growled to the waiter by one guest.
*That she could pass without trying was quipped by John.

(61) Nor can passivization apply to an expletive in object position that might be
linked to them:
*It was whispered by Morris that night was falling.
*It was shrieked by some kid that Manchester had scored two goals.
*It was growled to the waiter by one guest that an hour was too long.
*It was quipped by John that she could pass without trying.

(62) Nor can they be topicalized:

*That night was falling Morris whispered.
*That Manchester had scored two goals some kid shrieked.
*That an hour was too long one guest growled to the waiter.
*That she could pass without trying John quipped.

(63) Nor can they be focused in cleft sentences:

*It was that night was falling that Morris whispered.
*It was that Manchester had scored two goals that some kid shrieked.
*It was that an hour was too long that the guest growled.
*It was that she could pass without trying that John quipped.

Manner of speaking verbs contrast with many epistemic and emotive verbs, whose
finite complements bind phonetically null expletives in object position (49ii): admit,
believe, conclude, fear, know, realize, regret, understand, etc. (Recall that expletives are
semantically null.)
(64) Morris admitted / believed [DP Ø] that night was falling.
Some fans concluded / feared [DP Ø] that Manchester had scored two goals.
One guest knew / realized [DP Ø] that an hour was too long.
John regretted / understood [DP Ø] that she could pass without trying.

These expletives can move to subject position by passivization, where they

asymmetrically bind their verbal complement associate in extrapositon, and hence
must be overt (it), as (49i) requires:

(65) It was admitted / believed by Morris [DP Ø] that night was falling.
It was concluded / feared [DP Ø] by some fans that Manchester had scored two goals.
It was known / realized [DP Ø] by one guest that an hour was too long.
It was regretted / understood [DP Ø] that she could pass without trying.
36 From Theory to Practice 2013

Alternatively, if the verbal complements of these transitive verbs are in pre-subject

position, as is allowed in main clauses, they asymmetrically bind the expletives in object
position, which must then remain null, by (49ii):

(66) That night might fall soon Morris had to admit / refused to believe [DP Ø].
That Manchester had gone ahead many fans already concluded / feared [DP Ø].
That an hour was too long at least one guest already knew / realized [DP Ø].
That she could pass without trying John really regretted / fully understood [DP Ø].

It can be observed that pragmatic felicity of such topicalized object clauses often
requires various kinds of adverbs, modals, etc. Elucidating such requirements is beyond
the scope of this paper.
It would be tedious to repeat the above diagnostics with the six further classes of
complements given in (57d-i) in the Table. The topicalization and cleft focus tests applied
to all these subclasses of clausal complements fully replicate the results above, namely
that none of these verbal clause complements pass these DP diagnostics. Already in
Rosenbaum (1967), some of these tests are applied to some of the verb classes in (57).
One can easily get the impression from early generative literature that selection of
complement clauses is somehow always related to their being or at least being like
noun phrases, both in distribution and in interpretation. However, many different verb
classes select clausal complements which are in no way DPS, and a fortiori these verbs
cannot select DP gerunds. We can verify that many types of verbs in (57) that select only
verbal clauses are incompatible with gerund objects, exactly as Gerund Distribution (5)
predicts:

(67) *The teachers failed satisfying their classes.

*We hesitated flying by way of Chicago.
*Morris whispered night’s being about to fall.
*Some kid shrieked Manchester’s having scored two goals.
*A guest growled an hour’s being too long to the waiter.
*John quipped her being able to pass without trying.

I thus conclude that English finite and infinitival clausal complements can be selected
by the clausal subcategorization frames ___CP, ___IP and ___VP, and that these verbal
clause complements have nothing to do with the syntax or selection of the category DP.
These verbal clauses undergo movement neither in the clause in which they originate
nor out of that clause. Nor are they in any way noun phrases (DPs). As opposed to
clauses interpreted via expletives in DP argument positions, seen in (64), all the clausal
complements selected by the verb classes in Table (57) appear only in the clause-final
position where they are lexically selected and not elsewhere.
Although there is no evidence against concluding that many types of verbal
clauses are neither DPs nor co-indexed with expletives, a persistent tendency in both
 Joseph E. Emonds 37

generative and traditional grammar has been to assume that every clausal complement
is somehow a DP (noun phrase) or associated with one. This tendency has to do with
a longstanding, conscious or unconscious identification of clausal embedding with a
notion of embedded predication in symbolic logic. The “arguments” of predicates in
logic are felt to “correspond” to noun phrases (DPs), so that if some proposition (a
“verbal clause”) is an argument of a higher predicate, then that clause must also be
a DP. As a non-metaphorical justification for a grammatical analysis, such thinking is
baseless, as shown by the properties of the array of constructions in Table (57).

Acknowledgement
I thank Ludmila Veselovská and Kateřina Havranová for careful readings of the text,
and many helpful suggestions.

Works Cited
Andrews, Avery. 1971. “Case Agreement of Predicate Modifiers in Ancient Greek.”
Linguistic Inquiry 2: 127–51.
Chomsky, Noam. 1981. Lectures on Government and Binding: The Pisa Lectures.
Dordrecht: Foris.
Chomsky, Noam. 1986. Barriers. Cambridge: MIT Press.
Emonds, Joseph E. 1976. A Transformational Approach to English Syntax: Root,
Structure-preserving, and Local Transformations. New York: Academic Press.
Emonds, Joseph E. 1985. A Unified Theory of Syntactic Categories. Dordrecht: Foris.
Emonds, Joseph E. 2000. Lexicon and Grammar: The English Syntacticon. Berlin:
Mouton de Gruyter.
Emonds, Joseph E. 2012. “Augmented Structure Preservation and the Tensed S
Constraint.” In Main Clause Phenomena: New Horizons, edited by Aelbrecht Loebke,
Liliane Haegeman, and Rachel Nye, 21–46. Amsterdam: John Benjamins.
Haegeman, Liliane. 2009. “The Movement Analysis of Temporal Adverbial Clauses.”
English Language and Linguistics 13: 385–408.
Haegeman, Liliane. 2011. “The Movement Derivation of Conditional Clauses.”
Linguistic Inquiry 41: 595–621.
Higgins, Francis. 1973. “The Pseudo-cleft Construction in English.” PhD diss.,
Massachusetts Institute of Technology, Cambridge, MA.
Kallulli, Dalina 1999. “The Comparative Syntax of Albanian: On the Contribution of
Syntactic Types to Propositional Interpretation.” PhD diss., University of Durham.
Rizzi, Luigi. 1997. “The Fine Structure of the Left Periphery.” In Elements of Grammar,
edited by Liliane Haegeman, 281–337. Dordrecht: Kluwer.
Rosenbaum, Peter S. 1967. The Grammar of English Predicate Complement
Constructions. Cambridge, MA: MIT Press.
Ross, John R. 1967. “Constraints on Variables in Syntax.” PhD diss., Massachusetts
Institute of Technology, Cambridge, MA.
 38 From Theory to Practice 2013

Stowell, Timothy A. 1981. “The Origins of Phrase Structure.” PhD diss., Massachusetts
Institute of Technology, Cambridge, MA.
Zwicky, Arnold. 1971. “In a Manner of Speaking.” Linguistic Inquiry 2: 223–33.

View publication stats