The series of volumes of Syntax of Dutch presents a synthesis of Comprehensive

Grammar Comprehensive Grammar Resources

currently available syntactic knowledge of Dutch. The principal Resources
concern of the volumes is with the description of the language, not
Henk van Riemsdijk & István Kenesei, series editors
with linguistic theory. Syntax of Dutch will be an indispensable resource
for researchers and advanced students of languages and linguistics.

The three volumes of Verbs and Verb Phrases discuss the internal

Syntax of
make-up and distribution of verb phrases and clauses. After a general
introduction covering various issues including tense marking, Volume 1
provides an extensive discussion of argument structure and verb frame

Dutch
alternations. Volume 2 is devoted to various types of verbal/clausal
complements in complex clauses. Volume 3 contains discussions of
adverbial modification and the organization (word order) of the clause.

Syntax of
Hans Broekhuis is a researcher at the Meertens Institute in

Dutch
Amsterdam, Norbert Corver is professor of Dutch Linguistics
at the University of Utrecht.

“reliance
The work is agreeably language- and theory-independent while its
Verbs and
Volume 2
Verb Phrases
Verbs and
on the solid basement of theoretical and empirical advances
in generative linguistics is palpable throughout the pages. The authors
manage to harmonize the demands of depth and breadth, and they draw
reasonable demarcation lines around the relevant domains they choose Verb Phrases
Volume 2
to describe. [...] The Syntax of Dutch project [..] has definitely become a
model for comprehensive grammatical description [...].
”
Gábor Alberti, University of Pécs, Department of Linguistics, and
Judit Farkas, Hungarian Academy of Sciences, Research Institute for Linguistics

Broekhuis
Corver
Hans Broekhuis
Norbert Corver

AUP.nl
Syntax of Dutch
Verbs and Verb Phrases
Volume 2
Comprehensive Grammar Resources

Editors:
Henk van Riemsdijk
István Kenesei
Syntax of Dutch
Verbs and Verb Phrases
Volume 2

Hans Broekhuis
Norbert Corver

With the cooperation of:

Hans Bennis
Frits Beukema
Carole Boster
Crit Cremers
Henk van Riemsdijk

Amsterdam University Press

The publication of this book is made possible by grants and financial support from:
Netherlands Organisation for Scientific Research (NWO)
Center for Language Studies
University of Tilburg
Truus und Gerrit van Riemsdijk-Stiftung
Meertens Institute (KNAW)

This book is published in print and online through the online OAPEN library
(www.oapen.org).

 
 
 
 
 
 
 
 
 
 
 
Cover design: Studio Jan de Boer, Amsterdam
Layout: Hans Broekhuis
 
ISBN 978 90 8964 731 3
e-ISBN 978 90 4852 483 9 (pdf)
NUR 616 / 624
 
 
 
 

 
Creative Commons License CC BY NC
(http://creativecommons.org/licenses/by-nc/3.0)
 
 Hans Broekhuis/Amsterdam University Press, Amsterdam 2015 
 
Some rights reserved. Without limiting the rights under copyright reserved above,
any part of this book may be reproduced, stored in or introduced into a retrieval
system, or transmitted, in any form or by any means (electronic, mechanical,
photocopying, recording or otherwise).
Contents
Volume 1
Abbreviations and symbols v

Preface and acknowledgments vii

1. General introduction vii
2. Main objective vii
3. Intended readership vii
4. Object of description viii
5. Organization of the material xiii
6. History of the project and future prospects xvii
7. Acknowledgments xix

Introduction 1

Chapter 1
Characterization and classification 13
Introduction 14
1.1. General characterization 14
1.2. Verb classifications 19
1.3. Inflection 62
1.4. Mood 79
1.5. Tense, epistemic modality and aspect 102

Chapter 2
Projection of verb phrases I: Argument structure 181
Introduction 182
2.1. Nominal arguments 185
2.2. Complementives (secondary predicates) 239
2.3. PP-complements (prepositional objects) 284
2.4. AP-complements 329
2.5. Special verbs 332
2.6. Bibliographical notes 397

Chapter 3
Projection of verb phrases II: Verb frame alternations 401
Introduction 402
3.1. Main types 402
3.2. Alternations involving the external argument 407
3.3. Alternations of noun phrases and PPs 514
3.4. Some apparent cases of verb frame alternation 591
3.5. Bibliographical notes 595
Volume 2
Chapter 4
Projection of verb phrases IIIa: Selection of clauses/verb phrases 597
Introduction 598
4.1. Semantic types of finite argument clauses 601
4.2. Finite and infinitival argument clauses 604
4.3. Control properties of verbs selecting an infinitival clause 608
4.4. Three main types of infinitival argument clauses 610
4.5. Non-main verbs 624
4.6. The distinction between main and non-main verbs 630

Chapter 5
Projection of verb phrases IIIb: Argument and complementive clauses 639
Introduction 640
5.1. Finite argument clauses 641
5.2. Infinitival argument clauses 765
5.3. Complementive clauses 936

Chapter 6
Projection of verb phrases IIIc: Complements of non-main verbs 945
Introduction 946
6.1. Characteristics and typology of non-main verbs 946
6.2. Non-main verbs selecting a participle 951
6.3. Non-main verbs selecting a te-infinitive 1003
6.4. Non-main verbs selecting a bare infinitive 1019

Chapter 7
Projection of verb phrases IIId: Verb clusters 1049
Introduction 1050
7.1. Recognizing verb clusters 1051
7.2. The hierarchical order of verbs in verb clusters 1062
7.3. The linear order of verbs in verb clusters 1091
7.4. Permeation of verb clusters 1112
7.5. Bibliographical notes 1117

Glossary (Volumes 1 & 2) 1119

Subject index (Volumes 1 & 2) 1153
References (Volumes 1 & 2) 1173
Volume 3 (may be subject to change)

Chapter 8
Projection of verb phrases IV: adverbial modification
Introduction
8.1. Semantic types of adverbial modifiers
8.2. Categorial types of adverbial modifiers
8.3. The order of adverbial modifiers
8.4. Bibliographical notes

Chapter 9
Word order in the clause I: General introduction
Introduction
9.1. The overall organization of the clause
9.2. The position of the verbs
9.3. The clause-initial position
9.4. The postverbal field
9.5. The middle field
9.6. Conclusion

Chapter 10
Word order in the clause II: Placement of the finite verb
Introduction
10.1. Placement of the finite verb
10.2. The behavior of X+V collocations under verb-first/second
10.3. Special cases

Chapter 11
Word order in the clause III: Clause-initial position
Introduction
11.1. The formation of V1- and V2-clauses
11.2. Clause-initial position remains (phonetically) empty
11.3. Clause-initial position is filled

Chapter 12
Word order in the clause IV: Postverbal positions
Introduction
12.1. Arguments
12.2. Modifiers
12.3. Parts of constituents
Chapter 13
Word order in the clause V: The middle field of the clause
Introduction
13.1. The global architecture of the middle field
13.2. Nominal argument shift
13.3. Focus movement
13.4. Negation Movement
13.5. Movement of weak proforms

Chapter 14
Main-clause external elements
Introduction
14.1. Left dislocation
14.2. Right dislocation
14.3. Afterthoughts
14.4. Vocatives
14.5. Discourse particles

Glossary (All Syntax of Dutch volumes)

Subject index (Verbs and verb phrases 1-3)
References (Verbs and verb phrases 1-3)
Abbreviations and symbols

This appendix contains a list of abbreviations and symbols that are used in this
volume. Sometimes, conventions are adopted that differ from the ones given in this
list, but if this is the case this is always explicitly mentioned in the text.

References to the other volumes of the Syntax of Dutch.

References to the chapters and sections to the other volume in the series Syntax of
Dutch are preceded by a letter: N + section # refers to the two volumes on nouns
and noun phrases, A + section # refers to the volume on Adjectives and adjective
Phrases, and P+section # refers to the volume on Adpositions and adpositional
phrases. For example, refers to Section P3.2. in Hans Broekhuis (2013). Syntax of
Dutch: Adpositions and adpositional phrases. Amsterdam: AUP.

Symbols and abbreviation used in the main text

°xxx refers to the XXX glossary
Domain D Domain of discourse

Abbreviations used in both the main text and the examples

AP Adjectival Phrase
CP Complementizer Phrase
DP Determiner phrase
NP Noun Phrase
Noun phrase used when the NP-DP distinction is not relevant
NumP Numeral Phrase
PP Prepositional Phrase
PO-verb Verb with a prepositional object
QP Quantifier Phrase
TP Tense Phrase
VP Verb Phrase
Aux2-Main1 Verb cluster. The numeral indices indicate the hierarchical order of
the verbs: Vn+m is superior to Vn. the en-dash indicates linear
order: the element to the left precedes the element to the right in
the surface order of the sentence: see Section 7.2, sub I, for details.

Symbols, Abbreviations and conventions used in the examples

e Phonetically empty element
Ref Referent argument (external °thematic role of nouns/adjectives)
Rel Related argument (internal thematic role of relational nouns)
OP Empty operator
PG Parasitic gap
PRO Implied subject in, e.g., infinitival clauses
PROarb Implied subject PRO with arbitrary (generic) reference
t Trace (the original position of a moved element)
XXX Small caps indicates that XXX is assigned contrastive accent
Abbreviations used as subscripts in the examples
1p/2p/3p 1st, 2nd, 3rd person pl Plural
acc Accusative poss Possessor
dat Dative pred Predicate
ben Beneficiary rec Recipient
nom Nominative sg Singular
Abbreviations used in the glosses of the examples
AFF Affirmative marker
COMP Complementizer: dat ‘that’ in finite declarative clauses, of ‘whether/if’
in finite interrogative clauses, and om in infinitival clauses
prt. Particle that combines with a particle verb
PRT Particle of different kinds
REFL The short form of the reflexive pronoun, e.g., zich; the long form
zichzelf is usually translated as himself/herself/itself
XXX Small caps in other cases indicates that XXX cannot be translated
Diacritics used for indicating acceptability judgments
* Unacceptable
*?
Relatively acceptable compared to *
??
Intermediate or unclear status
?
Marked: not completely acceptable or disfavored form
(?)
Slightly marked, but probably acceptable
no marking Fully acceptable
%
Varying judgments among speakers
#
Unacceptable under intended reading
$
Special status: old-fashioned, archaic, very formal, semantically
incoherent, degraded/unacceptable for non-syntactic reasons, etc. The
nature of the deviation is normally explained in the main text.
Other conventions
xx/yy Acceptable both with xx and with yy
*xx/yy Unacceptable with xx, but acceptable with yy
xx/*yy Acceptable with xx, but unacceptable with yy
(xx) Acceptable both with and without xx
*(xx) Acceptable with, but unacceptable without xx
(*xx) Acceptable without, but unacceptable with xx
.. <xx> Alternative placement of xx in an example
.. <*xx> .. Impossible placement of xx in an example
 Necessarily implies
 Does not necessarily imply
XX ... YY Italics indicate binding
XXi ... YYi Coindexing indicates coreference
XXi ... YYj Counter-indexing indicates disjoint reference
XX*i/j Unacceptable with index i, acceptable with index j
XXi/*j Unacceptable with index j, acceptable with index i
[XP ... ] Constituent brackets of a constituent XP
Chapter 4 Projection of verb phrases IIIa:
Selection of clauses/verb phrases

Introduction 598
4.1. Semantic types of finite argument clauses 601
4.2. Finite and infinitival argument clauses 604
4.3. Control properties of verbs selecting an infinitival clause 608
4.4. Three main types of infinitival argument clauses 610
4.4.1. Om + te-infinitivals 610
4.4.2. Bare infinitivals 612
4.4.3. Te-infinitivals 619
4.5. Non-main verbs 624
4.6. The distinction between main and non-main verbs 630
598 Syntax of Dutch: Verbs and verb phrases

Introduction
Chapter 5 to Chapter 7 will discuss constructions in which a verb select a verbal
projection, that is, a clause or some smaller (extended) °projection of some other
verb. The present chapter provides the necessary background for these chapters by
providing a review of a number of issues in this domain. In a sense, Chapter 5 can
be seen as a continuation of the discussion in Chapter 2 on °argument structure: it
discusses cases in which main verbs select a verbal projection, that is, a finite or
infinitival argument clause. The reason why we did not discuss this type of
°complementation in Chapter 2 is that in essence it does not alter the syntactic verb
classification that was developed there: for example, many verbs taking an internal
°argument have the option of choosing between a nominal and a clausal
complement. This is illustrated in (1) for the transitive verb zien ‘to see’ and the
ditransitive verb vertellen ‘to tell’.
(1) a. Jan zag het schilderij.
Jan saw the painting
a. Jan zag [dat het regende].
Jan saw that it rained
b. Peter vertelde Marie een leuk verhaal.
Peter told Marie a nice story
b. Peter vertelde Marie [dat Jan ziek was].
Peter told Marie that Jan ill was
‘Peter told Marie that Jan was ill.’
If a specific verb resists a nominal object, pronominalization of the clausal
complement shows that the verb in question is in principle able to take a nominal
complement and to assign case to it. The acceptability of pronominalization in (2b),
for instance, shows that the verb betogen ‘to argue’ is simply a transitive verb and
that the infelicitousness of the use of the nominal object die stelling ‘that thesis’ is a
matter of semantics, not syntax; complements of verbs like betogen must simply
have propositional content. This is confirmed by the fact illustrated in (2c) that
there are acceptable cases of nominal complementation with noun phrases like het
tegendeel ‘the opposite’, which are propositional in nature.
(2) a. Jan betoogt [dat dit boek een mislukking is].
Jan argues that this book a failure is
‘Jan argues that this book is a failure.’
b. Jan betoogt dat/$die stelling.
Jan argues that/that thesis
c. Els zegt [dat dit boek een meesterwerk is] maar Jan betoogt het tegendeel.
Els says that this book a masterpiece is but Jan argues the opposite
‘Els says that this book is a masterpiece but Jan argues the opposite.’
The examples in (3) show that clausal complements of PO-verbs can normally also
be pronominalized or be replaced by a noun phrase. This illustrates again that
clausal complements do not essentially affect the syntactic classification of verbs,
and that the omission of clausal complements from our discussion of argument
structure in Chapter 2 is therefore relatively innocuous.
 Selection of clauses/verb phrases 599

(3) a. Jan twijfelt (erover) [of hij de juiste beslissing genomen heeft].
Jan doubts about.it whether he the right decision taken has
‘Jan isnʼt sure (about it) whether he has taken the right decision.’
b. Jan twijfelt daarover/over zijn beslissing.
Jan doubts about.it/about his decision
‘Jan isnʼt sure about that/about his decision.’
The reason for devoting a separate discussion to clausal/verbal arguments is that
these arguments exhibit various special properties and introduce a number of
complicating factors that have been investigated extensively in the literature. A
discussion of these special properties and complicating factors would seriously
interfere with the main line of argumentation in Chapter 2: it is better to discuss
these properties in their own right. The present chapter will point at some of the
topics that need special attention.
After having read the general discussion in this chapter, the reader will be
sufficiently equipped to read the next three chapters, which we briefly review here
for convenience. Chapter 5 starts by showing that main verbs can take a number of
different types of clausal/verbal arguments: the examples in (4) show that such
argument clauses may be finite or infinitival: finite argument clauses are discussed
in Section 5.1 and the various types of infinitival clauses in Section 5.2.
(4) a. Jan vertelde me dat Marie in Utrecht woont. [finite]
Jan told me that Marie in Utrecht lives
‘Jan told me that Marie lives in Utrecht.’
b. Jan verzocht me om naar Amsterdam te komen. [infinitival]
Jan asked me COMP to Amsterdam to come
‘Jan asked me to come to Amsterdam.’
Section 5.3 concludes Chapter 5 by investigating whether finite and infinitival
clauses can function as °complementives in copular and vinden-constructions.
Examples such as (5a) seem to point in this direction but the fact that such
examples occur alongside examples such as (5b), in which the finite clause clearly
functions as the subject of the construction, shows that this cannot be taken for
granted.
(5) a. Een feit is [dat hij te lui is].
a fact is that he too lazy is
‘A fact is that heʼs too lazy.’
b. Het is een feit [dat hij te lui is].
it is a fact that he too lazy is
‘It is a fact that heʼs too lazy.’
Chapter 6 discusses the various types of verbal complements of non-main verbs.
Although such complements do not function as arguments in the sense of predicate
calculus, they can still be said to be selected by the non-main verbs: the examples in
(6) show that perfect auxiliaries like hebben ‘to have’ select past participles,
whereas aspectual verbs like gaan ‘to go’ select infinitives.
600 Syntax of Dutch: Verbs and verb phrases

(6) a. Jan heeft dat boek gelezen.

Jan has that book read
‘Jan has read that book.’
b. Jan gaat dat boek lezen.
Jan goes that book read
‘Jan is going to read that book.’
Constructions with embedded non-finite clauses/verbal projections may exhibit
monoclausal behavior in the sense that the °matrix verb (that is, the verb that selects
the clause/verbal projection and thus heads the matrix clause) and the verb heading
the non-finite complement form a °verb cluster, that is, a more or less impermeable
sequence of verbs. This may give rise to what we will refer to as CLAUSE
SPLITTING; the infinitival clause becomes discontinuous in the sense that the matrix
verb separates the infinitival verb from its dependents (like arguments and
modifiers). The phenomenon of verb clustering (which is often referred to as VERB
RAISING in the formal linguistic literature) and concomitant clause splitting is
illustrated in (7a): the verb zien ‘to see’ selects the infinitival complement Peter dat
boek lezen, which surfaces as a discontinuous phrase due to clustering of the verbs
zien ‘to see’ and lezen ‘to read’. Example (7b) has been added to show that verb
clustering is often obscured in main clauses because they require movement of the
finite verb into second position; see Section 9.2 for discussion.
(7) a. dat Jan Peter dat boek ziet lezen.
that Jan Peter that book sees read
‘that Jan sees Peter read that book.’
b. Jan ziet Peter dat boek lezen.
Jan sees Peter that book read
‘Jan sees Peter read that book.’
Constructions with non-main verbs typically exhibit monoclausal behavior; they
always involve verb clustering, as shown in (8) by the embedded counterparts of the
examples in (6).
(8) a. dat Jan dat boek heeft gelezen.
that Jan that book has read
‘that Jan has read that book.’
b. dat Jan dat boek gaat lezen.
that Jan that book goes read
‘that Jan is going to read that book.’
Chapter 7 is devoted especially to verb clustering given that this is a recurring topic
in the literature. The reader must be aware, however, that it is often not a priori
clear what counts as a case of verb clustering. This is due to the facts listed in (9),
which are established in the sections indicated; we refer the reader to these sections
for detailed discussion.
 Selection of clauses/verb phrases 601

(9) a. Projections headed by a participle are not only used as verbal complements
of auxiliaries but can also be used as adjectival complementives; see A9.
b. Projections headed by an infinitive are not only used as infinitival clauses,
but can also be used as (i) adjectival complementives (this holds especially
for te-infinitives; see A9) or (ii) nominalizations (this holds especially for
bare infinitives; see N1.3.1.2 and N2.2.3.2).
The facts in (9) appear not always to have been taken into account in the existing
literature, which has led to confusion and, what is worse, an inaccurate and
unnecessarily complex empirical description of verb clustering. In order to avoid
this here, Chapter 5 will also discuss the disputable cases of verb clustering, which
we will subsequently eliminate these from the discussion, so that Chapter 7 can
focus on the true cases of verb clustering and formulate a small number of relatively
simple and, in our view, descriptively adequate generalizations.

4.1. Semantic types of finite argument clauses

The examples in (10) show that finite verbal argument clauses come in at least two
different forms, and that the choice between the two is largely dependent on the
matrix verb: the verbs zeggen ‘to say’ and vragen ‘to ask’ differ in that the former
takes declarative clauses as its complement, whereas the latter takes interrogative
clauses (that is, yes/no- or wh-questions) as its complement.
(10) a. Jan zegt [dat/*of Peter ziek is]. [declarative clause]
Jan says that/whether Peter ill is
‘Jan says that Peter is ill.’
b. Jan vraagt [of/*dat Peter ziek is]. [yes/no-question]
Jan asks whether/that Peter ill is
‘Jan asks whether Peter is ill.’
b. Jan vraagt [wie er ziek is]. [wh-question]
Jan asks who there ill is
‘Jan asks who is ill.’
Although we occasionally find similar differences in the domain of nominal
complementation (cf. Jan stelde een vraag/*antwoord ‘Jan asked a question’ versus
Jan gaf een antwoord/*vraag ‘Jan gave an answer’), this distinction is quite basic
when it comes to complementation by finite clauses.
Since Grimshaw (1979) it has often been claimed that verbs are subcategorized
for specific semantic types of complement clauses: embedded declarative clauses
such as (10a) are of the type “proposition” and embedded questions are of the type
“interrogative”. Grimshaw adds the type of “wh-exclamative”, which is found in the
examples in (11); the wh-phrases in these examples are not interrogative but express
“high degree” modification, just as in the exclamative main clauses given in the
primed examples. Observe that there are a number of differences between the main
and embedded clause (e.g. concerning word order and the form of the wh-word),
which we will ignore for the moment, but to which we will return in Section 11.3.5.
602 Syntax of Dutch: Verbs and verb phrases

(11) a. Ik was vergeten wat een ontzettend aardige vrouw Marie is. [exclamative]
I was forgotten what a very nice woman Marie is
‘Iʼd forgotten what a very nice woman Marie is.’
a. Wat is Marie een ontzettend aardige vrouw!
what is Marie a very nice woman
‘What a very nice woman Marie is!’
b. Ik was vergeten hoe ontzettend aardig Marie is. [exclamative]
I was forgotten hoe very nice Marie is
‘Iʼd forgotten how very nice Marie is.’
b. Wat is Marie ontzettend aardig!
what is Marie very nice
‘How very nice Marie is!’
The fact that Grimshaw (1979) includes exclamatives suggests that the list of
semantic types is open-ended in the sense that it would be possible to add more
semantic types to it; so it seems desirable to restrict it by imposing principled
constraints on the set of possible types. An attempt to do this can be found in Nye
(2013), who proposes that complement clauses are selected on the basis of two
binary features: [±WH] and [±FACTIVE]. These features characterize the four
different constructions in (12) provided we adopt the following definition of
factivity: FACTIVITY refers to constructions with verbs which take a complement
clause, and where the speaker presupposes the truth of some proposition expressed
by that clause; see Kiparsky & Kiparsky (1970) in the interpretation of Broekhuis &
Nye (2013). In the (a)-examples the relevant proposition is expressed by the full
complement clause, whereas in the (b)-examples it is expressed by the non-wh part
of the complement clause. For the two types of wh-questions, see also Groenendijk
& Stokhof (1984:91ff.) who define the distinction in terms of pragmatic
implicatures, that is, the speaker’s presupposition instead of factivity.
(12) a. Jan denkt dat Els morgen vertrekt.   Els vertrekt morgen.
Jan thinks that Els tomorrow leaves Els leaves tomorrow
‘Jan thinks that Els is leaving tomorrow.   Els is leaving tomorrow.’
a. Jan betreurt dat Els morgen vertrekt.  Els vertrekt morgen.
Jan regrets that Els tomorrow leaves Els leaves tomorrow
‘Jan thinks that Els is leaving tomorrow.  Els is leaving tomorrow.’
b. Jan vroeg wie er vertrekt.   Er vertrekt iemand.
Jan asked who there leaves there leaves someone
‘Jan asked who is leaving.   someone is leaving.’
b. Jan weet wie er vertrekt.  Er vertrekt iemand.
Jan knows who there leaves there leaves someone
‘Jan knows who is leaving.  someone is leaving.’
The binary feature approach thus gives rise to the four construction types in Table
1, which now includes the new class of factive interrogatives illustrated in (12b).
 Selection of clauses/verb phrases 603

Table 1: Complement clause selection

[-WH] [+WH]
[-FACTIVE] non-factive declarative (12a) non-factive interrogative (12b)
[+FACTIVE] factive declarative (12a) factive interrogative (12b)
wh-exclamative (11)

Another advantage of adopting the binary features [±WH] and [±FACTIVE] is that
they enable us to account for the fact that betreuren ‘to regret’ and weten ‘to know’
impose different selection restrictions on their complement; the unacceptability of
(13a) shows that the verb betreuren is only compatible with declarative clauses,
whereas the acceptability of (13b) shows that weten is compatible both with
declarative and with interrogative clauses. This can be expressed by assuming that
betreuren selects a [-WH,+FACTIVE] complement clause, but that weten does not
impose restrictions on the [wh]-feature and thus simply selects a [+FACTIVE]
complement clause. Providing a similar account in a non-ad hoc fashion seems
harder if we adopt Grimshaw’s claim that verbs select semantic types like
proposition, interrogative or exclamative.
(13) a. *Jan betreurt wanneer Els vertrekt. [cf. example (12a)]
Jan regrets when Els leaves
b. Jan weet dat Els morgen vertrekt. [cf. example (12b)]
Jan knows that Els tomorrow leaves
‘Jan knows that Els is leaving tomorrow.’
Note in passing that examples like Ik betreur [wat je hier schrijft] ‘I regret what
you write here’ are not relevant in this context: the bracketed part is a free relative,
therefore we are dealing with a nominal complement and not a complement clause.
In a similar way, we might account for the fact that verbs like betwijfelen ‘to
doubt’ in (14) can be combined with an embedded yes/no-question, but not with an
embedded wh-question by claiming that its interrogative complement clause must
be [-FACTIVE]—although it should be noted that this still leaves open why the
embedded wh-question in (14) cannot be interpreted as non-factive. Again,
providing a similar account is not possible under Grimshaw’s proposal where
yes/no- and wh-questions are claimed to be of the same semantic type.
(14) Jan betwijfelt of/*wanneer Marie vertrekt.
Jan doubts whether/when Marie leaves
‘Jan doubts whether Marie will leave.’
For completeness’ sake, it should be noted that a less fortunate aspect of a binary
feature approach is that it does not account for the fact that factive verbs like weten
can also take yes/no-questions: Jan weet (niet) of Marie morgen komt ‘Jan
knows/does not know whether Marie is coming tomorrow’, which can never be
used to express a non-null proposition. This, as well as the problem noted for
example (14), shows that the binary feature approach is still in need of some fine-
tuning, but we leave this issue for future research.
The new class of [+FACTIVE,+WH] verbs does not seem to be restricted to
factive interrogative constructions. If we assume that the feature [+WH] does not
604 Syntax of Dutch: Verbs and verb phrases

refer to a semantic feature but to the formal (syntactic/morphological) feature that

wh-elements have in common and that enables them to undergo wh-movement, it
may also include verbs taking exclamative complements; cf. the primeless
examples in (11) above. Another construction that may be included, which is
discussed in Nye (2013), is the one illustrated in (15a); the complement clause in
this construction, which is especially found in narrative contexts, is introduced by
the wh-word hoe ‘how’ but seems to be more or less semantically equivalent with
the factive declarative dat-clause in (15b).
(15) a. Ik herinner me goed hoe hij daar altijd stond te kletsen.
I remember me well how he there always stood to chat
‘I well remember how he always stood chatting there.’
b. Ik herinner me goed dat hij daar altijd stond te kletsen.
I remember me well that he there always stood to chat
‘I well remember that he always stood chatting there.’
This section has shown that the semantic selection restrictions on finite complement
clauses exceed the dichotomies between (i) declarative and interrogative clauses
and (ii) yes/no- and wh-questions normally found in descriptive grammars. In
addition, we have shown that Nye’s (2013) binary-feature approach to the selection
of complement clauses has certain advantages compared to Grimshaw’s (1979)
approach based on semantic types.

4.2. Finite and infinitival argument clauses

We need to distinguish between finite and infinitival argument clauses. The
examples in (16) show that the choice depends on the matrix verb: whereas
propositional verbs like zeggen ‘to say’ or beweren ‘to claim’ can take either a
finite or an infinitival clause, an °irrealis verb like proberen ‘to try’ is only
compatible with an infinitival clause. Note that we assume that the infinitival
clauses in the primed examples contain a phonetically empty pronominal element
°PRO that functions as the implied subject of the infinitival clause; we will discuss
this element in more detail in Section 4.3.
(16) a. Jan beweert/zegt [dat hij morgen komt]. [finite clause]
Jan claims/says that he tomorrow comes
‘Jan claims that heʼll come tomorrow.’
a. Jan beweert /zegt [PRO morgen te komen]. [te-infinitival]
Jan claims/says tomorrow to come]
‘Jan claims to come tomorrow.’
b. *Jan probeert [dat hij morgen komt]. [finite clause]
Jan tries that he tomorrow comes
b. Jan probeert [PRO morgen te komen]. [te-infinitival]
Jan tries tomorrow to come]
‘Jan tries to come tomorrow.’
Replacing finite interrogative clauses by infinitival ones does not seem to be always
possible. Although example (17a) is acceptable, it belongs to a more formal
register of the language—in speech we only find more or less fixed forms like Hij
 Selection of clauses/verb phrases 605

vroeg me wat te doen ‘he asked me what he was supposed to do’. Example (17b) is
also acceptable, but the verb does not introduce a question; instead the embedded
clause has the function of a request.
(17) a. Jan vroeg me [hoe hij die auto kon repareren]. [finite clause]
Jan asked me how he that car could repair
‘Jan asked me how he could repair that car.’
a. Jan vroeg me [hoe PRO die auto te repareren]. [te-infinitival]
Jan asked me how that car to repair
b. Jan vroeg Marie [of ze kwam]. [finite clause]
Jan asked Marie whether she came
‘Jan asked (= inquired from) Marie whether sheʼd come.’
b. Jan vroeg Marie [PRO te komen]. [te-infinitival]
Jan asked Marie to come
‘Jan asked (= requested from) Marie to come.’
That the verb determines the form of the clausal argument is also clear from the
examples in (18), which show that while perception verbs like zien ‘to see’ exhibit
an alternation between finite and non-finite complement clauses, the
causative/permission verb laten allows infinitival clauses only.
(18) a. Jan zag [dat Marie vertrok]. [finite clause]
Jan saw that Marie left
a. Jan zag [Marie (*te) vertrekken]. [bare infinitival]
Jan saw Marie to leave
‘Jan saw Marie leave.’
b. *Jan laat [dat Marie vertrekt]. [finite clause]
Jan lets that Marie leaves
b. Jan laat [Marie (*te) vertrekken]. [bare infinitival]
Jan lets Marie to leave
A comparison of the primed examples in (18) with those in (16) shows us that the
verb also determines the type of infinitival clause; whereas the verbs in (16) take te-
infinitivals, the verbs in (18) take bare infinitivals. Section 5.2 will provide a brief
introduction to the different types of infinitival clauses.
It seems that there are only few verbs that can be combined with a finite but not
with an infinitival declarative object clause. Manner of speech verbs seem to prefer
a finite clause as their complement, but judgments on the corresponding infinitival
constructions appear to differ from case to case and from person to person. This is
clear from a Google search (3/16/2012) on the string [Vfinite * te zullen] for various
tense forms of the matrix verbs roepen ‘to call’, schreeuwen ‘to yell’ and huilen ‘to
cry’. while there are countless examples in which these verbs are followed by a
finite declarative clause, our search resulted in only a small number of cases in
which they were followed by an infinitival clause. We found a relatively large
number of examples such as (19a) with the verb roepen, but only two examples
such as (19b) with the verb schreeuwen, and no examples such as (19c) with the
verb huilen ‘to cry’. Nevertheless, the primed examples all seem passable for at
least some of the speakers we have consulted.
606 Syntax of Dutch: Verbs and verb phrases

(19) a. Hij riep jarenlang [dat hij nooit zou trouwen].

he called for.years that he never would marry
‘He called for years that he would never marry.’
a. Hij riep jarenlang [PRO nooit te zullen trouwen].
he called for.years never to will marry
b. Ze schreeuwden [dat ze hem zouden vermoorden].
they yelled that they him would kill
‘They yelled that they would kill him.’
%
b. Ze schreeuwden [PRO hem te zullen vermoorden].
they yelled him to will kill
c. Het jongetje huilde [dat hij gevallen was].
the boy cried that he fallen was
‘The boy cried that heʼd fallen.’
%
c. Het jongetje huilde [PRO gevallen te zijn].
the boy cried fallen to be
At first sight, the (a)-examples in (20) seem to contradict the claim that there are
few verbs that can be combined with a finite declarative clause only, but the (b)-
examples show that we should be careful not to jump to conclusions.
(20) a. Jan merkte/ontdekte [dat hij loog].
Jan noticed/discovered that he lied
a. $Jan merkte/ontdekte [PRO te liegen].
Jan noticed/discovered to lie
b. Jan merkte/ontdekte [dat hij honger had].
Jan noticed/discovered that he hunger had
‘Jan noticed/discovered that he was hungry.’
b. Jan merkte/ontdekte [PRO honger te hebben].
Jan noticed/discovered hunger to have
‘Jan noticed/discovered that he was hungry.’
The contrast between the two primed examples seems to be related to the preferred
interpretation of the implied subject PRO of the infinitival clause. First, consider the
primeless examples with a finite clause: the most prominent reading of (20a) is that
Jan noticed or discovered something about someone else, that is, the subject
pronoun hij ‘he’ of the embedded clause is preferably interpreted as referring to
some individual not mentioned in the sentence; example (20b), on the other hand, is
also compatible with a reading in which Jan noticed or discovered something about
himself, that is, in which the subject pronoun hij takes the subject of the matrix
clause as its antecedent. The contrast between the primed examples can now be
accounted for by referring to the fact that the implied subject PRO of the infinitival
clause differs from the subject pronoun hij in that it must be interpreted as
coreferential with the subject of the matrix clause; this makes the interpretation of
(20a) as unusual as that of (20a) if the pronoun hij is taken to be coreferential with
the subject of the matrix clause. See Section 4.3 for a brief introduction to the
restrictions on the interpretation of the PRO-subject of infinitival argument clauses.
It seems that, besides restrictions imposed by the interpretation of PRO, there
are various other factors that may affect the acceptability of infinitival argument
 Selection of clauses/verb phrases 607

clauses. The examples in (21), for instance, suggest that the verb voorkomen ‘to
prevent’ can only select finite clauses; the pronoun hij in the (a)-examples can
without difficulty be interpreted as coreferential with the subject of the main clause
but nevertheless the primed examples are severely degraded.
(21) a. Jan voorkwam net op tijd [dat hij zijn bril vermorzelde].
Jan prevented just in time that he his glasses crushed
‘Jan prevented just in time that he crushed his glasses.’
a. *?Jan voorkwam net op tijd [PRO zijn bril te vermorzelen].
Jan prevented just in time his glasses to crush
b. Jan voorkwam net op tijd [dat hij viel].
Jan prevented just in time that he fell
‘Jan prevented just in time that he fell.’
b. *?Jan voorkwam net op tijd [PRO te vallen].
Jan prevented just in time to fall
The examples in (22) show, however, that the primed examples improve a great
deal when we add an adverbial phrase indicating causation or manner.
(?)
(22) a. Jan voorkwam hierdoor net op tijd [PRO zijn bril te vermorzelen].
Jan prevented by.this just in time his glasses to crush
b. (?)Jan voorkwam zo net op tijd [PRO te vallen].
Jan prevented thus just in time to fall
The primed examples in (23), which are adaptations of actually occurring sentences
on the internet, further show that examples like these become even better if the
embedded clause is a passive or copular construction. Observe that the relevant
issue is not that subjects of passive and copular constructions are internal arguments
because the same thing holds for the subjects of °unaccusative verb like vallen ‘to
fall’ in the (b)-examples above.
(23) a. Hierdoor voorkwam hij [dat hij gedeporteerd werd naar Duitsland].
by.this prevented he that he deported was to Germany
‘In this way he prevented that he was deported to Germany.’
a. Hierdoor voorkwam hij [PRO gedeporteerd te worden naar Duitsland].
by.this prevented he deported to be to Germany
b. Zo voorkwam ik [dat ik zeeziek werd].
thus prevented I that I seasick became
‘In this way I prevented that I became seasick.’
b. Zo voorkwam ik [PRO zeeziek te worden].
thus prevented I seasick to become
Although there are more potential counterexamples to the claim that there are only
few verbs that can be combined with a finite but not with an infinitival declarative
clause, we will conclude by pointing out the contrast between the two examples in
(24). The reason why (24a) does not have an infinitival counterpart might simply be
that we are dealing with an idiomatic expression (which is also listed as such in
dictionaries); the options for substituting the finite clause in (24a) are very limited.
608 Syntax of Dutch: Verbs and verb phrases

(24) a. Ik maakte [dat ik wegkwam].

I made that I away-came
‘I got out as quickly as I could.’
b. *Ik maakte [PRO weg te komen].
I made away to come
The examples in (25) suggest that there are also few verbs that can be combined
with a finite but not with an infinitival prepositional object clause: the (a)-examples
show that verbs such as verwachten ‘to expect’ that normally take finite PO-clauses
also allow infinitival complements; verbs such as vragen ‘to request’ that normally
take infinitival PO-clauses give rise to a degraded result with finite complements.
(25) a. Jan verwacht [dat hij wordt uitgenodigd].
Jan expects that he is prt-invited
‘Jan expects that heʼll be invited.’
a. Jan verwacht uitgenodigd te worden.
Jan expects prt.-invited to be
‘Jan expects to be invited.’
b. Jan vraagt Marie [PRO te vertrekken].
Jan asks Marie to leave
‘Jan asks Marie to leave.’
b. ??Jan vraagt Marie [dat zij vertrekt].
Jan asks Marie that she leaves
To our knowledge the factors affecting the acceptability of infinitival argument
clauses have not been studied in detail but we provisionally conclude on the basis of
the discussion above that verbs selecting a finite declarative argument clause may
also take an infinitival clause in the normal course of things, but not necessarily
vice versa. Future research must show whether this conclusion is tenable.

4.3. Control properties of verbs selecting an infinitival clause

Section 4.2 has already shown that there are different restrictions on the
interpretation of referential subject pronouns of finite clauses on the one hand, and
°PRO-subjects of infinitival complement clauses on the other. The former can
freely take some antecedent from the °matrix clause or refer to some entity that is
part of the domain of discourse, whereas the latter must be coreferential with some
noun phrase in the matrix clause. We illustrate this again by showing that
passivization of the primeless examples in (26) gives rise to different results in
acceptability: the subject pronoun hij ‘he’ of the embedded finite clause in (26a)
can readily take some antecedent from the discourse domain, whereas the PRO-
subject of the infinitival clause in (26b) cannot.
(26) a. Jan ontdekte [dat hij honger had].
Jan discovered that he hunger had
‘Jan discovered that he was hungry.’
a. Er werd ontdekt [dat hij honger had].
there was discovered that he hunger had
‘It was discovered that he was hungry.’
 Selection of clauses/verb phrases 609

b. Jan ontdekte [PRO honger te hebben].

Jan discovered hunger to have
‘Jan discovered that he was hungry.’
b. *Er werd ontdekt [PRO honger te hebben].
there was discovered hunger to have
Intended reading: ‘It was discovered that he was hungry.’
The restrictions on the interpretation of PRO-subjects of infinitival complement
clauses have become known as °control theory. In many cases, it is required that
PRO should be controlled, that is, bound by some antecedent in the matrix clause.
The examples in (27) show, however, that PRO cannot take just any antecedent; in
(27a) PRO can only be controlled by the subject and in (27b) it can only be
controlled by the object of the matrix clause. The available readings are indicated
by means of referential indices.
(27) a. Jani beloofde Peterj [PROi/*j te komen]. [subject control]
Jan promised Peter to come
‘Jan promised Peter to come.’
b. Jani vroeg Peterj [PROj/*i te komen]. [object control]
Jan asked Peter to come
‘Jan asked Peter to come.’
The examples in (27) suggest that the interpretation of PRO is determined by the
matrix verb: accordingly, verbs like beloven ‘to promise’ have become known as
subject control verbs, and verbs like vragen ‘to ask’ as object control verbs.
However, the situation is more complex given that the contents of the embedded
clause may also affect the control options; adding a deontic modal verb like mogen
‘to be allowed’ to the infinitival clauses in (27), for example, reverses the
interpretation possibilities of PRO, a phenomenon known as CONTROL SHIFT.
(28) a. Jani beloofde Peterj [PROj/*i te mogen komen]. [object control]
Jan promised Peter to be.allowed.to come
‘Jan promised Peter to be allowed to come.’
b. Jani vroeg Peterj [PROi/*j te mogen komen]. [subject control]
Jan asked Peter to be.allowed.to come
‘Jan asked Peter to be allowed to come.’
The examples in (27) and (28) show that the interpretation of PRO can be affected
by properties of both the matrix verb and the infinitival clause. Moreover, it would
seem that these restrictions are not syntactic in nature but related to our knowledge
of the world; the interpretation of example (27a), for instance, is related to the fact
that the speaker has the ability to promise that he will perform a certain action
himself but he cannot promise that the addressee will perform that action; the
interpretation of example (28a), on the other hand, is based on the fact that the
speaker may grant permission to the addressee to do something, whereas it is much
less likely that he will or needs to grant such permission to himself. Consequently,
it is not at all surprising that we find similar shifts when the verbs beloven and
vragen take finite clauses as their complement.
610 Syntax of Dutch: Verbs and verb phrases

(29) a. Jani beloofde Peterj [dat hiji/*j zou komen].

Jan promised Peter that he would come
‘Jan promised Peter that he (≠ Peter) would come.’
a. Jani beloofde Peterj [dat hijj/*i mocht komen].
Jan promised Peter that he be.allowed.to come
‘Jan promised Peter that he (≠ Jan) would be allowed to come.’
b. Jani vroeg Peterj [of hijj/*i kwam].
Jan asked Peter whether he came
‘Jan asked Peter whether he (≠ Jan) was willing to come.’
b. Jani vroeg Peterj [of hiji/*j mocht komen].
Jan asked Peter whether he was.allowed.to come
‘Jan asked Peter whether he (≠ Peter) was allowed to come.’
For the moment we provisionally conclude that the PRO-subject of infinitival
argument clauses must be controlled by some antecedent in the matrix clause, but
that the actual choice of the antecedent must be compatible with our knowledge of
the world. Section 5.2 will show, however, that there are circumstances under
which the PRO-subject may be exempt from the requirement that it should be
bound.

4.4. Three main types of infinitival argument clauses

Sections 4.2 and 4.3 have shown that we need to distinguish between verbs taking
finite and verbs taking infinitival complement clauses, and that these types of
complements introduce a number of issues of a more semantic or pragmatic nature.
On top of this, this section will show that infinitival clauses do not constitute a
single syntactic category but can be subdivided into at least the three formally
different categories illustrated in (30): Om + te-infinitivals, te-infinitivals and bare
infinitivals.
(30) a. Jan beloofde [om PRO het boek naar Els te sturen]. [om + te-infinitival]
Jan promised COMP the book to Els to send
‘Jan promised to send the book to Els.’
b. Jan beweerde [PRO het boek naar Els te sturen]. [te-infinitival]
Jan claimed the book to Els to send
‘Jan claimed to send the book to Els.’
c. Jan wilde [PRO het boek naar Els sturen]. [bare infinitival]
Jan wanted the book to Els send
‘Jan wanted to send the book to Els.’
The following sections will briefly introduce the three subtypes in (30). For reasons
of presentation we begin with om + te-infinitivals and bare infinitivals, because we
will see that te-infinitivals again fall into various subgroups, some of which behave
more like om + te-infinitivals and some of which behave more like bare infinitivals.

4.4.1. Om + te-infinitivals
Om + te-infinitivals are formally characterized by the presence of the infinitival
marker te and the complementizer-like element om. There are at least two analyses
 Selection of clauses/verb phrases 611

available for the infinitival marker te. According to some proposals te is a bound
morpheme prefixed to the infinitival verb, just like ge- is a bound morpheme in past
participle forms like ge-pak-t ‘taken’. This may account for the fact that both te and
ge- are normally adjacent to the stem of the verb. An alternative proposal is that te
is the T(ense) °head of the functional projection TP. We refer the reader to Section
1.3, sub IIIA1, and references cited there for a more extensive discussion of these
proposals.
One reason for assuming that the element om is a complementizer, and not a
preposition, is that infinitival complement clauses introduced by this element
behave like finite complement clauses and not like PP-complements in that they
must be in extraposed position, that is, they obligatorily follow the °matrix verb in
clause-final position. This can be illustrated by means of the embedded and the
perfect-tense counterparts of example (30a), which are given in (31).
(31) a. dat Jan beloofde [om PRO het boek naar Els te sturen].
that Jan promised COMP the book to Els to send
‘that Jan promised to send the book to Els.’
a. *dat Jan [om PRO het boek naar Els te sturen] beloofde.
b. Jan heeft beloofd [om PRO het boek naar Els te sturen].
Jan has promised COMP the book to Els to send
‘Jan has promised to send the book to Els.’
b. *Jan heeft [om PRO het boek naar Els te sturen] beloofd.
Further grounds for assuming that om is a complementizer are that it can often be
omitted, as illustrated in (32a). This would be quite surprising for a preposition, but
it is attested for complementizers in many languages: cf. John promised (that) he
would send Elisabeth the book. Another reason for assuming we are not dealing
with a PP-complement is that the infinitival clause is not pronominalized by means
of the pronominal PP erom but by the pronoun dat; this is illustrated in (32b).
(32) a. Jan heeft beloofd [(om) PRO het boek naar Els te sturen].
Jan has promised COMP the book to Els to send
‘Jan has promised to send the book to Els.’
b. Jan heeft dat/*erom beloofd.
Jan has that/P+it promised
‘Jan has promised that.’
It should be noted, however, that the omission of om is not syntactically innocuous;
the examples in (33) show that it may make the infinitival clause transparent for
extraction of the object to a position to the left of the matrix verb in the matrix
clause; see Section 4.4.3 for more detailed discussion. The percentage sign in (33b)
is added because some speakers object to such examples.
(33) a. *Jan heeft het boeki beloofd [om PRO ti naar Els te sturen].
Jan has the book promised COMP to Els to send
b. %Jan heeft het boeki beloofd [PRO ti naar Els te sturen].
Jan has the book promised to Els to send
612 Syntax of Dutch: Verbs and verb phrases

The fact that this type of extraction is excluded from finite clauses such as (34a)
suggests that om + te-infinitivals and finite clauses are of the same categorial type;
they are CPs. Infinitival clauses without om, on the other hand, are likely to be less
extended verbal projections, which would make TP a likely candidate. See Section
9.1 for an introduction to the functional categories CP and TP.
(34) a. Jan heeft beloofd [CP dat hij het boek naar Els zal sturen].
Jan has promised that he the book to Els will send
‘Jan has promised that heʼll send the book to Els.’
b. *Jan heeft het boeki beloofd [dat hij ti naar Els zal sturen].
For completeness’ sake, note that the string Jan heeft het boek beloofd dat hij naar
Els zal sturen is acceptable if the postverbal clause is interpreted as a relative clause
modifying het boek (“John promised the book that he will bring to Els”), but this is
of course irrelevant here.

4.4.2. Bare infinitivals

This section discusses a number of formal properties of so-called bare infinitivals
and shows that we should distinguish at least three different subcategories.

I. The infinitive verb is not preceded by te

Bare infinitivals are characterized by the fact that they contain neither the
complementizer-like element om nor the infinitival marker te, that is, the infinitive
is bare in the sense of not being accompanied by any of the elements that we may
find in the two other types of infinitival clauses. The question as to whether a verbal
complement may appear as a bare infinitival depends on the matrix verb; a verb like
willen ‘to want’, for example, may take a finite clause or a bare infinitival, but not
an (om +) te-infinitival. Note in passing that English to want crucially differs from
Dutch willen in selecting a to-infinitival, not a bare infinitival.
(35) a. Jan wil [dat Peter het boek naar Els stuurt].
Jan wants that Peter the book to Els sends
‘Jan wishes that Peter will bring the book to Els.’
b. *Jan wil [(om) PRO het boek naar Els te sturen].
Jan wants COMP the book to Els to send
c. Jan wil [PRO het boek naar Els sturen].
Jan wants the book to Els send
‘Jan wants to send the book to Els.’

II. Verb clustering

Customarily, the bare infinitive forms a °verb cluster with the verb selecting the
bare infinitival complement. This is clear from the fact that the two verbs cluster in
clause-final position and that, as a result, the infinitival clause may be split:
example (36a) shows that whereas the bare infinitive follows the matrix verb in
clause-final position, all other constituents of the infinitival clause must precede it.
For convenience, we italicize the infinitival clauses in the examples below.
 Selection of clauses/verb phrases 613

(36) a. dat Jan het boek naar Els wil sturen.

that Jan the book to Els wants send
‘that Jan wants to send the book to Els.’
b. %dat Jan het boek wil naar Els sturen.
b. %dat Jan wil het boek naar Els sturen.
The percentage signs in the two (b)-examples in (36) are added to indicate that
certain southern varieties of Dutch also allow parts of the remaining part of the
embedded infinitival clause to follow the matrix verb; we will ignore this for the
moment and refer the reader to Section 5.2.3 for an extensive discussion of this.
III. The infinitivus-pro-participio (IPP) effect
Monoclausal behavior in the sense indicated in the previous subsection is typically
signaled by the so-called °infinitivus-pro-participio effect, that is, the phenomenon
that a verb does not surface in its expected past participial form when governed by a
perfect auxiliary, but as an infinitive. That constructions with bare infinitival
complements exhibit monoclausal behavior can be shown by comparing the perfect-
tense constructions in (37): if the matrix verb willen selects a finite clause, as in
(37a), it behaves as expected by appearing as a past participle in perfect-tense
constructions, but when it selects a bare infinitival complement, it must appear as
an infinitive in such constructions.
(37) a. Jan had gewild/*willen [dat Peter het boek naar Els had gestuurd].
Jan had wanted/want that Peter the book to Els had sent
‘Jan had wished that Peter would have sent the book to Els.’
b. Jan had het boek naar Els willen/*gewild sturen.
Jan had the book to Els want/wanted send
‘Jan had wanted to send the book to Els.’
IV. Three subtypes of bare infinitival clauses
Bare infinitival complements may occur in at least three different syntactic
environments, which differ in the way their subject is realized in the surface
structure: the subject can be realized as an accusative noun phrase in an °AcI-
construction, the phonetically empty element PRO in a °control construction, or as
the subject of the matrix clause in a °subject raising construction. In the following
examples the infinitival clauses are italicized and their subjects are underlined.
(38) a. Jan zag Marie/haar op de hei lopen. [AcI-infinitival]
Jan saw Marie/her on the heath walk
‘Jan saw Marie/her walk on the heath.’
b. Jan wil PRO een boek kopen. [control infinitival]
Jan wants a book buy
‘Jan wants to buy a book.’
c. Marie/Zij kan vertraagd zijn. [subject raising infinitival]
Marie/she may delayed be
‘Marie/She may be delayed.’
We will refer to these infinitival constructions by means of the names given in
straight brackets, for reasons that will become clear in the following subsections.
614 Syntax of Dutch: Verbs and verb phrases

A. Accusativus-cum-infinitivo infinitivals
Bare infinitival complement clauses selected by perception verbs like zien ‘to see’
or the causative/permissive verb laten ‘to make/let’ exhibit an ACCUSATIVUS-CUM-
INFINITIVO effect: the subjects of the bare infinitival clauses do not appear as the
phonetically empty element PRO, as would normally be the case in infinitival
clauses, but as an accusative noun phrase. This is illustrated in (39), in which the
subject of the infinitival clause is underlined.
(39) a. dat Jan het meisje/haar een lied hoorde zingen.
that Jan the girl/her a song heard sing
‘that Jan heard the girl/her sing a song.’
b. dat Jan het meisje/haar een lied liet zingen.
that Jan the girl/her a song made/let sing
‘that Jan made/let the girl/her sing a song.’
It is generally assumed that the subject of the infinitival complement is case-marked
by the matrix verb, that is, that we are dealing with so-called exceptional case-
marking across the boundary of an infinitival clause. That it is the matrix verb
which assigns case to the subject of the embedded clause is, however, not so easy to
prove for Dutch because the examples in (40) show that matrix verbs of AcI-
constructions cannot be passivized. We are therefore not able to provide evidence
that the underlined noun phrases in (39) are indeed assigned °accusative case by the
active matrix verbs. This claim must therefore be motivated by appealing to the fact
that there is simply no other element available that could be held responsible for
case-assignment.
(40) a. *dat het meisje/zij een lied werd gehoord/horen zingen.
that the girl/she a song was heard/hear sing
b. *dat het meisje/zij een lied werd gelaten/let zingen.
that the girl/she a song was made/make sing
That the underlined phrases in (39) are not selected by the matrix verbs but function
as the subjects of the bare infinitival clauses seems undisputed and can be supported
by means of pronominalization; the fact that the accusative noun phrase cannot be
realized in (41a) shows that it is not selected by the matrix verb horen ‘to hear’ but
must be part of the infinitival clause pronominalized by dat ‘that’. Unfortunately,
(41b) shows that pronominalization cannot readily be used as a test in the case of
the verb laten ‘to make/let’, as it is at best marginally acceptable with this verb
under its permissive reading and completely excluded under its causative reading.
(41) a. dat Jan (*het meisje/*haar) dat hoorde. [perception verb]
that Jan the girl/her that heard
‘that Jan heard that.’
b. dat Jan ??(*het meisje/*haar) dat liet. [permissive verb]
that Jan the girl/her that let
b. *dat Jan (het meisje/haar) dat liet. [causative verb]
that Jan the girl/her that let
 Selection of clauses/verb phrases 615

Accusativus-cum-infinitivo constructions of the type discussed here exhibit

monoclausal behavior. First, as is indicated by italics in (39) above, the bare
infinitival complements are normally split; whereas the bare infinitives normally
follow the matrix verbs in clause-final position, their arguments must precede them.
Second, the examples in (42) show that they exhibit the IPP-effect; the matrix verb
cannot surface as a past participle in perfect-tense constructions, but must be
realized as an infinitive.
(42) a. dat Jan het meisje/haar een lied heeft horen/*gehoord zingen.
that Jan the girl/her a song has hear/heard sing
‘that Jan has heard the girl/her sing a song.’
b. dat Jan het meisje/haar een lied heeft laten/*gelaten zingen.
that Jan the girl/her a song has make/made sing
‘that Jan has made/let the girl/her sing a song.’

B. Control infinitivals
A bare infinitival clause selected by a so-called root/deontic modal like kunnen ‘to
be able’, mogen ‘to be allowed’ or willen ‘to want’, or a verb like leren ‘to
teach/learn’ has its subject realized as the phonetically empty pronominal-like
element PRO. As in the case of (om +) te-infinitivals, the PRO-subject of a bare
infinitival can be either controlled by the subject or by the object of the matrix
clause. The choice again depends on the matrix verb: whereas de deontic modals
and intransitive leren ‘to learn’ require PRO to be controlled by their subjects,
transitive leren ‘to teach’ requires that PRO be controlled by its object. Again, we
have italicized the bare infinitival clause and underlined its subject.
(43) a. dat Jani PROi het boek naar Marie kan brengen.
that Jan the book to Marie is.able bring
‘Jan is able to bring the book to Marie.’
b. dat [zijn dochtertje]i PROi piano leert spelen.
that his daughter piano learns play
‘that his daughter is learning to play the piano.’
b. dat Jani [zijn dochtertje]j PROj/*i piano leert spelen.
that Jan his daughter piano teaches play
‘that Jan teaches his daughter to play the piano.’
Control constructions of the sort discussed here exhibit monoclausal behavior. First,
the constructions in (43) show once more that the bare infinitival can be split; as is
indicated by italics, the arguments of the bare infinitival precede the matrix verb in
clause-final position, whereas the bare infinitive normally follows it. Second, the
examples in (44) show that the construction exhibits the IPP-effect; the matrix verbs
cannot appear as past participles in perfect-tense constructions, but must surface as
infinitives.
(44) a. dat Jan PRO het boek naar Marie heeft kunnen/*gekund brengen.
that Jan the book to Marie has be.able/been.able bring
‘that Jan has been able to bring the book to Marie.’
616 Syntax of Dutch: Verbs and verb phrases

b. dat zijn dochtertje PRO piano heeft leren/*geleerd spelen.

that his daughter piano has learn/learned play
‘that his daughter has learnt to play the piano.’
b. dat Jan zijn dochtertje PRO piano heeft leren/*geleerd spelen.
that Jan his daughter piano has teach/ taught play
‘that Jan has taught his daughter to play the piano.’
That the noun phrases Jan in (43a) and zijn dochtertje in (43b&b) do not function
as subjects of the bare infinitivals is clear from the fact illustrated in (45) that they
must also be present when the infinitival clauses are pronominalized; this shows
that these noun phrases are assigned °thematic roles by the matrix verbs. The agent
role of the bare infinitive must therefore be assigned to some independent
argument, which motivates the postulation of a PRO-subject in these examples.
(45) a. Jan kan dat.
Jan is.able that
‘Jan is able to do that.’
b. Zijn dochtertje leert dat.
his daughter learns that
‘His daughter is learning that.’
b. Jan leert zijn dochtertje dat.
Jan teaches his daughter that
‘Jan is teaching that to his daughter.’
Note in passing that, if we adopt the conclusion from Section 4.6 that the quality of
being predicational is a defining property of main verbs, the fact that the root modal
kunnen ‘to be able’ in the (a)-examples above is able to license the noun phrase Jan
independently of the embedded infinitival shows that the traditional assumption that
root modal verbs are non-main verbs cannot be maintained and that they must
instead be seen as regular transitive verbs; cf. Klooster (1984/2001). We return to
this issue in Section 4.5.

C. Subject raising infinitivals

The previous subsection has put on hold the fact that examples such as (43a) are
actually ambiguous: the matrix verb can not only receive a deontic/root reading but
also a so-called epistemic reading. Although the most prominent reading of (43a) is
the deontic one, the ambiguity can be brought out by putting this example in the
perfect tense; if the modal verb is realized as a non-finite verb, it can only be
interpreted deontically as “to be able to”, but if it is realized as the finite verb it can
only be interpreted epistemically as “may”.
(46) a. dat Jan PRO het boek naar Marie heeft kunnen brengen. [deontic]
that Jan the book to Marie has be.able bring
‘that Jan has been able to bring the book to Marie.’
b. dat Jan het boek naar Marie kan hebben gebracht. [epistemic]
that Jan the book to Marie may have brought
‘that Jan may have brought the book to Marie.’
 Selection of clauses/verb phrases 617

That constructions with epistemic modals exhibit monoclausal behavior cannot be

demonstrated by the IPP-effect as the perfect auxiliary is now part of the infinitival
complement of the modal verb (see below), but it is still clear from the fact that the
bare infinitival can be split: the arguments of the infinitival clause precede the
modal verb in clause-final position whereas the bare infinitive normally follows it.
The underlining in (46) suggests entirely different structures for the two
constructions: if the modal verb has a deontic interpretation, the subject of the
infinitival clause is realized as the phonetically empty pronominal element PRO; if
the modal verb has an epistemic interpretation. the subject surfaces as °nominative
subject of the entire sentence by being promoted to subject (“raised to the subject
position”) of the matrix clause. Grounds for this assumption are again related to
pronominalization of the infinitival clause; example (47a) illustrates again that the
nominative subject is not affected by pronominalization if the modal verb is
deontic, whereas (47b) shows that the nominative argument cannot be realized if
the modal is epistemic and should therefore be assumed to belong to the
pronominalized infinitival clause. We have added example (47b) in order to
support our earlier claim that the perfect auxiliary in the epistemic constructions in
(46b) is part of the infinitival complement, which is pronominalized by dat in the
(b)-examples in (47).
(47) a. Jan heeft dat gekund. [deontic/*epistemic]
Jan has that been.able
‘Jan has been able to do that.’
b. Dat kan. [epistemic/*deontic]
that may.be.the.case
b. *Dat kan hebben.
that may.be.the.case have
There is another good reason for assuming that the nominative subject in the
epistemic example in (46b) originates as the subject of the infinitival complement
clause. This immediately accounts for the fact that in passive constructions such as
(48b) the internal argument of the bare infinitive stelen ‘to steal’ surfaces as the
nominative subject of the sentence; passivization of the bare infinitive first
promotes the noun phrase die auto ‘that car’ to subject of the infinitival clause, and
subject raising subsequently promotes it to subject of the matrix clause.
(48) a. Jan kan de auto/hem gestolen hebben.
Jan may that car/him stolen have
‘Jan may have stolen that car/it.’
b. Die auto/Hij kan gestolen zijn.
that car/he may stolen have.been
‘That car/It may have been stolen.’
Under the alternative hypothesis that the nominative noun phrases in the examples
in (48) originate as arguments of the epistemic modal kunnen, we can only account
for the pattern in (48) by adopting the highly unlikely assumption that passivization
of the embedded verb affects the selectional properties of the matrix verb.
618 Syntax of Dutch: Verbs and verb phrases

A final argument we mention here is that example (49a), with a subject clause
introduced by the °anticipatory pronoun het ‘it’, is semantically more or less
equivalent to (49b), at least with respect to the thematic relations between the
italicized elements. If we assume that the nominative subject in (49b) originates as
the subject of the infinitival clause and is subsequently promoted to subject of the
matrix clause, the observed semantic equivalence follows straightforwardly.
(49) a. Het kan dat Jan gevallen is.
it may.be.the.case that Jan fallen is
‘It may be the case that Jan has fallen.’
b. Jan kan gevallen zijn
Jan may fallen be
‘Jan may have fallen.’
Each of the examples in (47) to (49) strongly suggests that nominative subjects
in epistemic constructions such as (46b) originate as the subject of the infinitival
clause and are subsequently “raised” to the subject position of the matrix clause.
We can formally derive this by assuming that the subject of the infinitival clause
cannot be assigned accusative case and must therefore be assigned nominative case
by being promoted to subject of the matrix clause in a fashion similar to objects in
passive constructions.
(50) a. ___ Vepistemic [NP .... Vinfinitive ] [underlying structure]
b. NPi Vepistemic [ti .... Vinfinitive ] [Subject Raising]
Note that the analysis in (50) implies that epistemic modals do not assign an
external thematic role. They must be able to assign an internal thematic role,
however, which is clear from the fact that the finite complement clause in (49a), or
the anticipatory pronoun in subject position introducing it, must be semantically
licensed. Given the similarity in meaning between the two constructions in (49), we
may also assume that the infinitival clause in (49b) must likewise be assigned an
internal thematic role. If we adopt the conclusion from Section 4.6 that being
predicational is a defining property of main verbs, the conclusion that epistemic
modal verbs assign an internal thematic role would imply that the traditional view
that epistemic modal verbs are non-main verbs cannot be maintained; we should,
instead, consider them as °unaccusative main verbs.

D. Conclusion
The previous subsections have shown that bare infinitival clauses may occur in at
least three types of syntactic environment which affect the way their subject is
realized: the subject can be realized as an accusative noun phrase, the phonetically
empty element PRO, or it may be “raised”, that is, be promoted to subject of the
matrix clause and be assigned nominative case. What we did not discuss, and what
is in fact a still largely unresolved issue, is what the syntactic mechanisms are that
determine the form of the subject of the infinitival clause. For example, why is it
that the modal verb willen ‘want’ lacks the ability of perception verbs to assign
accusative case to the subject of their infinitival complement. Is this simply a
lexical property of the verbs involved, or are we dealing with different syntactic
 Selection of clauses/verb phrases 619

structures? And, why is it that the subject of the infinitival clause is realized as PRO
when the modal verb moeten is deontic but not when it is epistemic; cf. Klooster
(1986)?
(51) a. *Jan wil [Marie komen].
Jan wants Marie come
Intended reading: ‘Jan wants Marie to come.’
b. Jan moet [PRO om drie uur aanwezig zijn]. [deontic]
Jan must at three oʼclock present be
‘Jan must be present at 3.p.m.’
c. Jani moet [ti om drie uur aanwezig geweest zijn]. [epistemic]
Jan must at three oʼclock present be
‘Jan must have been present at 3.p.m.’
Since we do not have anything insightful to offer on the first question, we leave it
as an unsolved issue for future research. The second question poses a severe
problem for the traditional formulation of control theory in Chomsky (1981), which
in effect states that °traces of movement and PRO cannot occur in the same
syntactic configuration. The answer may lie in an appeal to the alternative proposal
in Koster (1978:ch.2) and, more specifically, Koster (1984a/1984b) that the
difference is a property of the antecedent of the empty category (trace/PRO); we
will briefly return to this issue in the conclusion of Section 5.2.2.1.

4.4.3. Te-infinitivals
This section shows that from an observational point of view clausal complements
headed by a te-infinitive can be divided into at least three subtypes: one type that
exhibits behavior similar to om + te infinitivals, one type exhibiting behavior
similar to bare infinitivals, and a third type that exhibits mixed behavior. The main
characteristics of the three types are given in (52).
(52)  Types of te-infinitivals
a. Opaque: no clause splitting and no IPP-effect
b. Transparent: clause splitting and IPP-effect
c. Semi-transparent: clause splitting and no IPP
The abbreviation IPP stands for the infinitivus-pro-participio effect, the
phenomenon that matrix verbs sometimes cannot appear as past participles in
perfect-tense constructions but must surface as infinitives. Section 4.4.2, sub III, has
shown that this effect is obligatory in constructions with bare infinitivals, but
Subsections I to III below will demonstrate that this does not hold for te-infinitivals;
obligatory IPP is only found with transparent te-infinitivals.
The term CLAUSE SPLITTING refers to the phenomenon that infinitival clauses
can be discontinuous: the infinitive and its arguments may surface on different sides
of the matrix verb in clause-final position. Evidence has been presented in section
4.4.2, sub II, that in the case of bare infinitivals clause splitting is a concomitant
effect of °verb clustering, that is, the formation of an impermeable series of verbs in
clause-final position. Subsection IV will show, however, that clause splitting is
probably not a uniform process in the case of te-infinitivals: transparent and semi-
620 Syntax of Dutch: Verbs and verb phrases

transparent te-infinitivals are different in that only the former involve verb
clustering in the technical sense given above.
The term OPAQUE (or INCOHERENT) as applied to the infinitival clause refers to
the fact that such clauses constitute an independent clausal domain in the sense that
they may block locally restricted syntactic dependencies like °NP-movement or
°binding of the weak reflexive zich ‘him/her/itself’. TRANSPARENT (or COHERENT)
infinitives, on the other hand, behave in certain respects as if they constitute a single
clause with the matrix clause: they do not block such dependencies.
Another term for transparency found in the literature is RESTRUCTURING, which
has a transformational background in that it was assumed that an underlying
biclausal structure is transformationally restructured such that the embedded
infinitival clause forms a monoclausal structure with the matrix clause; see Evers
(1975), Rizzi (1982:ch.1) and much subsequent work. Since several more recent
approaches do not adopt this transformational view, we will not use this notion in
this work in order to avoid unnecessary theoretical bias.

I. Opaque te-infinitivals
Verbs taking opaque te-infinitival complement clauses are, e.g., NOM-DAT verbs,
PO-verbs and particle verbs; see Evers (1975:39ff) and Den Besten et al. (1988).
The primeless examples in (53) show that such verbs do not allow clause splitting;
like om + te-infinitivals, the te-infinitival is in extraposed position, that is, placed in
a position following the matrix verb in clause-final position. The primed examples
further show that the matrix verbs appear as participles in the perfect tense, that is,
as in the case of infinitival clauses introduced by om, there is no IPP-effect. For
convenience, we will italicize the te-infinitivals in the examples below and refrain
from indicating their implied PRO-subject for the sake of simplicity.
(53) a. dat het hem <*het boek> berouwt <het boek> gekocht te hebben.
that it him the book regrets bought to have
‘that he regrets it that he has bought the book.’
a. Het heeft hem berouwd/*berouwen het boek gekocht te hebben.
it has him regretted/repent the book bought to have
‘He has regretted it that he has bought the book.’
b. dat Jan ertoe <*het boek> neigt <het boek> te kopen.
that Jan to.it the book inclines to buy
‘that Jan is inclined to buy the book.’
b. Jan is ertoe geneigd/*neigen het boek te kopen.
Jan is to.it inclined/incline the book to buy
‘Jan is inclined to buy the book.’
c. dat Peter Marie <??dat boek> opdraagt <dat boek> te kopen.
that Peter Marie that book prt.-ordered to buy
‘that Peter orders Marie to buy that book.’
c. Peter heeft Marie opgedragen/*opdragen dat boek te kopen.
Peter has Marie prt.-ordered/prt.-order that book to buy
‘Peter has ordered Marie to buy that book.’
 Selection of clauses/verb phrases 621

Opaque infinitivals appear to be characterized by the fact that they do not have the
syntactic function of direct object of the matrix verb, nor are they assigned a
°thematic role by it. The infinitival clauses in the (a)-examples above function as
subjects and may also be introduced by the anticipatory subject pronoun het ‘it’.
The infinitival clauses in the (b)-examples correspond to the nominal part of a PP-
complement of the matrix verb, as is clear from the fact that they can be introduced
by the °anticipatory pronominal PP ertoe ‘to it’. The infinitival clauses in the (c)-
examples, finally, are not arguments of the verb at all but licensed as °logical
SUBJECTs of the verbal particle op; see Section 2.2.1.

II. Transparent te-infinitivals

Verbs selecting a transparent infinitival complement often have a modal or
aspectual interpretation. Examples are the modal verbs schijnen ‘to seem’, lijken ‘to
appear’ and blijken ‘to turn out’. That the infinitival complements of these verbs are
transparent is clear from the fact that they are obligatorily split; whereas the te-
infinitive in (54) must follow the matrix verb in clause-final position, its object
must precede it.
(54) dat Jan <een nieuwe auto> schijnt <*een nieuwe auto> te kopen.
that Jan a new car seems to buy
‘that Jan seems to be buying a new car.’
That the infinitival complement in (54) is transparent is also clearly shown by the
fact that we are dealing with °subject raising, that is, promotion of the subject of the
infinitival clause to °nominative subject of the higher clause. This will become clear
when we consider the near-equivalent examples in (55): the subject of the finite
complement clause in (55a) appears as the nominative subject of the entire sentence
in (55b), in which the complement clause is infinitival.
(55) a. Het schijnt dat Jan een nieuwe auto koopt.
it seems that Jan a new car buys
‘It seems that Jan is buying a new car.’
b. dat Jan een nieuwe auto schijnt te kopen.
that Jan a new car seems to buy
‘that Jan seems to be buying a new car.’
Unfortunately, it is more difficult to illustrate that the modal verbs schijnen ‘to
seem’, lijken ‘to appear’ and blijken ‘to appear’ trigger the IPP-effect, for the
simple reason that not all speakers allow them to occur in perfect-tense
constructions, especially not if they take an infinitival complement. Speakers that
do allow IPP, however, normally prefer the use of an infinitive.
(56) dat Jan een nieuwe auto heeft %schijnen/*geschenen te kopen.
that Jan a new car has seem/seemed to buy
‘that Jan has seemed to buy a new car.’
Other examples of transparent te-infinitivals mentioned both by Evers (1975:5) and
Den Besten et al. (1988) are the somewhat formal/obsolete semi-modals dienen ‘to
be obliged to’, plegen ‘to be accustomed/tend’ and weten ‘to be able to’, which
622 Syntax of Dutch: Verbs and verb phrases

seem to have a deontic interpretation and are probably best analyzed as °control
structures. It is, however, hard to find support for this analysis given that the
infinitival clauses cannot be pronominalized without the loss of the modal
interpretation of the matrix verbs. The transparent nature of the te-infinitivals in
(57) is clear from the fact that clause splitting and the IPP-effect are obligatory in
these examples.
(57) a. dat Jan <dat boek> dient <*dat boek> te lezen.
that Jan that book is.obliged to read
‘that Jan has to read that book,’
a. dat Jan dat boek heeft dienen/*gediend te lezen.
that Jan that book has be.obliged/been.obliged to read
‘that Jan has had to read that book.’
b. dat Marie <dat boek> weet <*dat boek> te bemachtigen.
that Marie that book knows to obtain
‘that Marie is able (knows how) to obtain that book.’
b. dat Marie dat boek heeft weten/*geweten te bemachtigen.
that Marie that book has know/known to obtain
‘that Marie has been able to obtain that book.’

III. Semi-transparent te-infinitivals

Evers (1975) suggested that te-infinitivals functioning as theme arguments and
surfacing as direct objects can (in our terms) be either opaque or transparent, but he
also noted that some verbs, his class IIIb, are not very particular in the sense that
they can select either type. We illustrate this in (58) with perfect-tense constructions
containing the matrix verb proberen ‘to try’. The fact that the verb appears as a
participle in (58a) but as an infinitive in (58b) suggests that we are dealing with,
respectively, an opaque and a transparent infinitival clause in these examples. This
is also supported by the fact that the infinitival clause is split in (58b), but not in
(58a). Following the standard hypothesis of the time that Dutch has an underlying
OV-order, Evers accounted for this by assuming that the direct object clause is
base-generated to the left of the matrix verb, and that (58a) and (58b) are derived
by, respectively, EXTRAPOSITION of the entire clause and VERB RAISING of the
infinitival verb te lezen ‘to read’.
(58) a. dat Jan heeft ti geprobeerd [PRO dat boek te lezen]i. [opaque]
that Jan has tried that book to read
‘that Jan has tried to read that book.’
b. dat Jan [PRO dat boek tte lezen] heeft proberen te lezen. [transparent]
that Jan that book has try to read
‘that Jan has tried to read that book.’
The examples in (59) suggest, however, that it is not sufficient to assume that
certain verbs optionally trigger extraposition or verb raising. The unacceptability of
example (59a) first shows that extraposition indeed requires that the matrix verb
surfaces as a past participle in perfect-tense constructions; there are no extraposition
constructions that involve IPP in Standard Dutch (but see Barbiers et al., 2008:
Section 2.3.6.1.3, for a number of Flemish and Frisian dialects that do accept
 Selection of clauses/verb phrases 623

examples such as (59a)). Den Besten et al. (1988) found, however, that clause
splitting is very common when the matrix verb appears as a participle, that is,
clause splitting does not require IPP as is clear from the fact that it is easy to find
example (59b) alongside (58b); cf. Gerritsen (1991: Map 25), Haeseryn et al.
(1997:950-2), Barbiers et al. (2008), and taaladvies.net/taal/advies/vraag/674.
(59) a. *dat Jan heeft proberen dat boek te lezen.
that Jan has try that book to read
‘that Jan has tried to read that book.’
b. dat Jan dat boek heeft geprobeerd te lezen. [semi-transparent]
that Jan that book has tried to read
‘that Jan has tried to read that book.’
Note in passing that the verb proberen is special in that it seems equally acceptable
with opaque, transparent and semi-transparent infinitival complement clauses.
Many verbs are more restrictive in this respect (although there is always some
variation in what speakers do or do not accept): besluiten ‘to decide’, for example,
can only take opaque or semi-transparent te-infinitivals, as is clear from the fact
illustrated in (60b) that it is incompatible with the IPP-effect.
(60) a. dat Jan <dat boek> besloot <dat boek> te lezen.
that Jan that book decided to read
‘that Jan has decided to read that book.’
b. dat Jan <dat boek> heeft <dat boek> besloten te lezen. [opaque/semi-tr.]
that Jan that book has decided to read
‘that Jan decided to read that book.’
b. *dat Jan dat boek heeft besloten/*?besluiten te lezen. [transparent]
that Jan that book has decided/decide to read
‘that Jan has decided to read that book.’

IV. Potential problems with the classification of te-infinitivals

The main conclusion to be drawn from Subsections I to III is that from an
observational point of view we can distinguish the three types of te-infinitivals in
(61) on the basis of whether or not clause splitting and IPP are possible.
(61)  Types of te-infinitivals
a. Opaque: no clause splitting and no IPP-effect
b. Transparent: clause splitting and IPP-effect
c. Semi-transparent: clause splitting and no IPP
It should be pointed out, however, that semi-transparent te-infinitivals differ from
transparent ones in that the former do not require that all non-verbal constituents of
the infinitival clause precede the matrix verb; cf. the contrast between the two
examples in (62). This bears out that clause splitting of semi-transparent te-
infinitivals is not the result of verb clustering in the technical sense defined in the
introduction to this section, that is, the formation of an impermeable series of verbs
in clause-final position.
624 Syntax of Dutch: Verbs and verb phrases

(62) a. dat Marie die jongen <een kus> heeft proberen <*een kus> te geven.
that Marie that boy a kiss has try to give
b. dat Marie die jongen <een kus> heeft geprobeerd <een kus> te geven.
that Marie that boy a kiss has tried to give
‘that Marie has tried to give that boy a kiss.’
Constructions with semi-transparent te-infinitivals like (59b) and (62b) were
referred to as the third construction in Den Besten et al. (1988), but have become
known later as the REMNANT EXTRAPOSITION construction. Den Besten et al. (1988)
derived the construction by a combination of °extraposition of the te-infinitival and
leftward movement of one or more of its constituents. As a result, the extraposed
phrase consists of merely a remnant of the original te-infinitival (see also Reuland
1981). If we adopt the leftward movement analysis (while leaving open the question
as to whether extraposition involves rightward movement of the infinitival clause),
the representations of (59b) in (62b) are as given in (63).
(63) a. dat Jan dat boeki heeft geprobeerd [PRO ti te lezen].
b. dat Marie die jongeni een kusj heeft geprobeerd [PRO ti tj te geven].
b. dat Marie die jongeni heeft geprobeerd [PRO ti een kus te geven].
The fact that the direct object een kus ‘a kiss’ in (62b) may either precede or follow
the clause-final verbs implies that the postulated leftward movement is optional.
This means that it is no longer obvious that the te-infinitivals in examples such as
(64) should be considered opaque as they can also be analyzed as semi-transparent
clauses without the postulated leftward movements in (63).
(64) a. dat Jan heeft geprobeerd dat boek te lezen.
that Jan has tried that book to read
‘that Jan has tried to read that book.’
b. dat Marie heeft geprobeerd die jongen een kus te geven.
that Marie has tried that boy a kiss to give
‘that Marie has tried to give that boy a kiss.’
All of this might indicate that Den Besten et al. (1988) were wrong in assuming that
there are opaque te-infinitivals, and that rather we have to assume that all te-
infinitivals are (semi-)transparent. If so, the “opaque” cases discussed in Subsection I
cannot be described by appealing to the label “clause type”. Since
(semi-)transparent infinitival clauses differ crucially from the opaque infinitival
clauses discussed in Subsection I in that they (i) are selected as internal arguments
of a verb and (ii) have the syntactic function of direct object, this may be the key to
the solution. This will be one of the topics addressed in our more extensive
discussion of te-infinitivals in Section 5.2.2.

4.5. Non-main verbs

Non-main verbs differ from main verbs: they do not denote states of affairs, but
express additional (temporal, modal, aspectual, etc.) information about states of
affairs denoted by main verbs. This implies that non-main verbs are normally
accompanied by the projection of a main verb. Moreover, constructions with non-
 Selection of clauses/verb phrases 625

main verbs are characterized by the fact that the main verbs in them are never finite.
The examples in (65) also show that the form of the non-finite main verb depends
on the type of non-main verb: perfect and passive auxiliaries, for example, combine
with past/passive participles, modal/aspectual verbs combine with bare infinitivals,
and semi-aspectual verbs combine with te-infinitivals.
(65)  Types of non-main verbs
a. Jan heeft dat boek gelezen. [perfect auxiliary]
Jan has that book read
‘Jan has read that book.’
b. Het boek werd me (door Peter) toegestuurd. [passive auxiliary]
the book was me by Peter prt.-sent
‘The book was sent to me (by Peter).’
c. Jan wil/gaat dat boek kopen. [modal/aspectual verb]
Jan wants/goes that book buy
‘Jan wants/is going to buy that book.’
d. Jan zit dat boek te lezen. [semi-aspectual verb]
Jan sits that book to read
‘Jan is reading that book.’
Although the set of non-main verbs traditionally assumed is substantially larger
than the four groups mentioned in (65), we will confine ourselves to these verbs for
the purpose of illustration; Section 5.2 will provide a more exhaustive discussion.

I. Perfect and passive auxiliaries

Auxiliaries like hebben and zijn are temporal in the sense that the perfect-tense
constructions they are part of situate the state of affairs prior to a specific point in
time. Example (66a), for instance, situates the arrival of Marie prior to the speech
time (which is the default value), as the fact that it can be modified by the time
adverbial gisteren ‘yesterday’ but not by the time adverbial morgen ‘tomorrow’
makes quite clear. In addition, perfect-tense constructions may under certain
conditions also have aspectual implications by expressing that the state of affairs
denoted by the main verb is completed in the sense that some logically implied
endpoint has been reached: example (66b), for example, can only be used when Jan
has told the full story. We refer the reader to Section 1.5.1 for a more detailed
discussion of the semantics of the perfect tense.
(66) a. Marie is (gisteren/*morgen) gearriveerd.
Marie is yesterday/tomorrow arrived
‘Marie arrived/Marie arrived yesterday.’
b. Jan heeft me het verhaal (gisteren/*morgen) verteld.
Jan has me the story yesterday/tomorrow told
‘Jan has told me the story (yesterday).’
Participles are also used in combination with the auxiliaries worden ‘to be’ and zijn
‘to have been’ in regular passive constructions like (67a&b) and the auxiliary
krijgen ‘to get’ in so-called krijgen-passive constructions such as (67c).
626 Syntax of Dutch: Verbs and verb phrases

(67) a. Het boek werd me (door Peter) toegestuurd.

the book was me by Peter prt.-sent
‘The book was sent to me (by Peter).’
b. Het boek is me (door Peter) toegestuurd.
the book has.been me by Peter prt.-sent
‘The book has been sent to me (by Peter).’
c. Ik kreeg het boek (door Peter) toegestuurd.
I got the book by Peter prt.-sent
‘I was sent the book (by Peter).’
Note in passing that the auxiliary verb zijn in (67b) is sometimes analyzed not as a
passive but as a perfect auxiliary given that the passive participle geworden can at
least marginally be added to such examples. If correct, this means that worden and
krijgen would exhaust the set of passive auxiliaries, but see Section 6.2.2 for a
potential problem for this conclusion.
That the auxiliaries discussed in this section are only instrumental in creating
perfect tense or passive constructions immediately accounts for the fact that they
cannot be used as heads of clauses (although zijn ‘to be’ and worden ‘to become’ do
occur as copulas, and hebben ‘to have’ and krijgen ‘to get’ can also be used as main
verbs of possession).

II. Modal/aspectual verbs

The examples in (68) show that modal and aspectual verbs like willen and gaan
differ from temporal and passive auxiliaries in that they do not combine with
participles but require the main verb to take the form of a bare infinitive.
(68) a. Jan wil dat boek morgen kopen.
Jan wants that book tomorrow buy
‘Jan wants to buy that book tomorrow.’
b. Jan gaat morgen dat boek kopen.
Jan goes tomorrow that book buy
‘Jan is going to buy that book tomorrow.’
The primeless examples in (69) show that modal and aspectual verbs also differ
from main verbs in that they exhibit the infinitivus-pro-participio (IPP) effect; they
do not take the form of a participle in perfect-tense constructions, but of an
infinitive. The primed examples have been added to show that willen and gaan do
appear as participles are used as main verbs.
(69) a. Jan heeft dat boek altijd al willen/*gewild kopen.
Jan has that book always already want/wanted buy
‘Jan has always wanted to buy that book.’
a. Jan heeft dat boek altijd al gewild/*willen.
Jan has that book always already wanted/want
‘Jan has always wanted have that book.’
 Selection of clauses/verb phrases 627

b. Jan is dat boek gaan/*gegaan kopen.

Jan has that book go/gone buy
‘Jan has gone to buy that book.’
b. Jan is naar de winkel gegaan/*gaan.
Jan is to the shop gone/go
‘Jan has gone to the shop.’
If modal and aspectual verbs supplement the event expressed by the main verb with
specific modal/aspectual information, we expect that these verbs cannot be used
without a main verb. This is indeed borne out in the case of the aspectual verbs, but
not in the case of the modal verbs: the (a)-examples in (70) show that the string een
ijsje kopen can be pronominalized by means of het ‘it’ or dat ‘that’; the (b)-
examples show that this is not possible with aspectual verbs (although speakers do
accept °left dislocation constructions like Een ijsje kopen, dat gaan we zeker!
‘Buying ice cream we certainly will!’; we refer the reader to Section 4.6, sub II, for
reasons for assuming that this does not involve pronominalization).
(70) a. Jan wil [een ijsje kopen]. a. Jan wil het/dat.
Jan wants an ice.cream buy Jan wants it/that
‘Jan wants to buy an ice cream.’
b. Jan gaat [een ijsje kopen]. b. *Jan gaat het/dat.
Jan goes an ice.cream buy Jan goes it/that
‘Jan is going to buy an ice cream.’
Of course, one might try to solve this problem with modal verbs by assuming that
example (70a) in fact contains a phonetically empty verb that corresponds to the
semantically light verb doen ‘to do’ in (71a), but this would leave unexplained why
this verb cannot co-occur with the aspectual verb gaan.
(71) a. Jan wil het/dat doen. a. Jan wil het/dat 
Jan wants it/that do Jan wants it/that 
b. Jan gaat het/dat doen. b. *Jan gaat het/dat .
Jan goes it/that do/ Jan goes it/that 
Furthermore, this line of thinking might lead us to expect the modal verb willen to
exhibit the IPP-effect irrespective of whether the clause contains the verb doen or
its phonetically empty counterpart . The examples in (72) show that this
expectation is not borne out: the effect does not occur when doen is not present.
(72) a. Jan heeft het/dat willen/*gewild doen.
Jan has it/that want/wanted do
b. Jan heeft het/dat gewild /*willen .
Jan has it/that wanted/want 
Finally, the fact illustrated in (73) that modal verbs differ from aspectual verbs like
gaan in that they can be combined with a nominal object is problematic for the
view that the former is a non-main verb.
628 Syntax of Dutch: Verbs and verb phrases

(73) a. Jan wil een ijsje.

Jan wants an ice.cream
‘Jan want to have an ice cream.’
b. *Jan gaat een ijsje.
Jan goes an ice.cream
The examples above are intended to bring out that it is not a priori clear that the
question as to whether or not a specific verb can be used as the only verb of a clause
is cast iron proof for establishing whether or not that specific verb is a main verb.
We will return to this issue in Section 4.6.

III. Semi-aspectual verbs

Semi-aspectual verbs correspond to main verbs like zitten ‘to sit’, liggen ‘to lie’,
hangen ‘to hang’ and staan ‘to stand’ in (74), which refer to a certain posture or
position of the subject of the clause, and certain verbs of movement like lopen ‘to
walk’.
(74) a. Het boek staat in de kast.
the book stands in the bookcase
‘The book is in the bookcase.’
b. Het boek ligt op tafel.
the book lies on table
‘The book is lying on the table.’
In their semi-aspectual use the lexical meaning of the main verb can but need not be
present; examples like those in (75) can be used comfortably when the speaker
cannot observe the referent of the subject of the clause and is thus not able to tell
whether this referent is actually sitting or walking at the moment of speech. The
primary function of the semi-aspectual verb is to indicate that we are dealing with
an ongoing event and we are thus dealing with a progressive construction
comparable to the English progressive construction: see the renderings given in
(75).
(75) a. Jan zit momenteel te lezen.
Jan sits at.present to read
‘Jan is reading at the moment.’
b. Els loopt momenteel over het probleem te piekeren.
Els walks at.present on that problem to worry
‘Els is worrying about that problem at the moment.’
The examples in (75) also show that semi-aspectual verbs differ from the modal and
aspectual verbs in (68) in that they do not combine with bare infinitivals but with
so-called te-infinitivals: leaving out the infinitival marker te leads to
ungrammaticality. This is, however, not the case in the corresponding perfect-tense
constructions in (76), in which the marker te can be dropped. The examples in (76)
make it clear as well that the semi-aspectual verbs exhibit the IPP-effect;
replacement of the infinitive zitten/lopen by the participle gezeten/gelopen leads to
ungrammaticality
 Selection of clauses/verb phrases 629

(76) a. Jan heeft de hele dag zitten (te) lezen.

Jan has the whole day sit to read
‘Jan has been reading the whole day.’
b. Els heeft de hele dag over het probleem lopen (?te) piekeren.
Els has the whole day on the problem walk to worry
‘Els has been worrying about that problem all day.’

IV. Non-main verbs are part of a verbal complex

The previous subsections have shown that non-main verbs impose certain
restrictions on the morphological form of the main verb: temporal and passive
auxiliaries select participles, modal/aspectual verbs select bare infinitivals, and
(finite) semi-aspectual verbs select te-infinitivals. What we have not shown yet is
that non-main verbs and the main verb they select obligatorily form a °verbal
complex, in which the main verb refers to some state of affairs and the non-main
verbs function as modifiers providing supplementary information. This is clear
from the fact that an embedded main verb cannot normally be the °head of an
independent finite clause introduced by the complementizer dat ‘that’ or an
infinitival clause introduced by the complementizer om. We illustrate this in (77)
for the aspectual verb gaan and the semi-aspectual verb zitten; the number sign # is
used to indicate that (77b) is possible if zitten is construed as a main verb and the
infinitival clause is an adverbial purpose clause: “Jan sits in order to read the book”.
(77) a. *Jan gaat dat hij het boek leest.
Jan goes that he the book reads
b. #Jan zit om dat boek te lezen.
Jan sits COMP that book to read
Observe that this test shows again that a modal verb like willen ‘to want’ can be
used as a main verb; see the discussion of (72) in Subsection II. We will return to
the issue in Section 4.6.
(78) a. Jan wil op tijd komen.
Jan wants in time arrive
‘Jan wants to arrive there on time.’
b. Jan wil dat hij op tijd komt.
Jan want that he in time arrives
‘Jan wants that heʼll arrive there on time.’

V. Placement of the non-main verb in the clause

All examples in the subsections above have been presented as main with the non-
main verb in second position. In most varieties of Dutch spoken in the Netherlands
the auxiliaries cluster with the main verb in clause-final position; the arguments of
the main verb must precede the non-main verb even when the main verb follows it.
This clausal split is illustrated in (79) for the perfect auxiliary hebben ‘to have’, the
modal verb willen ‘to want’ and the semi-aspectual verb zitten ‘to sit’.
630 Syntax of Dutch: Verbs and verb phrases

(79) a. dat Jan <het boek> heeft <*het boek> gelezen.

that Jan the book has read
‘that Jan has read the book.’
b. dat Jan <het boek> wil <%het boek> lezen.
that Jan the book wants read
‘that Jan wants to read the book.’
c. dat Jan <het boek> zit <%het boek> te lezen.
that Jan the book sits to read
‘that Jan is reading the book.’
We should note, however, that certain southern varieties of Dutch (including the
standard variety spoken in Belgium) do allow the object to intervene between non-
main verbs and (te-)infinitives, hence the use of the percentage sign in (79b&c). See
Barbiers (2008: Section 2.3.1) and Chapter 7 for more detailed information.

VI. Conclusion
The previous subsections have shown that auxiliaries must be accompanied by a
main verb in the same clause. Furthermore, non-main verbs place restrictions on the
form of the main verb they select: temporal and passive auxiliaries select
participles, modal and aspectual verbs select bare infinitivals, and (finite) semi-
aspectual verbs select te-infinitivals. Non-main verbs do not combine with clauses
introduced by the complementizer dat or om, which strongly suggests that non-main
verbs must form a single verbal complex with a main verb. Finally, we have seen
that in the varieties of Dutch spoken in the Netherlands, clauses with non-main
verbs exhibit monoclausal behavior in the sense that they trigger °verb clustering,
as a result of which the projection of the main verb must be split if the non-main
verb is in clause-final position.

4.6. The distinction between main and non-main verbs

Although native speakers normally have little difficulty in distinguishing main from
non-main verbs, there are cases in which making a decision is not so
straightforward; see the remarks on the behavior of the modal verb willen in Section
4.5, sub II and IV. The question now arises what the crucial differences between
main and non-main verbs are. We will consider two options: (i) the question as to
whether the non-main and the infinitival main verb enter a °verbal complex in the
complex resulting in monoclausal behavior, and (ii) the question as to whether the
verb can be considered predicational in nature. We will argue that the second option
is to be preferred despite the fact that this will give rise to a somewhat different
dividing line between non-main and main verbs than traditionally assumed; cf., e.g.,
Haeseryn et al. (1997).

I. Mono- versus biclausal behavior

Main and non-main verbs play a different semantic role in the clause. The former
function semantically as n-place predicates and are therefore typically the (semantic
and syntactic) °head of a clause; sentences that contain two main verbs are thus
normally biclausal. The fact that the addition of a non-main verb to a clause such as
 Selection of clauses/verb phrases 631

(80a) does not affect the number of arguments that can be expressed is normally
taken as evidence that non-main verbs are not predicates. Instead, they are assumed
to add, e.g., temporal, aspectual or modal information to the meaning expressed by
the main verb.
(80) a. Jan leest het boek. [main verb only]
Jan reads the book
b. Jan heeft het boek gelezen. [perfect auxiliary]
Jan has the book read
c. Jan wil/gaat het boek lezen. [modal/aspectual verb]
Jan wants/goes the book read
d. Jan zit het boek te lezen. [semi-aspectual verb]
Jan sits the book to read
Let us therefore for the moment assume that non-main verbs must, but main verbs
cannot combine with another main verb in a structure exhibiting monoclausal
behavior, and that we can test this for infinitival constructions by assuming that
mono- and biclausal structures systematically differ with respect to °verb clustering
and the infinitivus-pro-participio (IPP) effect in the way indicated in Table 2.
Table 2: Structures exhibiting mono- and biclausal behavior
MONOCLAUSAL BICLAUSAL
VERB CLUSTERING + —
INFINITIVUS-PRO-PARTICIPIO + —

The examples in (81) illustrate the monoclausal properties of structures containing

the semi-aspectual verb zitten. First, example (81a) shows that the semi-aspectual
verb and the main verb lezen form a verb cluster, as a result of which the infinitival
verb zitten is separated from its direct object het boek ‘the book’. Second, example
(81b) shows that the IPP-effect is obligatory.
(81) a. dat Jan <dat boek> zit <%dat boek> te lezen. [verb clustering]
that Jan that book sits to read
‘that Jan is reading that book.’
b. Jan heeft dat boek zitten/*gezeten te lezen. [IPP]
Jan has that book sit/sat to read
‘Jan has been reading that book.’
We should note, however, that verb clustering is somewhat obscured in the varieties
of Dutch spoken in Belgium since these allow permeation of the verb cluster by
various elements; for example, the order in (81a) marked by a percentage sign is
acceptable in some of these varieties. Further, we should note that passive
constructions are exempt from the IPP-effect; we will ignore this here but return to
it in Section 6.2.2.
The examples in (82) illustrate the biclausal properties of structures containing
the main verb beweren ‘to claim’: example (82a) shows that the object het boek ‘the
book’ of the verb lezen ‘to read’ can intervene between beweren and lezen ‘to read’
and (82b) shows that the IPP-effect does not arise.
632 Syntax of Dutch: Verbs and verb phrases

(82) a. dat Jan beweert dat boek te lezen. [no verb clustering]
that Jan claims that book to read
b. Jan heeft beweerd/*beweren dat boek te lezen. [no IPP]
Jan has claimed/claim that book to read
Now consider example (83a), in which the verb proberen ‘to try’ semantically
functions as a two-place predicate with an agentive subject and an infinitival direct
object clause. That we are dealing with a regular direct object clause is clear from
the fact illustrated in (83b) that the infinitival clause can be pronominalized or be
replaced by a referential noun phrase.
(83) a. Jan probeerde (om) dat boek te lezen.
Jan tried COMP that book to read
‘Jan tried to read that book.’
b. Jan probeerde het/een nieuw merk sigaretten.
Jan tried it/ a new brand [of] cigarettes
‘Jan tried it/a new brand of cigarettes.’
Example (83a) also shows that the infinitival complement of proberen can be either
an om + te-infinitival or a te-infinitival without the complementizer om. We will
see shortly that these infinitival complements exhibit a somewhat different
behavior, but, first, the examples in (84) show that the two types of infinitival
clause may be placed after the verb proberen in clause-final position, and that
proberen must occur as a past participle in the corresponding perfect-tense
construction. This is fully consistent with the earlier claim that proberen is a main
verb.
(84) a. dat Jan probeert (om) dat boek te lezen.
that Jan tries COMP that book to read
‘that Jan is trying to read that book.’
b. dat Jan heeft geprobeerd/*proberen (om) dat boek te lezen.
that Jan has tried/try COMP that book to read
‘that Jan has tried to read that book.’
The examples in (85) show, however, that the te-infinitival without om is special in
that it is also compatible with the IPP-effect, provided that the object of the
infinitival verb lezen precedes proberen: the word order in (85b) is unacceptable.
(85) a. dat Jan dat boek heeft proberen te lezen.
that Jan that book has tried to read
‘that Jan has tried to read that book.’
b. *Jan heeft proberen dat boek te lezen.
This shows that proberen may also trigger monoclausal behavior, from which we
may conclude that it does not always behave like a run-of-the mill main verb, but
may be of a hybrid nature in the sense that it also exhibit properties of non-main
verbs. The fact that proberen is not an isolated case and that there are more
unsuspected main verbs which can enter a verbal complex and thus trigger
monoclausal behavior strongly suggests that having this option is not a defining
 Selection of clauses/verb phrases 633

property of non-main verbs. This is confirmed by the fact that constructions with
bare infinitivals always exhibit monoclausal behavior, irrespective of whether the
selecting verb is a main or a non-main verb: this is illustrated in (86) for the
aspectual verb gaan and the main verb horen ‘to hear’.
(86) a. dat hij een liedje gaat zingen. [verb clustering]
that he a song goes sing
‘that heʼs going to sing a song’
a. dat hij een liedje is gaan zingen. [infinitivus-pro-participio]
that he a song is gone sing
‘that he has started singing a song’
b. dat ik hem een liedje hoor zingen. [verb clustering]
that I him a song hear sing
‘that I hear him sing a song.’
b. dat ik hem een liedje heb horen zingen. [infinitivus-pro-participio]
that I him a song have heard sing
‘that Iʼve heard him sing a song.’
All of this implies that the hypothesis that main verbs differ from non-main verbs in
that they cannot combine with another main verb in a structure that exhibits
monoclausal behavior is refuted, and, consequently, that we have to look for other
means to distinguish main from non-main verbs.

II. The predicational nature of the verb

This subsection investigates two other syntactic properties that seem related to the
predicational nature of main versus the non-predicational nature of non-main verbs.
The predicational nature of main verbs like beweren ‘to claim’ and proberen ‘to
try’ is clear from the fact that they do not require a °projection of a main verb as
their complement; the primed examples in (87), in which the italicized infinitival
clauses of the primeless examples are pronominalized or replaced by a noun phrase,
unambiguously show that we are dealing with two-place predicates, that is, regular
transitive main verbs.
(87) a. Jan beweerde dat boek te lezen.
Jan claimed that book to read
a. Jan beweerde het/de vreemdste dingen.
Jan claimed it/the weirdest things
b. Jan probeert dat boek te lezen.
Jan tried that book to read
b. Jan probeerde het/een stickie.
Jan tried it/a joint
Non-main verbs like the aspectual verb gaan in the (a)-examples in (88), on the
other hand, are clearly not predicational, as is clear from the fact that they normally
do not allow pronominalization of the projection of the infinitival main verb: the
verb gaan is not able to license the subject and the object pronoun, which clearly
shows that it does not behave like a transitive verb. A potential problem is,
however, that the (b)-examples show that modal verbs exhibit unexpected behavior
634 Syntax of Dutch: Verbs and verb phrases

in this respect; example (88b) shows that pronominalization is possible (see also
Section 4.5, sub II, where the same point was made).
(88) a. Jan gaat het boek lezen. a. *Jan gaat het/dat.
Jan goes the book read Jan goes it/that
‘Jan is going to read the book.’
b. Jan wil het boek lezen. b. Jan wil het/dat.
Jan wants the book read Jan wants it/that
‘Jan wants to read the book.’
Another potential problem is that we wrongly expect that main verbs always allow
pronominalization of their infinitival complement. Consider the (b)-examples in
(89) with the causative/permissive verb laten. Example (89b) shows that laten adds
an argument to those selected by the embedded main verb lezen in (89a), from
which we may safely conclude that it is a two-place predicate that selects a nominal
subject and an object clause. Example (89b) shows, however that laten does not
allow pronominalization of the embedded infinitival clause. The (c)-examples are
added to show that perception verbs such as zien ‘to see’ do behave as expected by
allowing pronominalization of the embedded clause.
(89) a. Jan leest het boek.
Jan reads the book
b. Zij laat Jan het boek lezen. b. *Zij laat het/dat.
she makes Jan the book read she makes it/that
‘She makes/lets Peter read the book.’
c. Zij zag Jan het boek lezen. c. Zij zag het/dat.
she saw Jan the book read she saw it/that
‘She saw Jan read the book.’
We have seen that there are two ways to establish whether a verb that combines
with an infinitival verb is propositional in nature. The easiest way is to investigate
whether it is able to introduce an argument that is not licensed by the embedded
main verb; if this is the case, the °matrix verb clearly has an argument structure of
its own. The second way is to investigate whether the projection of the infinitival
verb can be pronominalized; if so, we may conclude that the pronoun must be
semantically licensed and therefore functions as an argument of the verb. Table 3
provides the results of these tests for the verbs in (88) and (89).
Table 3: A comparison of aspectual, modal and causative verbs
VERB TYPE ADDITIONAL ARGUMENT PRONOMINALIZATION EXAMPLE
aspectual — — (88a)
modal — + (88b)
causative + — (89a)
perception + + (89b)

Assuming that the distinction between main and non-main verbs is really
determined by the question as to whether the verb is predicational in nature, we
have to conclude that of the four verb types discussed here, only the aspectual verbs
 Selection of clauses/verb phrases 635

can be considered non-main verbs. This implies that the dividing line between these
two sets will be slightly different than normally assumed in more traditional
grammars. For example, whereas modal verbs are normally considered non-main
verbs, we are bound to conclude that they are main verbs; see Klooster
(1984/1986).
For completeness’ sake, we conclude by noting that the pronominalization test
must be applied with care; not all structures containing the pronoun dat/het can be
used to show that the verb under investigation is predicational in nature. There
appear to be two complications. First, the examples in (90) show that secondary
predicates can also be pronominalized by the pronoun dat; the intended
interpretation of the pronoun is indicated by means of coindexing. The acceptability
of the second conjunct in these examples does not show that the copular verb zijn is
a two-place predicate; as Section 2.2 has shown, it is simply a verb taking a
predicative small-clause complement.
(90) a. Jan is slimi en Marie is dati ook.
Jan is smart and Marie is that too
b. Jan is [een goede leerling]i en Marie is dati ook.
Jan is an apt pupil and Marie is that too
Second, the examples in (91) show that °left dislocation constructions should also
be set aside. The fact illustrated in (91a) that the pronoun dat can be used to refer to
the left-dislocated participle phrase does not show that the perfect auxiliary hebben
is a two-place predicate. In fact, if we took example (91a) as evidence for assuming
that the perfect auxiliary hebben is two-place predicate, we would be forced to
conclude on the basis of examples like (91b&c) that it can also be a three- or even a
four-place predicate, a conclusion that is clearly untenable.
(91) a. [Boeken gelezen]i dati heeft hij niet.
books read that has he not
‘He hasnʼt read books.’
b. [Gelezen]i dati heeft hij dat boek niet.
read that has he that book not
‘He hasnʼt read that book.’
c. [Gegeven]i dati heeft hij Peter dat boek niet.
given that has he Peter that book not
‘He hasnʼt given Peter that book.’

III. Why we discuss non-main verb constructions as subcases of complementation

We normally use the term complement as equivalent with the term °internal
argument; it refers, e.g., to arguments of verbs that are assigned a °thematic role
like goal or theme. Given that Section 4.6 has argued that main and non-main verbs
differ in that only the former are predicational in nature, and that the latter are not
able to select any arguments, we could restrict the term verbal complement such
that it only refers to verbal arguments of main verbs. Nevertheless, we will adopt a
somewhat looser notion of verbal complements that also includes the verbal
projections in the domain of non-main verbs. The main reason for doing so is that
we have seen that non-main verbs impose certain morphosyntactic selection
636 Syntax of Dutch: Verbs and verb phrases

restrictions on the main verb: perfect auxiliaries, for example, must combine with
past participles, aspectual verbs only combine with bare infinitivals, and semi-
aspectual verbs normally combine with te-infinitivals. By stating that non-main
verbs select the projection of the main verb as their complement, these selection
restrictions can be accounted for.
(92) a. Jan heeft dat boek gelezen. [perfect auxiliary]
Jan has that book read
‘Jan has read that book.’
b. Jan gaat dat boek kopen. [modal/aspectual verb]
Jan goes that book buy
‘Jan is going buy that book.’
d. Jan zit dat boek te lezen. [semi-aspectual verb]
Jan sits that book to read
‘Jan is reading that book.’
By discussing verbal complements of main and non-main verbs in a single chapter,
it will also become easier to compare the behavior of such verbal complements.
That this is desirable is clear from the fact that Subsection II has shown that besides
clear-cut cases of main and non-main verbs, there are also verbs that are of a more
hybrid nature; we will see numerous other cases in Section 5.2.2 and 5.2.3.

IV. Conclusion
This section discussed a number of properties of main and non-main verbs. Main
verbs function semantically as n-place predicates and are therefore typically the
(semantic and syntactic) head of some clause; if the sentence contains two main
verbs, they are prototypically expressed in a biclausal structure. Non-main verbs, on
the other hand, are not predicates but provide additional information to the meaning
expressed by the main verb. As a result, non-main verbs must combine with a main
clause in a verbal complex and thus trigger monoclausal behavior; they exhibit the
two properties indicated in Table 4, repeated from Subsection I.
Table 4: Structures exhibiting mono- and biclausal behavior
MONOCLAUSAL BICLAUSAL
VERB CLUSTERING + —
INFINITIVUS-PRO-PARTICIPIO + —

It is nevertheless not always easy to determine whether we are dealing with a main
or a non-main verb, given that some verbs exhibit a somewhat hybrid behavior.
Subsection II was devoted to the question as to how we can distinguish main from
non-main verb. We argued that it is not sufficient to show that a verb enters into a
verbal complex with an infinitival main verb and then draw the conclusion that we
are dealing with a non-main verbs, given that main verbs like proberen ‘to try’ also
have this property. Therefore we decided that we need to investigate the
predicational nature of the verb in question: if addition of this verb results in the
addition of an argument that is not licensed by the non-finite main verb, or if the
projection of the non-finite main verb can be pronominalized, we are dealing with a
 Selection of clauses/verb phrases 637

main verb. This leads to a classification slightly different from what is normally
assumed in descriptive grammars. We illustrated this for modal verbs like willen,
which are normally classified a non-main verbs but must be considered to be main
verbs according to our criterion. Section 5.2 will show that this also holds for a
number of other verb types.
Chapter 5 Projection of verb phrases IIIb:
Argument and complementive clauses

Introduction 640
5.1. Finite argument clauses 641
5.1.1. General introduction 642
5.1.2. Direct object clauses 649
5.1.2.1. Selection restrictions on finite direct object clauses 650
5.1.2.2. The placement of finite object clauses 661
5.1.2.3. Factive versus non-factive complement clause constructions 669
5.1.2.4. Reported speech 684
5.1.3. Subject clauses 717
5.1.4. Prepositional object clauses? 725
5.1.5. Fragment clauses 728
5.1.6. Wh-extraction from argument clauses 756
5.1.7. Independently used argument clauses 763
5.1.8. Bibliographical notes 765
5.2. Infinitival argument clauses 765
5.2.1. Om + te-infinitivals 766
5.2.1.1. The distribution of om + te-infinitivals 767
5.2.1.2. The categorial status of the element om 774
5.2.1.3. The implied subject PRO in om + te-infinitivals 776
5.2.2. Te-infinitivals 802
5.2.2.1. Control infinitivals 804
5.2.2.2. Subject raising infinitivals 818
5.2.2.3. Extraposition and verb clustering 846
5.2.3. Bare infinitivals 872
5.2.3.1. The verb leren ‘to teach/learn’ 874
5.2.3.2. Modal verbs 879
5.2.3.3. Perception verbs 899
5.2.3.4. Causative laten ‘to make/let’ and doen ‘to make’ 918
5.2.3.5. Hebben ‘to have’ + infinitive 931
5.2.4. Bibliographical notes 935
5.3. Complementive clauses 936
640 Syntax of Dutch: Verbs and verb phrases

Introduction
This chapter provides an exhaustive discussion of dependent clauses functioning as
°arguments or °complementives. Section 5.1 starts with finite argument clauses; we
will consider in detail subject, direct object, and prepositional clauses.
(1) a. dat duidelijk is [dat Marie de nieuwe voorzitter wordt]. [subject]
that clear is that Marie the new chairman becomes
‘that it is clear that Marie will be the new Chair.’
b. dat Jan niet gemeld heeft [dat hij weg zou zijn]. [direct object]
that Jan not reported has that he away would be
‘that Jan hasnʼt reported that heʼd be away.’
c. dat Peter erover klaagt [dat het regent]. [prepositional object]
that Peter about.it complains that it rains
‘that Peter is complaining about that it is raining.’
Section 5.1 also includes a discussion of fragment clauses and wh-extraction. A
typical example of fragment clauses is given in (2a), in which the wh-word who is
interpreted in the same way as the embedded clause in Ik weet niet wie Jan gisteren
heeft bezocht ‘I do not know who Jan has visited yesterday.’ Wh-extraction is
illustrated in (2b) by means of wh-movement of the direct object of the
°complement clause; the °trace ti indicates the normal position of the direct object.
(2) a. Jan heeft gisteren iemand bezocht, maar ik weet niet wie.
Jan has yesterday someone visited but I know not who
‘Jan visited someone yesterday but I donʼt know who.’
b. Wati denk je [CLAUSE dat Marie ti morgen zal kopen]?
what think you that Marie tomorrow will buy
‘What do you think that Marie will buy tomorrow?’
Section 5.2 discusses three types of formally different types of infinitival clauses:
Om + te-infinitivals, te-infinitivals and bare infinitivals. Some typical examples are
given in (3), which typically have an implicit (phonetically empty) subject pronoun,
normally represented as PRO; an important issue will be what the conditions on the
interpretation of °PRO are (°control theory).
(3) a. Jan beloofde [om PRO het boek naar Els te sturen]. [om + te-infinitival]
Jan promised COMP the book to Els to send
‘Jan promised to send the book to Els.’
b. Jan beweerde [PRO het boek naar Els te sturen]. [te-infinitival]
Jan claimed the book to Els to send
‘Jan claimed to send the book to Els.’
c. Jan wilde [PRO het boek naar Els sturen]. [bare infinitival]
Jan wanted the book to Els send
‘Jan wanted to send the book to Els.’
Section 5.2 also discusses °subject raising and accusativus-cum-infinitivo infinitivals
such as (4). We will give reasons for assuming that the °nominative subject in (4a)
is extracted from the infinitival clause and that the subject of the infinitival clause in
 Argument and complementive clauses 641

(4b) functions as the subject of the infinitival clause but is assigned °accusative case
by the °matrix verb horen ‘to hear’.
(4) a. Jani schijnt [ti een nieuwe auto te kopen]. [subject raising]
Jan seems a new car to buy
‘Jan seems to buy a new car.’
b. Els hoorde [henacc een liedje zingen]. [accusativus-cum-infinitivo]
Els heard them a song sing
‘Els heard them sing a song.’
Section 5.3 concludes with a discussion of complementives, that is, clauses that
function as secondary predicates; examples that are sometimes (perhaps incorrectly)
analyzed as involving complementive clauses are the copular constructions in (5).
(5) a. Een feit is [dat hij te lui is].
a fact is that he too lazy is
‘Itʼs a fact is that heʼs too lazy.’
b. dat boek is moeilijk [(om) te lezen].
that book is hard COMP to read
‘that book is hard to read.’

5.1. Finite argument clauses

Section 5.1.1 starts with a number of more general remarks concerning finite
°argument clauses. Sections 5.1.2 through 5.1.4 discuss in more detail the use of
finite clauses as direct objects, subjects and prepositional objects. Section 5.1.5
continues with a discussion of fragment clauses. A prototypical case of the type of
fragment clauses we have in mind is provided by the so-called sluicing construction
in (6b), which can be used as a reaction to example in (6a). Sluicing constructions
are arguably derived by partial deletion of the phonetic contents of a finite clause,
which is indicated here by means of crossing-out.
(6) a. Jan heeft gisteren iemand bezocht. [speaker A]
Jan has yesterday someone visited
‘Jan visited someone yesterday.’
b. Kan je me ook zeggen wie Jan gisteren bezocht heeft? [speaker B]
can you me also tell who Jan yesterday visited has
‘Can you also tell me who (Jan visited yesterday)?’
Section 5.1.6 concludes with a brief discussion of wh-extraction from finite clauses,
which is illustrated in (7) by means of wh-movement of a direct object; the wh-
phrase wat in (7b) arguably originates in the same position as the direct object dit
boek in (7a); consequently, the embedded clause in (7b) contains an interpretative
gap, which we have indicated by means of the °trace ti.
(7) a. Ik denk [CLAUSE dat Marie dit boek morgen zal kopen].
I think that Marie this book tomorrow will buy
b. Wati denk je [CLAUSE dat Marie ti morgen zal kopen]?
what think you that Marie tomorrow will buy
‘What do you think that Marie will buy tomorrow?’
642 Syntax of Dutch: Verbs and verb phrases

Wh-extraction is only possible from complement clauses of a limited set of verbs,

and our discussion will focus especially on the properties that a °matrix verb must
have in order to license wh-extraction. For a more general and extensive discussion
of the restrictions on wh-movement the reader is referred to Section 11.3.1.

5.1.1. General introduction

This section provides a brief introduction to a number of more general issues
concerning finite argument clauses. We begin with a brief discussion of the
syntactic functions that argument clauses may have. This is followed by some
remarks on their form, with special attention to the position of the finite verb and
the form of their complementizer. We then investigate the anticipatory pronominal
elements that can be used to introduce finite argument clauses. We conclude this
introduction with a brief discussion of free relatives, which are sometimes also
analyzed as argument clauses.

I. The syntactic function of finite argument clauses

Finite clauses regularly occur as arguments of verbs: they can be used as subject,
direct object or as part of a prepositional object. Indirect objects are normally
nominal, which seems related to the fact that they typically refer to living entities or
institutions, not to propositions. The examples in (8) show that argument clauses are
normally placed after the verbs in clause-final position. The reason for calling the
embedded clause in (8c) a prepositional object and not a direct object is that it
cannot be pronominalized by means of the pronoun het, but must be replaced by the
pronominal PP erover. The properties of the three types of argument clauses in (8)
will be discussed in greater detail in Sections 5.1.2 to 5.1.4.
(8) a. dat duidelijk is [dat Marie de nieuwe voorzitter wordt]. [subject]
that clear is that Marie the new chairman becomes
‘that it is clear that Marie will be the new Chair.’
b. dat Jan niet gemeld heeft [dat hij weg zou zijn]. [direct object]
that Jan not reported has that he away would be
‘that Jan hasnʼt reported that he wouldnʼt be there.’
c. dat Peter klaagt [dat het regent]. [prepositional object]
that Peter complains that it rains
‘that Peter is complaining that it is raining.’
c. dat Jan erover/*het klaagt.
that Jan about.it/it complains

II. The form of finite argument clauses

Finite argument clauses normally take the form of an embedded clause, that is, a
clause with the finite verb in clause-final position, as in the indirect reported speech
example in (9a). Possible exceptions to this general rule are found in the direct and
semi-direct reported speech examples in (9b-c), in which the apparent dependent
clause appears in main clause order, that is, with the finite verb in second position.
For this reason cases of direct and semi-direct speech deserve special attention and
they will therefore be discussed separately in Section 5.1.2.4.
 Argument and complementive clauses 643

(9) a. Jan zei [dat hij Marie ging bezoeken]. [indirect reported speech]
Jan said that he Marie went visit
‘Jan said that he was going to visit Marie.’
b. Jan zei: “Ik ga Marie bezoeken.” [direct reported speech]
Jan said I go Marie visit
‘Jan said: “Iʼm going to visit Marie”.’
c. Jan zei hij ging Marie bezoeken. [semi-direct reported speech]
Jan said he went Marie visit
‘Jan said he was going to visit Marie.’
Examples (10a&b) show that declarative argument clauses are obligatorily
introduced by the complementizer dat ‘that’, that is, unlike English that, Dutch dat
cannot be omitted. Example (10c) further shows that Dutch also differs from
German in that it does not allow embedded clauses without a complementizer and
with °verb-second; see Haider (1985) for a discussion of verb-second in embedded
clauses in German and Barbiers et al. (2005: Section 1.3.1.8) for a number of Dutch
(especially eastern) dialects that may also have this construction. Observe that
example (10c) is acceptable as a case of direct reported speech, but this is, of
course, not the reading intended here.
(10)  Declarative argument clauses
a. Jan zegt [dat Peter ziek is]. [with complementizer]
Jan says that Peter ill is
‘Jan says that Peter is ill.’
b. *Jan zegt [Ø Peter ziek is]. [without complementizer and without V2]
Jan says that Peter ill is
‘Jan says Peter is ill.’
c. *Jan zegt [Peter is ziek]. [without complementizer and with V2]
Jan says Peter is ill
Interrogative argument clauses are introduced either by the complementizer of
‘whether’ or by a wh-phrase. In speech (but not in written language) it is also
common that the wh-phrase in embedded wh-questions is followed by a
complementizer: the complementizer of is used in the northern, whereas the
complementizer dat is more common in the southern varieties; some (mainly
northern) speakers even use the combination of dat; we refer the reader to Barbiers
(2005: Section 1.3.1.5) for details on the geographical distribution of these options;
see also Hoekstra & Zwart (1994), Sturm (1996) and Zwart & Hoekstra (1997) on
the question as to whether of dat should be analyzed as a compound or as two
separate words.
(11)  Interrogative argument clauses
a. Jan vraagt [of Peter ziek is]. [yes/no-question]
Jan asks whether Peter ill is
‘Jan asks whether Peter is ill.’
b. Jan vraagt wie (of/dat) er ziek is. [wh-question]
Jan asks who whether/that there ill is
‘Jan asks who is ill.’
644 Syntax of Dutch: Verbs and verb phrases

If two embedded yes/no questions are coordinated by means of the disjunction of

‘or’, as in (12a), the complementizer of the second clause does not occur as of but
as dat in order to avoid a sequence of two (homophonous) occurrences of of. That
this is a surface phenomenon is clear from the fact illustrated in (12b) that the
second complementizer must be realized as of when we replace the disjunction of
by the more formal disjunction dan wel ‘or’; see Haeseryn et al. (1997:547).
(12) a. Ik weet niet [of hij nog komt] of [dat/*of hij thuis blijft].
I know not whether he still comes or that/whether he home stays
‘I donʼt know whether heʼs still coming or whether heʼll stay at home.’
b. Ik weet niet [of hij nog komt] dan wel [of/*dat hij thuis blijft].
I know not whether he still comes or whether/that he home stays
‘I donʼt know whether heʼs still coming or whether heʼll stay at home.’
There is a small set of cases in which what would seem to be an argument clause is
introduced by the conjunction als ‘if/when’; cf. Haeseryn et al. (1997:1136&1153).
The primeless examples in (13) show that such als-clauses are especially common
in constructions with a subject/object experiencer, although the primed examples
show that the experiencer may also remain implicit; observe that het functions as an
°anticipatory pronoun associated with the als-clause. To our knowledge als-clauses
of this type have received little attention in the literature so far, and, in fact, it
remains to be demonstrated whether they do indeed function as argument clauses in
these cases; this is why Section 5.1.2.2, sub IV, investigates them in more detail.
(13)  Argument clauses introduced by als ‘if/when’?
a. JanExperiencer waardeert het [als je hem helpt]. [subject experiencer]
Jan appreciates it if one him helps
‘Jan appreciates it if you help him.’
a. Het wordt gewaardeerd [als je hem helpt]. [implicit experiencer]
it is appreciated if you him helps
‘Itʼs appreciated if you help him.’
b. Het irriteert me [als je zingt]. [object experiencer]
it annoys me when you sing
‘Your singing annoys me.’
b. Het is irritant [als je zingt]. [implicit experiencer]
it is annoying when you sing
‘Your singing annoys me.’
It is important to observe that the distinction between declarative and
interrogative embedded clauses is formal rather than semantic: the embedded clause
in (14a) is called declarative despite the fact that we are clearly not dealing with an
assertion, and the embedded clauses in (14b&c) are called interrogative despite the
fact that we are not dealing with true questions. Notwithstanding this, we will
simply accept the traditional terminology.
(14) a. Jan vermoedt [dat hij ziek is]. [declarative clause]
Jan suspects that he ill is
‘Jan suspects that heʼs ill.’
 Argument and complementive clauses 645

b. Jan betwijfelt [of hij op tijd zal aankomen]. [yes/no-question]

Jan doubts whether he on time will arrive
‘Jan doubts whether heʼll arrive in time.’
c. Els onderzoekt [wie het boek gestolen heeft]. [wh-question]
Els investigates who the book stolen has
‘Els is investigating who has stolen the book.’

III. The anticipatory pronominal elements het ‘it’ and er + P ‘P + it’

The examples in (15) show that finite argument clauses may be introduced by an
anticipatory pronominal element (given in italics), which appears to the left of the
clause-final verbs.
(15) a. dat het duidelijk is [dat Marie de nieuwe voorzitter wordt]. [subject]
that it clear is that Marie the new chairman becomes
‘that it is clear that Marie will be the new Chair.’
b. dat Jan het niet gemeld heeft [dat hij weg zou gaan]. [direct object]
that Jan it not reported has that he away would go
‘that Jan didnʼt report it that heʼd go away.’
c. dat Peter erover klaagt [dat het regent]. [prepositional object]
that Peter about.it complains that it rains
‘that Peter complains about it that it rains.’
The distribution of anticipatory pronominal elements is rather complex: Sections
5.1.2 to 5.1.4 will show that in many cases it is optional, but there are also cases in
which it must or cannot occur. In addition, the presence or absence of the
pronominal element may affect the syntactic behavior of argument clauses: example
(16b), for example, shows that object clauses only allow wh-extraction if there is no
anticipatory pronoun; see, e.g., Bennis (1986:ch.2)
(16) a. dat Jan (het) zei [dat Peter een nieuwe auto gekocht had].
that Jan it said that Peter a new car bought had
‘that Jan said (it) that Peter had bought a new car.’
b. Wati zei Jan (*het) [dat Peter ti gekocht had]?
what said Jan it that Peter bought had
‘What did Jan say that Peter had bought?’
If the anticipatory pronominal element is optional, its presence may trigger a
somewhat different reading: sentence (16a) without the pronoun het presents the
proposition expressed by the embedded clause as new information; (16a) with the
pronoun, on the other hand, presents the embedded proposition as old information
and adds to this that Jan was the source of the information. In cases such as (17),
the presence of the anticipatory pronoun may trigger a factive reading of the object
clause: example (17a) simply presents the proposition expressed by the embedded
clause as new information, which may or may not be true, whereas (17b) presents
this proposition as familiar truthful information.
646 Syntax of Dutch: Verbs and verb phrases

(17) a. Jan heeft me gisteren verteld [dat hij decaan wordt].

Jan has me yesterday told that he dean becomes
‘Jan told me yesterday that heʼll become dean of the faculty.’
b. Jan heeft het me gisteren verteld [dat hij decaan wordt].
Jan has it me yesterday told that he dean becomes
‘Jan told me yesterday that heʼll become dean of the faculty.’
A similar contrast can be found in the passive counterparts of the examples in (17)
in (18): the impersonal passive with the °expletive er ‘there’ in (18a) presents the
proposition expressed by the embedded clause as new information that may be true
or false, whereas the personal passive with the anticipatory subject pronoun het ‘it’
in (18b) presents it as familiar and true; see Haeseryn et al. (1997:1138) for similar
intuitions. A more detailed description of the distribution of expletive er ‘there’ and
the anticipatory subject pronoun het ‘it’ will be provided in Section 5.1.3, sub III.
(18) a. Er werd me gisteren verteld [dat hij decaan wordt].
there was me yesterday told that he dean becomes
‘I was told yesterday that heʼll become dean of the faculty.’
b. Het werd me gisteren verteld [dat hij decaan wordt].
it was me yesterday told that he dean becomes
‘I was told yesterday that heʼll become dean of the faculty.’
The question as to whether a factive reading arises is, however, more complex than
the examples in (17) and (18) suggest. Examples (19a&b) show that regardless of
the presence or absence of the anticipatory pronoun, the truth of propositions
expressed by clausal objects of typically factive verbs like betreuren ‘to regret’ will
normally be presupposed by the speaker, whereas the truth of propositions
expressed by clausal objects of a typically non-factive verb like beweren ‘to claim’
will normally be left open. It is only with neutral verbs like vertellen ‘to tell’, which
can be used both as factive and as non-factive verbs, that the presence of the
anticipatory pronoun het will normally trigger the factive reading.
(19) a. Jan betreurt (het) [dat Marie ontslagen is]. [factive]
Jan regrets it that Marie fired is
‘Jan regrets (it) that Marie has been fired.’
b. Jan beweert (het) [dat Marie ontslagen is]. [non-factive]
Jan claims it that Marie fired is
‘Jan claims (it) that Marie has been fired.’
c. Jan vertelde me [dat Marie ontslagen is]. [non-factive]
Jan told me that Marie fired is
‘Jan told me that Marie has been fired.’
c. Jan vertelde het me [dat Marie ontslagen is]. [factive]
Jan told it me that Marie fired is
‘Jan told it to me that Marie has been fired.’
Because the semantic effect of the anticipatory pronoun het is sometimes difficult to
pinpoint even with neutral verbs like vertellen, we will not digress on this issue and
leave further investigation of it to future research.
 Argument and complementive clauses 647

Observe finally that the frequency of the anticipatory pronoun het is much
higher with typically factive verbs like betreuren ‘to regret’ than with non-factive
verbs like beweren ‘to claim’; neutral verbs like vertellen ‘to tell’ take up an
intermediate position. This is shown in Table (20) by the results of a Google search
(12/9/2011) on the strings [V-t (het) dat] and [V-de (het) dat].
(20) The realization of the anticipatory pronoun het ‘it’
ANTICIPATORY PRONOUN PRESENT ANTICIPATORY PRONOUN ABSENT
FACTIVE betreurt het dat ... 1.300.000 betreurt dat ... 300.000
regrets it that 81 % regrets that 19%
betreurde het dat ... 112.000 betreurde dat ... 42.400
regretted it that 72 % regretted that 28 %
NON-FACTIVE beweert het dat ... 120.000 beweert dat ... 1.250.000
claims it that 9% claims that 91 %
beweerde het dat ... 15.600 beweerde dat ... 548.000
claimed it that 3% claimed that 97 %
NEUTRAL vertelt het dat ... 360.000 vertelt dat ... 1.290.000
tells it that 22% claims that 78 %
vertelde het dat ... 162.000 vertelde dat ... 174.000
told it that 48 % told that 52 %

IV. Free relatives

Haeseryn et al. (1997) assume that argument clauses may also take the form of free
relative clauses. The reason for this is that we are clearly dealing with non-main
clauses functioning as arguments. That we are dealing with non-main clauses is
easily recognizable from the fact that the finite verb appears in clause-final position;
that we are dealing with arguments is clear from the fact that free relatives may
function as subject, direct object and part of a prepositional object.
(21) a. [Wie dit leest] is gek. [subject]
who this reads is crazy
‘Anyone who reads this is crazy.’
b. Jan prijst [wie hij bewondert]. [direct object]
Jan praises who he admires
‘Jan praises whoever he admires.’
c. Jan wil wachten [op wat Els te zeggen heeft]. [PO-object]
Jan wants.to wait for what Els to say has
‘Jan wants to wait for whatever Els has to say (about it).’
The question we want to raise here, however, is whether free relatives exhibit the
behavior typical of argument clauses. There may be good reasons for answering this
question in the negative and for assuming that free relatives are nominal in nature.
The first reason is that they normally refer to entities and not to propositions. This
would also account for the fact that free relatives can readily be used as indirect
objects, whereas declarative and interrogative argument clauses cannot.
648 Syntax of Dutch: Verbs and verb phrases

(22) a. Jan gaf [wie erom vroeg] een gesigneerde foto.

Jan gave who for.it asked a signed picture
‘Jan gave a signed picture to anyone who asked for it.’
b. Jan gaf een gesigneerde foto aan [wie erom vroeg].
Jan gave a signed picture to who for.it asked
‘Jan gave a signed picture to anyone who asked for it.’
Secondly, the examples in (23) show that free relatives may occur in the argument
positions in the °middle field of the clause, which are normally not available to
declarative and interrogative argument clauses.
(23) a. dat [wie dit leest] gek is. [subject]
that who this reads crazy is
‘that anyone who reads this is crazy.’
b. dat Jan [wie hij bewondert] prijst. [direct object]
that Jan who he admires praises
‘that Jan praises whoever he admires.’
c. dat Jan [op wat Els te zeggen heeft] wil wachten. [PO-object]
that Jan for what Els to say has wants.to wait
‘that Jan wants to wait for whatever Els has to say (about it).’
Thirdly, the examples in (24) show that the use of the anticipatory elements het and
erop is impossible.
(24) a. *dat heti gek is [wie dit leest]i. [subject]
that it crazy is who this reads
b. *dat Jan heti prijst [wie hij bewondert]i. [direct object]
that Jan it praises who he admires
c. *dat Jan eri op wacht [wat Els te zeggen heeft]i. [PO-object]
that Jan there for waits what Els to say has
Fourthly, the examples in (25) show that °extraposition of the free relatives only
yields an acceptable result if they function as direct objects. Not also that the
prepositional object clause may only be in extraposed position if it pied-pipes the
preposition, although this would normally give rise to a marked result with finite
prepositional object clauses; cf. ??dat Jan wacht op dat Els iets zegt ‘that Jan is
waiting for that Els says something’.
??
(25) a. dat gek is [wie dit leest]. [subject]
that crazy is who this reads
b. dat Jan prijst [wie hij bewondert]. [direct object]
that Jan praises who he admires
c. dat Jan <*op> wacht <op> [wat Els te zeggen heeft]. [PO-object]
that Jan for waits what Els to say has
The behavior displayed in examples (22)-(24) is what we attribute to nominal but
not to clausal arguments. The only fact that is perhaps not immediately expected is
that free relatives functioning as direct objects may follow the clause-final verbs,
but this would follow if we assume that free relatives exhibit similar extraposition
 Argument and complementive clauses 649

behavior as the regular relative clauses with an overt antecedent (here: iedereen and
hetgeen) in (26). However, this suggestion leaves unexplained why (25c) is
unacceptable with the preposition op stranded in preverbal position.
?
(26) a. dat iedereen gek is [REL-CLAUSE die dit leest].
that everyone crazy is who this reads
‘that everyone who reads this is mad.’
b. dat Jan iedereen prijst [REL-CLAUSE die hij bewondert].
that Jan everyone praises who he admires
‘that Jan praises everyone whom he admires.’
c. dat Jan op hetgeen wacht [REL-CLAUSE dat Els te zeggen heeft].
that Jan for the.things waits that Els to say has
‘that Jan is waiting for the things that Els has to say.’
We conclude from the discussion that free relatives are nominal in nature and
should therefore not be included in our discussion of argument clauses. We refer to
Section N3.3.2.2 for a discussion of free relatives.

5.1.2. Direct object clauses

This section investigates finite direct object clauses. Section 5.1.2.1 deals with a
number of verb classes that may select such object clauses, as well as the semantic
restrictions that may be imposed on them by the individual members of these
classes; example (27) shows, for instance, that verbs like zeggen ‘to say’ and
vragen ‘to ask’ differ in that the former selects declarative while the latter selects
interrogative clauses.
(27) a. Jan zei [dat de bibliotheek gesloten was]. [declarative]
Jan said that the library closed was
‘Jan said that the library was closed.’
b. Peter vroeg [of de bibliotheek open was]. [interrogative]
Peter asked whether the library open was
‘Peter asked whether the library was open.’
The unmarked position of object clauses is in clause-final position, but Section
5.1.2.2 will show that they may also occur in sentence-initial position. The only
option blocked in clauses with a neutral intonation pattern is that they are placed in
the °middle field of the °matrix clause (the order in (28c) improves when
contrastive accent is placed on the adjective zwanger).
(28) a. Jan heeft daarnet nog beweerd [dat Marie zwanger is].
Jan has just.now still claimed that Marie pregnant is
‘Jan has claimed only just now that Marie is pregnant.’
b. [Dat Marie zwanger is] heeft Jan daarnet nog beweerd.
c. *?Jan heeft [dat Marie zwanger is] daarnet nog beweerd.
Factive constructions, that is, constructions in which the truth of the embedded
clause is presupposed by the speaker, are a systematic exception to this general rule,
as shown in (29). Since factivity deserves closer attention, it will be investigated in
greater detail in Section 5.1.2.3.
650 Syntax of Dutch: Verbs and verb phrases

(29) a. Marie zal vanmiddag bekend maken [dat zij zwanger is].
Marie will this.afternoon public make that she pregnant is
‘This afternoon, Marie will make public that sheʼs pregnant.’
b. [Dat zij zwanger is] zal Marie vanmiddag bekend maken.
c. Marie zal [dat zij zwanger is] vanmiddag bekend maken.
Section 5.1.2.4 concludes the discussion of finite direct object clauses by dealing
with the three types of reported speech illustrated in (30). Contrary to what is
frequently claimed, there are reasons for assuming that direct and semi-direct
quotes do not necessarily function as direct object clauses of the matrix verbs.
(30) a. Marie zei [dat zij zwanger is]. [indirect reported speech]
Marie said that she pregnant is
‘Marie said that sheʼs pregnant.’
b. Marie zei: “Ik ben zwanger.” [direct reported speech]
Marie said I am pregnant
c. Marie zei ze was zwanger. [semi-direct reported speech (erlebte rede)]
Marie said she was pregnant

5.1.2.1. Selection restrictions on finite direct object clauses

Finite direct object clauses can be selected by a wide range of verbs. Providing an
exhaustive enumeration is virtually impossible, but example (31) serves to provide
a small, but representative sample of verbs that can do so.
(31)  Verb types that take a finite direct object clause
a. Verbs of communication: aankondigen ‘to announce’, beloven ‘to promise’,
bevelen ‘to command’, mailen ‘to text’, roepen ‘to call’, schrijven ‘to write’,
melden ‘to report’, smeken ‘to beg’, vertellen ‘to tell’, verzoeken ‘to request’,
vragen ‘to ask’, zeggen ‘to say’
b. Verbs of perception: horen ‘to hear’, kijken ‘to look’, luisteren ‘to listen’,
proeven ‘to taste’, ruiken ‘to smell’, voelen ‘to feel’, zien ‘to see’
c. Verbs of cognition: betwijfelen ‘to doubt’, begrijpen ‘to understand’,
doorhebben ‘to see through’, geloven ‘to believe’, overwegen ‘to consider’,
voorzien ‘to expect’, vermoeden ‘to suspect’, verwachten ‘to expect’, vinden
‘to be of the opinion’, weten ‘to know’, zich inbeelden ‘to imagine’, zich
realiseren ‘to realize’, zich afvragen ‘to wonder’
d. Verbs of investigation and discovery: aantonen ‘to show’, nagaan ‘to
examine’, onderzoeken ‘to investigate’, ontdekken ‘to discover’
e. Verbs of wishing: hopen ‘to hope’, wensen ‘to wish’, willen ‘to want’
f. Verbs with subject experiencers: betreuren ‘to regret’, haten ‘to hate’,
verafschuwen ‘to loathe’, waarderen ‘to appreciate’
Direct object clauses also occur in sentences with verbs like achten and vinden ‘to
consider’, where they are semantically licensed as the °SUBJECT of an adjectival or
nominal °complementive. Note in passing that such object clauses are regularly
introduced by the anticipatory pronoun het ‘it’.
 Argument and complementive clauses 651

(32) a. Jan acht het belangrijk [dat zijn kleren netjes zijn].
Jan considers it important that his clothes neat are
‘Jan considers it important that his clothes are neat.’
a. Jan vindt het vervelend [dat zijn schoenen vies zijn].
Jan considers it annoying that his shoes dirty are
‘Jan considers it annoying that his shoes are dirty.’
b. Jan acht het een voordeel [dat zijn project later start].
Jan considers it an advantage that his project later starts
‘Jan considers it an advantage that his project starts later.’
b. Jan vindt het een schande [dat zijn project geen aandacht krijgt].
Jan considers it a disgrace that his project no attention gets
‘Jan considers it a disgrace that his project doesnʼt get any attention.’
Finite direct object clauses normally take the form of a declarative clause
introduced by the complementizer dat ‘that’, an interrogative clause with the
complementizer of ‘whether’ or an interrogative clause introduced by a wh-phrase;
examples are given in (33). The following subsections show that providing a simple
and straightforward answer to the question what determines the distribution of these
clause types is not easy: it appears to be determined by various factors, which all
seem to have a semantic component, however.
(33) a. dat Jan hoopt [dat Marie morgen komt].
that Jan hopes that Marie tomorrow comes
‘that Jan hopes that Marie will come tomorrow.’
b. dat Peter weet [of/wanneer Marie komt].
that Peter knows whether/when Marie comes
‘that Peter knows whether/when Marie will come.’
Subsection I to VI will investigate the selection restrictions imposed by the verb
types in (31) and discuss a number of factors that seem to determine these
restrictions. Subsection VII concludes with a discussion of examples such as (32)
which illustrate object clauses functioning as a SUBJECT of a complementive.
I. Verbs of communication
At first sight, it seems relatively straightforward to determine whether a verb of
communication selects a declarative or an interrogative clause. The former are
selected by verbs like zeggen ‘to say’ and aankondigen ‘to announce’, which are
used in the (a)-examples in (34) to report something that was said/announced, while
the latter are selected by ditransitive verbs like vragen ‘to ask’ and smeken ‘to beg’,
which are used in the (b)-examples to report something that was asked/requested. In
short, the choice between declarative and interrogative clauses is determined by the
speech act reported by the speaker.
(34) a. Jan zei [dat Peter ziek was].
Jan said that Peter ill was
‘Jan said that Peter was ill.’
a. Marie kondigde aan [dat Els ontslag zou nemen].
Marie announced prt. that Els resignation would take
‘Marie announced that Els would resign.’
652 Syntax of Dutch: Verbs and verb phrases

b. Jan vroeg Marie [of Peter ziek was].

Jan asked Marie whether Peter ill was
‘Jan asked Marie whether Peter was ill.’
b. Marie smeekt Els [of ze nog wat langer wil blijven].
Marie begs Els whether she yet a.bit longer want.to stay
‘Marie begged Els that she would stay a bit longer.’
Closer inspection reveals the situation to be more complex than this. The above
only holds in cases of indirect reported speech; in other contexts verbs like zeggen
and aankondigen may also select interrogative clauses, as is shown by the two (b)-
examples in (35). The choice between the three examples depends on the speaker’s
knowledge state. Example (35a) is used when the speaker knows that there will be a
reorganization, but does not know whether Marie has made this public. Example
(35b) is normally used when the speaker does not know for certain whether or not
there will be a reorganization, and (35b) is used when he knows that there will be a
reorganization but does not know when it will take place.
(35) a. Heeft Marie gezegd [dat het instituut gereorganiseerd zal worden]?
has Marie said that the institute reorganized will be
‘Did Marie say that the institute will be reorganized?’
b. Heeft Marie gezegd [of het instituut gereorganiseerd zal worden]?
has Marie said whether the institute reorganized will be
‘Did Marie say whether the institute will be reorganized?’
b. Heeft Marie gezegd [wanneer het instituut gereorganiseerd zal worden]?
has Marie said when the institute reorganized will be
‘Did Marie say when the institute will be reorganized?’
The examples in (36) show that the speaker has a similar choice if the sentence is
negated. The choice between the three utterances again depends on the speaker’s
knowledge state. Example (36a) can be used to express that the speaker knows that
there will be a reorganization but that Marie did not make this public or to express
that the speaker expected that Marie would announce a reorganization but that this
expectation was not borne out. Example (36b) will typically be used when the
speaker does not know for certain whether or not there will be a reorganization, and
(36b) expresses that, while the speaker is convinced that there will be a
reorganization, Marie did not give more specific information about the time when it
will take place.
(36) a. Marie heeft niet gezegd [dat het instituut gereorganiseerd zal worden].
Marie has not said that the institute reorganized will be
‘Marie didnʼt say that the institute will be reorganized.’
b. Marie heeft niet gezegd [of het instituut gereorganiseerd zal worden].
Marie has not said whether the institute reorganized will be
‘Marie didnʼt say whether the institute will be reorganized.’
b. Marie heeft niet gezegd [wanneer het instituut gereorganiseerd zal worden].
Marie has not said when the institute reorganized will be
‘Marie didnʼt say when the institute will be reorganized.’
 Argument and complementive clauses 653

Not all verbs of communication are compatible with an interrogative argument

clause if they occur in an interrogative or negative sentence. The examples in (37),
for instance, show that the verb aankondigen ‘to announce’ in (37a) does not easily
allow it, which is probably due to the fact that it is factive in the sense discussed in
Section 5.1.2.3. Observe also that there is a contrast in acceptability between
yes/no- and wh-clauses and that the latter do occasionally occur on the internet.
(37) a. Heeft Marie aangekondigd [dat/*of Els ontslag neemt]?
has Marie prt.-announced that/whether Els resignation takes
‘Has Marie announced that/*whether Els will resign?’
a. ??Heeft Marie aangekondigd [waarom Els ontslag neemt]?
has Marie prt.-announced why Els resignation takes
b. Marie heeft niet aangekondigd [dat/*of Els ontslag neemt].
Marie has not prt.-announced that/whether Els resignation takes
‘Marie hasnʼt announced that/*whether Els will resign.’
b. ??Marie heeft niet aangekondigd [waarom Els ontslag neemt].
Marie has not prt.-announced why Els resignation takes
For completeness’ sake, the examples in (38) show that verbs like vragen can
sometimes also be used with declarative argument clauses, in which case we are
dealing with a request/demand rather than a question. The two meanings can be
distinguished easily: vragen with the meaning “to ask” takes a nominal object that
alternates with an aan-PP, whereas vragen with the meaning “to request/demand”
prefers a van-PP and admits a nominal object in formal/archaic contexts only.
(38) a. Jan vroeg (aan) Marie [of/*dat Peter ziek was].
Jan asked to Marie whether/that Peter ill was
‘Jan asked Marie whether/*that Peter was ill.’
b. Jan vroeg ?(van) zijn team [dat het altijd beschikbaar was].
Jan asked of his team that it always available was
‘Jan asked of his team that they would always be available.’
II. Verbs of (direct) perception
The examples in (39) show that the perception verbs proeven ‘to taste’, ruiken ‘to
smell’ and voelen ‘to feel’ may select either a declarative or an interrogative clause.
The meaning of the verbs in the primed and the primeless examples differs in that
in the former case the subject of the perception verb senses involuntarily (in the
sense of “without conscious control”) that the state of affairs expressed by the
embedded clause holds (Yuk, the soup has gone off!), whereas in the primed
examples the subject intentionally employs his/her senses to establish whether the
state of affairs expressed by the embedded clause holds (No, don’t worry, the soup
is still fine).
(39) a. Marie proefde/rook [dat de soep bedorven was]. [involuntary]
Marie tasted/smelled that the soup tainted was
‘Marie tasted/smelled that the soup had gone off.’
a. Marie proefde/rook [of de soep bedorven was]. [voluntary]
Marie tasted/smelled whether the soup tainted was
‘Marie tasted/smelled whether the soup had gone off.’
654 Syntax of Dutch: Verbs and verb phrases

b. Jan voelde [dat de was droog was]. [involuntary]

Jan felt that the laundry dry was
‘Jan felt that the laundry was dry.’
b. Jan voelde [of de was droog was]. [voluntary]
Jan felt whether the laundry dry was
‘Jan felt whether the laundry was dry.’
It does not seem to be the case that we are dealing with two uses of one and the
same verb but with real polysemy. The reason for assuming so is that in the domain
of vision and hearing there are two specialized verbs for the two meanings: zien ‘to
see’ and horen ‘to hear’ are used for involuntary perception, whereas kijken ‘to
look’ and luisteren ‘to listen’ are used for the active involvement of vision and
hearing.
(40) a. Marie zag [dat/*of de zon scheen]. [involuntary]
Marie saw that/whether the sun shone
‘Marie saw that the sun was shining.’
a. Marie keek [of/*dat de zon scheen]. [voluntary]
Marie looked whether/that the sun shone
‘Marie looked whether the sun was shining.’
b. Jan hoorde [dat/*of de deur klapperde]. [involuntary]
Jan heard that/whether the door rattled
‘Jan heard that the door was rattling.’
b. Jan luisterde [of/*dat de deur klapperde]. [voluntary]
Jan listened whether/that the door rattled
‘Jan listened whether the door was rattling.’
That the distinction between involuntary and voluntary perception is also relevant
for the polysemous verbs proeven, ruiken and voelen is clear from the fact that
imperatives, which imply voluntary action, require these verbs to take an embedded
question.
(41) a. Proef/ruik even [of/*dat de soep nog eetbaar is]! [voluntary]
taste/smell PRT whether/that the soup yet edible is
‘Just taste/smell whether the soup is still edible.’
b. Voel even [of/*dat de was al droog is]! [voluntary]
feel PRT whether/that the laundry already dry is
‘Just feel whether the laundry is dry.’
The contrast between involuntary and voluntary perception seems quite sharp if
the argument clause is introduced by the complementizer of, but more diffuse when
introduced by a wh-phrase. The examples in (42) seem to allow both readings:
example (42c), for example, does not require that Jan purposely feels how wet the
washing was, but that he may accidently it that while putting the washing in the
cupboard.
 Argument and complementive clauses 655

(42) a. Marie proefde [welke kruiden er in de soep zaten].

Marie tasted which herbs there in the soup were
‘Marie tasted which herbs were in the soup.’
b. De hond rook [welke man cannabis bij zich had].
the dog smelled which man cannabis with REFL had
‘The dog smelled which man was in possession of cannabis.’
c. Jan voelde [hoe nat de was nog was].
Jan felt how wet the laundry still was
‘Jan felt how wet the washing still was’
That verbs of involuntary perception are compatible with embedded wh-questions is
also shown by the acceptability of the examples in (43), which contrast sharply with
the primeless examples in (40).
(43) a. Jan zag onmiddellijk [welke boeken Marie geleend had].
Jan saw immediately which books Marie borrowed had
‘Jan immediately saw which books Marie had borrowed.’
b. Jan hoorde onmiddellijk [wie de kamer binnenkwam].
Jan heard immediately who the room entered
‘Jan immediately heard who entered the room.’
A warning flag is needed here, however, given that interrogative argument
clauses are generally possible with the verbs zien ‘to see’ and horen ‘to hear’ if we
are dealing with indirect perception, as is illustrated in (44). This means that
examples such as (43) can only be used for distinguishing verbs of voluntary and
involuntary perception if we are dealing with direct perception and not with indirect
perception (e.g., on the basis of empty spaces on the book shelves or the sound of
foot steps).
(44) a. Jan ziet (aan haar gezicht) onmiddellijk [dat/of ze vrolijk is].
Jan saw from her face immediately that/whether she merry is
‘Her face shows Jan immediately that/whether sheʼs merry.’
b. Jan hoort (aan de misthoorns) [dat/of het mistig is].
Jan hears from the foghorns that/whether it misty is
‘The blast of the foghorns tells Jan that/whether it is foggy.’
In addition, the examples in (45) show that zien ‘to see’ and horen ‘to hear’ are also
fully compatible with an embedded yes/no questions if they head an interrogative or
negative sentence; in this respect they behave just like non-factive verbs of
communication like zeggen ‘to say’ discussed in Subsection I.
(45) a. Heb je gezien [dat/of de zon scheen]?
have you seen that/whether the sun shone
‘Have you seen that/whether the sun was shining?’
a. Ik heb niet gezien [dat/of de zon scheen].
I have not seen that/whether the sun shone
‘I havenʼt seen that/whether the sun was shining.’
656 Syntax of Dutch: Verbs and verb phrases

b. Heb je gehoord [dat/of de deur klapperde]?

have you heard that/whether the door rattled
‘Have you heard that/whether the door was rattling?’
b. Ik heb niet gehoord [dat/of de deur klapperde].
I have not heard that/whether the door rattled
‘I havenʼt heard that/whether the door was rattling.’
The examples in (46) show that the addition of a modal verb can have a similar
effect on the selection restrictions.
(46) a. Jan kan zien [dat/of de zon schijnt].
Jan can see that/whether the sun shines
‘Jan can see that/whether the sun is shining.’
b. Jan kan horen [dat/of de deur klappert].
Jan can hear that/whether the door rattles
‘Jan can hear that/whether the door is rattling.’
We will return to verbs of involuntary and voluntary perception in Section 5.2.3.3
where we show that they differ in yet another way: the former but not the latter may
occur in °AcI-constructions: Jan zag/*keek de zon opkomen ‘Jan saw the sun rise’.

III. Verbs of cognition

Verbs of cognition can be divided into the four groups in (47) on the basis of the
question as to whether they select a declarative or an interrogative clause.
(47) a. zich afvragen ‘to wonder’
b. geloven ‘to believe’, voorzien ‘to anticipate’, verwachten ‘to expect’, vinden
‘to be of the opinion’, zich inbeelden ‘to imagine’
c. begrijpen ‘to understand’, vermoeden ‘to suspect’, zich realiseren ‘to realize’
d. weten ‘to know’, overwegen ‘to consider’ and betwijfelen ‘to doubt’
The verb zich afvragen ‘to wonder’ in (47a) cannot be combined with a declarative
argument clause; it only occurs with interrogative clauses introduced by the
complementizer of or some wh-phrase.
(48) a. Jan vraagt zich af [of/*dat Marie dat boek gelezen heeft].
Jan wonders REFL prt. whether/that Marie that book read has
‘Jan wonders whether Marie has read that book.’
b. Jan vraagt zich af [welk boek Marie gelezen heeft].
Jan wonders REFL prt. which book Marie read has
‘Jan wonders which book Marie has read.’
The verbs in (47b) take a declarative object clause introduced by the
complementizer dat ‘that’: interrogative clauses give rise to degraded results. This
is illustrated in (50) for the verb geloven ‘to believe’.
 Argument and complementive clauses 657

(49) a. Jan gelooft [dat/*of Marie morgen niet kan komen].

Jan believes that/whether Marie tomorrow not can come
‘Jan believes that/*whether Marie canʼt come tomorrow.’
b. *Jan gelooft [wanneer Marie niet kan komen].
Jan believes when Marie not can come
c. *Jan gelooft [waarom Marie morgen niet kan komen].
Jan believes why Marie tomorrow not can come
The situation is less clear for the verbs in (47c). The examples in (50) show that the
verb begrijpen ‘to understand’ cannot take an interrogative verb introduced by the
complementizer of ‘whether’, but that interrogative clauses introduced by a wh-
phrase yield a much better result—although example (50b) is definitely marked
without the °anticipatory pronoun het ‘it’, example (50c) is fully acceptable. The
verbs vermoeden ‘to suspect’ and zich realiseren ‘to realize’ show a similar
behavior here.
(50) a. Jan begrijpt (het) [dat/*of Marie morgen niet kan komen].
Jan understands it that/whether Marie tomorrow not can come
‘Jan understands that/*whether Marie canʼt come tomorrow.’
??
b. Jan begrijpt (het) [wanneer Marie niet kan komen].
Jan understands it when Marie not can come
‘Jan understands when Marie canʼt come.’
c. Jan begrijpt (het) [waarom Marie morgen niet kan komen].
Jan understands it why Marie tomorrow not can come
‘Jan understands why Marie canʼt come tomorrow.’
The examples in (51) show that verbs of the type geloven ‘to believe’ and the type
begrijpen ‘to understand’ also behave differently if they function as the °head of an
interrogative or negative sentence: whereas the former remain incompatible with
interrogative argument clauses, the latter readily accept them.
(51) a. Heeft Jan geloofd [dat/*of/*wanneer Marie komt]?
has Jan believed that/whether/when Marie comes
‘Did Jan believe that Marie would come?’
a. Jan gelooft niet [dat/*of/*wanneer Marie komt].
Jan believes not that/whether/when Marie comes
‘Jan doesnʼt believe that Marie will come.’
b. Heeft Jan begrepen [dat/of/wanneer Marie komt]?
has Jan understood that/whether/when Marie comes
‘Did Jan understand that/whether/when Marie will come?’
b. Jan begrijpt niet [dat/of/wanneer Marie komt].
Jan understands not that/whether/when Marie come
‘Jan doesnʼt understand that/whether/when Marie will come.’
Observe that example (51b) with a declarative clause will normally be used when
the speaker wants to check whether Jan did get the relevant information that Marie
will come, whereas the use of an interrogative clause suggests that the speaker
himself does not know whether/when Marie will come and would in fact like to
have more information about it (which might be available to Jan). Similarly,
658 Syntax of Dutch: Verbs and verb phrases

example (51b) with a declarative clause expresses that Jan does not grasp the
established fact that Marie will come, whereas the (time of) Marie’s coming is left
open when begrijpen takes an interrogative argument clause.
The verbs in group (47d) seem to be compatible both with declarative and
interrogative argument clauses. We illustrate this in (52) for the verb weten.
Example (52a) is used to express that Jan is cognizant of the fact that Marie is not
able to come, and the examples in (52b&c) are used to express that Jan is able to
provide further information about whether/when Marie will come.
(52) a. Jan weet [dat Marie niet kan komen].
Jan knows that Marie not can come
‘Jan knows that Marie isnʼt able to come.’
b. Jan weet [of Marie kan komen].
Jan knows whether Marie can come
‘Jan knows whether Marie is able to come.’
c. Jan weet [wanneer Marie niet kan komen].
Jan knows when Marie not can come
‘Jan knows when Marie isnʼt able to come.’
IV. Verbs of investigation and discovery
Verbs of investigation and discovery may differ with respect to whether they select
a declarative or an interrogative clause. The former seems to be the case for, e.g.,
aantonen ‘to show’, bewijzen ‘to prove’, suggereren ‘to suggest’ and ontdekken ‘to
discover’, and the latter for nagaan ‘to examine’ and onderzoeken ‘to investigate’.
The former verbs are used especially if the proposition expressed by the argument
clause refers to an established fact and the latter when the argument clause refers to
some open question. The selection restrictions remain unchanged in interrogative
and negative sentences.
(53) a. Jan heeft aangetoond [dat/*of vette vis gezond is].
Jan has prt.-shown that/whether oily fish healthy is
‘Jan has proved that oily fish is healthy.’
a. Jan ontdekte [dat/*of zijn fiets kapot was].
Jan discovered that/whether his bike broken was
‘Jan found out that his bike was broken.’
b. Jan onderzocht [of/*dat vette vis gezond is].
Jan investigated whether/that oily fish healthy is
‘Jan investigated whether oily fish is healthy.’
b. Jan ging na [of/*dat zijn fiets kapot was].
Jan checked prt. whether/that his bike broken was
‘Jan checked whether his bike was broken.’
Question formation, negation as well as the addition of a modal verb may change
the selection restriction of verbs like aantonen/bewijzen ‘to prove’, as is clear from
the fact that the examples in (54) seem acceptable with embedded yes/no-questions;
in this respect such verbs behave just like the verbs of involuntary perception.
 Argument and complementive clauses 659

(54) a. Heeft Jan aangetoond [dat/?of vette vis gezond is]?

has Jan prt.-shown that/whether oily fish healthy is
‘Has Jan proved that oily fish is healthy?’
b. Jan heeft niet aangetoond [dat/of vette vis gezond is].
Jan has not prt.-shown that/whether oily fish healthy is
‘Jan hasnʼt proved oily fish is healthy.’
c. Jan kan aantonen [dat/of vette vis gezond is].
Jan can prt.-show that/whether oily fish healthy is
‘Jan can prove that/whether oily fish is healthy.’
V. Verbs of wishing
Verbs of wishing like hopen ‘to hope’, wensen ‘to wish’, and willen ‘to want’ are
only compatible with declarative argument clauses, irrespective of whether the
sentence they head is declarative, interrogative or negative. This is illustrated in
(55) for the verb hopen.
(55) a. Jan hoopt [dat/*of Marie morgen komt]
Jan hopes that/whether Marie tomorrow comes
‘Jan hopes that Marie will come tomorrow.’
b. Hoopt Jan [dat/*of Marie morgen komt]?
hopes Jan that/whether Marie tomorrow comes
‘Does Jan hope that Marie will come tomorrow?’
c. Jan hoopt niet [dat/*of Marie morgen komt].
Jan hopes not that/whether Marie tomorrow comes
‘Jan doesnʼt hope that Marie will come tomorrow.’

VI. Subject experiencer verbs

The primeless examples in (56) show that verbs like betreuren ‘to regret’ or
waarderen ‘to appreciate’, which select an experiencer subject, take declarative
object clauses; interrogative clauses are excluded. The primed examples show that
interrogative object clauses are also excluded when the °matrix clause is
interrogative or negative. For the benefit of the discussion that will follow in
Section 5.1.2.3 it should be pointed out that the object clauses in the primeless
examples are introduced by the anticipatory pronoun het ‘it’.
(56) a. Jan betreurde het [dat/*of hij niet kon komen].
Jan regretted it that/whether he not could come
‘Jan regretted it that he couldnʼt come.’
a. Heeft Jan het betreurd [dat/*of hij niet kon komen]?
has Jan it regretted that/whether he not could come
‘Did Jan regret it that he couldnʼt come?’
a. Jan betreurde het niet [dat/*of hij niet kon komen].
Jan regretted it not that/whether he not could come
‘Jan didnʼt regret it that he couldnʼt come.’
660 Syntax of Dutch: Verbs and verb phrases

b. Peter waardeerde het [dat/*of Els hem wou helpen].

Peter appreciated it that/whether Els him wanted help
‘Peter appreciated it that Els was willing to help him.’
b. Heeft Peter het gewaardeerd [dat/*of Els hem wou helpen]?
has Peter it appreciated that/whether Els him wanted help
‘Did Peter appreciate it that Els was willing to help him?’
b. Peter waardeerde het niet [dat/*of Els hem wou helpen].
Peter appreciated it not that/whether Els him wanted help
‘Peter didnʼt appreciate it that Els was willing to help him.’
Haeseryn et al. (1997:1155) have claimed that subject experiencer verbs like
betreuren may also take an object clause introduced by the conditional
complementizer als ‘if’; some potential cases are given in (57). Section 5.1.2.2 will
show, however, that there are reasons to reject this claim.
(57) a. Jan zou het betreuren [als hij niet kan komen].
Jan would it regret if he not can come
‘Jan would regret it if he couldnʼt come.’
b. Jan waardeert het zeer [als Els hem wil helpen].
Jan appreciates it a.lot if Els him want help
‘Jan really appreciates it if Els is willing to help him.’

VII. Finite object clauses that function as the SUBJECT of a complementive

Finite object clauses occur not only as internal arguments of verbs, but also as
SUBJECTs of °complementives, that is, in vinden- and resultative constructions. The
examples in (58) show that clause-final object clauses in vinden-constructions are
normally introduced by the anticipatory pronoun het; omission of the pronoun
results in a degraded result. It should be noted, however, that the pronoun is
optional if the complementive is topicalized, and even excluded if the object clause
is topicalized; see Section 5.1.2.2 for a discussion of topicalization of object clauses
and Section 5.1.3 for a discussion of similar behavior of subject clauses.
(58)  Vinden-construction
a. Jan vindt *(het) leuk [dat Marie morgen komt].
Jan considers it nice that Marie tomorrow comes
‘Jan considers it nice that Marie will come tomorrow.’
a. Leuk vindt Jan (het) [dat Marie morgen komt].
a. [Dat Marie morgen komt] vindt Jan (*het) leuk].
b. Peter vond *(het) interessant [dat de bal zonk].
Peter considered it interesting that the ball sank
‘Peter considered it interesting that the ball sank.’
b. Interessant vond Jan (het) [dat de bal zonk].
b. [Dat de bal zonk] vond Jan (*het) interessant.
The primeless examples in (59) show that, in contrast to what we found in the
vinden-construction, the anticipatory pronoun in resultative constructions can
normally readily be omitted. The primed examples show that the vinden- and
 Argument and complementive clauses 661

resultative construction behave in a similar fashion when the complementive or the

object clause is topicalized.
(59)  Resultative construction
a. Marie maakte (het) bekend [dat er een reorganisatie komt].
Marie made it known that there a reorganization comes
‘Marie made it known that there will be reorganization.’
a. Bekend maakte Marie (het) [dat er een reorganisatie komt].
a. [Dat er een reorganisatie komt] maakte Marie (*het) bekend.
b. Jan hield (het) verborgen [dat hij ontslagen zou worden].
Jan kept it hidden that he fired would become
‘Jan kept it a secret that he would be fired.’
b. Verborgen hield Jan (het) [dat hij ontslagen zou worden].
b. [Dat hij ontslagen zou worden] hield Jan (*het) verborgen.

VIII. Conclusion
The previous subsections have shown that the choice between declarative and
interrogative object clauses is not simply a matter of lexical selection by the matrix
verb. Specifically, it has been shown that question formation and negation may
license interrogative object clauses with a subset of the matrix verbs taking
declarative object clauses in positive declarative clauses.

5.1.2.2. The placement of finite object clauses

This section discusses the placement of finite object clauses. The most common
position for such clauses is after the clause-final verbs but they can also occur in
sentence-initial position (observe that we do not use the notion clause-initial here
for the simple reason that the initial position of embedded clauses cannot be
occupied by non-wh-phrases). Normally, finite object clauses (with the possible
exception of factive clauses discussed in Section 5.1.2.3) do not occur in the
°middle field of the clause. Subsections I to III below discuss these three options in
more detail.
(60) a. Jan heeft (het) gisteren gezegd [dat Marie ziek is]. [clause-final]
Jan has it yesterday said that Marie ill is
‘Jan said yesterday that Marie is ill.’
b. *Jan heeft gisteren [dat Marie ziek is] gezegd. [clause-internal]
Jan has yesterday that Marie ill is said
c. [Dat Marie ziek is] (dat) heeft Jan gisteren gezegd. [sentence-initial]
that Marie ill is that has Jan yesterday said
‘That Marie is ill Jan said yesterday.’
The examples in (60a&c) also show that object clauses in clause-final and sentence-
initial position differ in that the former can be preceded by the anticipatory object
pronoun het, whereas the latter can be followed by the resumptive demonstrative
pronoun dat ‘that’. We take this as a hallmark of argument clauses, and will use it
as a test to determine whether or not we are dealing with object clauses. Subsection
IV will show that according to this test conditional clauses introduced by als, which
662 Syntax of Dutch: Verbs and verb phrases

are analyzed as object clauses in Haeseryn et al. (1997:1155), are in fact adverbial
°adjuncts.

I. Extraposed position
Finite direct object clauses differ from nominal direct objects in that they must
follow the verbs in clause-final position in neutral contexts. This is illustrated in
(61): whereas the primeless examples show that nominal direct objects must
precede the main verb in clause-final position, the primed examples show that
direct object clauses can follow it.
(61) a. Jan heeft Marie <zijn belevenissen> verteld <*zijn belevenissen>.
Jan has Marie his adventures told
‘Jan has told Marie his adventures.’
a. Jan heeft Marie verteld [dat hij beroofd was].
Jan has Marie told that he robbed was
‘Jan has told Marie that he was robbed.’
b. Els zal <de gebeurtenis> onderzoeken <*de gebeurtenis>.
Els will the event investigate
‘Els will investigate the event.’
b. Els zal onderzoeken [of Jan beroofd is].
Els will investigate whether Jan robbed is
‘Els will investigate whether Jan has been robbed.’
In fact, it seems that object clauses normally follow all non-clausal constituents of
their clause including those placed after the verbs in clause-final position. This is
illustrated in (62) for, respectively, a prepositional indirect object and a temporal
adverbial phrase. The unacceptable orders improve when the object clause is
followed by an intonation break, in which case the PP/adverbial phrase would
express an afterthought.
(62) a. Jan heeft verteld <aan Marie> [dat hij beroofd was] <*aan Marie>.
Jan has told to Marie that he robbed was
‘Jan has told Marie that he was robbed.’
b. Els zal onderzoeken <morgen> [of Jan beroofd is] <*morgen>.
Els will investigate tomorrow whether Jan robbed is
‘Els will investigate tomorrow whether Jan has been robbed.’
Direct object clauses are, however, followed by extraposed adverbial clauses. This
is illustrated in the primeless examples in (63) for adverbial clauses expressing time
and reason, respectively; the number signs preceding the primed examples indicate
that these examples are only acceptable if the adverbial clause is interpreted
parenthetically, in which case it must be preceded and followed by an intonation
break. Note in passing that example (63a) is actually ambiguous; the adverbial
clauses may in principle also be construed as part of the object clause, in which case
it does not refer to the time at which John told that he was robbed, but to the time at
which the robbery took place.
 Argument and complementive clauses 663

(63) a. Jan heeft verteld [dat hij beroofd was] [direct nadat hij thuis kwam].
Jan has told that he robbed was] right after he home came
‘Jan has said that he was robbed immediately after he came home.’
a. #Jan heeft verteld [direct nadat hij thuis kwam] [dat hij beroofd was].
b. Els zal onderzoeken [of Jan beroofd is] [omdat zij het niet gelooft].
Els will investigate whether Jan robbed is because she it not believes
‘Els will investigate whether Jan has been robbed since she doesnʼt believe it.’
b. #Els zal onderzoeken [omdat zij het niet gelooft] [of Jan beroofd is].
Direct object clauses can also be followed by elements that are not part of the
sentence, like the epithet in (64a) or the afterthought in (64b). Such elements are
normally preceded by an intonation break.
(64) a. Jan heeft Marie verteld [dat hij beroofd was], de leugenaar.
Jan has Marie told that he robbed was the liar
‘Jan has told Marie that he was robbed, the liar.’
b. Els zal onderzoeken [of Jan beroofd is], (en) terecht.
Els will investigate whether Jan robbed is and with.good.reason
‘Els will investigate whether Jan has been robbed, and rightly so.’
Finite object clauses in extraposed position can be introduced by the anticipatory
pronoun het ‘it’, which we have indicated here by means of subscripts; see also
5.1.1, sub III.
(65) a. Jan zal heti Marie morgen vertellen [dat hij beroofd was]i.
Jan will it Marie tomorrow tell that he robbed was
‘Jan will tell Marie tomorrow that he was robbed.’
b. Els zal heti morgen onderzoeken [of Jan beroofd is]i.
Els will it tomorrow investigate whether Jan robbed is
‘Els will investigate tomorrow whether Jan has been robbed.’

II. Middle field

The examples in (66) show that as a general rule direct object clauses do not
precede their matrix verb in clause-final position.
(66) a. Jan heeft gisteren beweerd [dat Els gaat emigreren].
Jan has yesterday claimed that Els goes emigrate
‘Jan claimed yesterday that Els is going to emigrate.’
a. *Jan heeft [dat Els gaat emigreren] gisteren beweerd.
b. Marie zal grondig onderzoeken [of het waar is].
Marie will thoroughly investigate whether it true is
‘Marie will investigate thoroughly whether it is true.’
b. *Marie zal [of het waar is] grondig onderzoeken.
There are, however, a number of potential counterexamples to this general rule.
First, the examples in (67) show that free relative clauses can generally either
precede of follow the verbs in clause-final position. We have seen in Section 5.1.1,
sub IV, that this is one of a large number of reasons for assuming that free relatives
should not be considered argument clauses but noun phrases. Thus, the surprising
664 Syntax of Dutch: Verbs and verb phrases

thing is that example (67a) is acceptable, but it can be accounted for by assuming
that free relatives can be in extraposed position just like relative clauses with an
overt antecedent: dat Jan de man prijst [die hij bewondert] ‘that Jan praises the
man who he admires’.
(67) a. dat Jan prijst [wie hij bewondert].
that Jan praises who he admires
‘that Jan praises whoever he admires.’
b. dat Jan [wie hij bewondert] prijst.
Secondly, we find similar ordering alternations with so-called factive verbs like
onthullen ‘to reveal’ and betreuren ‘to regret’. Although some speakers may judge
the primed examples as marked compared to the primeless examples, they seem
well-formed and are certainly much better than the primed examples in (66).
Barbiers (2000) suggests that the markedness of the primed examples is not related
to grammaticality issues but due to the fact that center-embedding of longer
constituents normally gives rise to processing difficulties.
(68) a. Jan heeft gisteren onthuld [dat Els gaat emigreren].
Jan has yesterday revealed that Els goes emigrate
‘Jan revealed yesterday that Els is going to emigrate.’
a. Jan heeft [dat Els gaat emigreren] gisteren onthuld.
b. Jan heeft nooit betreurd [dat hij taalkundige is geworden]
Jan has never regretted that he linguist has become
‘Jan has never regretted that he has become a linguist.’
b. Jan heeft [dat hij taalkundige is geworden] nooit betreurd.
The main difference between the (a)-examples in (66) and (68) is related to the truth
of the proposition expressed by the embedded clause; cf. Kiparsky & Kiparsky
(1970). Consider the examples in (69). Sentence (69a) shows that the proposition
expressed by the clausal complement of beweren ‘to claim’ in (66a) can be denied
by the speaker without any problem; the speaker does not commit himself to the
truth of the proposition, but instead attributes the responsibility for its truth to Jan.
Things are different when the speaker uses a factive verb like onthullen ‘to reveal’;
by using this verb the speaker presupposes that the proposition “Els is going to
emigrate” is true. This is clear from the fact that the denial in the second conjunct in
(69b) is surprising, to say the least.
(69) a. Jan heeft beweerd [dat Els gaat emigreren], maar dat is niet waar.
Jan has claimed that Els goes emigrate but that is not true
‘Jan has said that Els is going to emigrate, but that isnʼt true.’
b. $Jan heeft onthuld [dat Els gaat emigreren], maar dat is niet waar.
Jan has revealed that Els goes emigrate but that is not true
‘Jan has revealed that Els is going to emigrate, but that isnʼt true.’
The behavior of factive clauses deserves more attention, especially since it has been
suggested that they do not function as argument clauses. However, since discussing
this here would lead us to far afield and away from the present topic, we will return
to this in Section 5.1.2.3.
 Argument and complementive clauses 665

III. Sentence-initial position

The examples in (70) show that object clauses can readily occur in sentence-initial
position. In accordance with the general °verb-second requirement in Dutch, the
preposed clause must be immediately followed by the finite verb. Placement of
object clauses in sentence-initial position is impossible if the anticipatory pronoun
het ‘it’ is present, as will become clear by comparing the examples in (70) to those
in (65).
(70) a. [Dat hij beroofd was] zal Jan (*het) Marie morgen vertellen.
that he robbed was will Jan it Marie tomorrow tell
‘That he was robbed Jan will tell Marie tomorrow.’
b. [Of Jan beroofd is] zal Els (*het) morgen onderzoeken.
whether Jan robbed is will Els it tomorrow investigate
‘Whether Jan has been robbed Els will investigate tomorrow.’
The impossibility of het in (70) can be accounted for in at least two ways. One way
is to assume that the examples in (70) are in fact not derived by regular
topicalization, but in a similar way as the °left dislocation constructions in (71); cf.
Koster (1978).
(71) a. [Dat hij beroofd was], dat zal Jan (*het) Marie morgen vertellen.
that he robbed was that will Jan it Marie tomorrow tell
b. [Of Jan beroofd is], dat zal Els (*het) morgen onderzoeken.
whether Jan robbed is that will Els it tomorrow investigate
If we follow this line of thinking, the examples in (70) may involve a phonetically
empty pronoun pro with the same function and in the same position as the
resumptive demonstrative pronoun dat ‘that’ in (71). On this analysis, the
anticipatory pronoun cannot be realized since it is replaced by the pronoun dat/pro,
which is moved into sentence-initial position; the structures in (72) show that the
use of het is blocked because the clause-internal object position is occupied by the
°trace of the moved pronoun.
(72) a. [dat hij beroofd was]i [sentence dati zal Jan ti Marie morgen vertellen].
b. [dat hij beroofd was]i [sentence proi zal Jan ti Marie morgen vertellen].
The analysis suggested above is contested in Klein (1979), who points out that the
examples in (70) and (71) exhibit different intonation patterns: whereas the
examples in (71) involve an intonation break between the clause and the pronoun
dat, indicated here by means of a comma, the clauses in (70) are not likely to be
followed by such an intonation break. If one wants to conclude from this that the
examples in (70) must be derived by topicalization of the finite clause, we may
account for the impossibility of the pronoun het by assuming that the clause must be
moved via the regular object position in the middle field of the clause; under this
proposal the pronoun het cannot be realized because the regular object position
would be filled by a trace of the moved clause. An analysis like this raises the
question as to why finite clauses cannot surface in the regular object position; see
the discussion in Subsection II. One option would be to assume that there is a
surface condition that prohibits that argument positions are filled by non-nominal
666 Syntax of Dutch: Verbs and verb phrases

categories; see Stowell (1983), Hoekstra (1984a), and Den Dikken and Næss (1993)
for proposals to this effect. We will see in Section 5.1.3 that the same issue arises
with finite subject clauses.

IV. Apparent object clauses

Haeseryn et al. (1997:1155) claim that subject experiencer verbs like betreuren ‘to
regret’ and waarderen ‘to appreciate’ may take an object clause introduced by the
conditional complementizer als ‘if’; cf. the primeless examples in (73). As the
claim is simply postulated without any motivation, we can only guess why it is
proposed; one obvious argument in favor of this claim is that we can replace the
als-clauses by noun phrases that clearly function as direct objects; cf. the primed
examples in (73).
(73) a. Jan zou het betreuren [als zij niet kan komen].
Jan would it regret if she not can come
‘Jan would regret it if she couldnʼt come.’
a. Jan zou haar afwezigheid betreuren.
Jan would her non-attendance regret
‘Jan would regret her absence.’
b. Jan waardeert het zeer [als zij hem wil helpen].
Jan appreciates it a.lot if she him want help
‘Jan really appreciates it if sheʼs willing to help him.’
b. Jan zou haar hulp zeer waarderen.
Jan would her help a.lot appreciate
‘Jan would really appreciate her help a lot.’
There are, however, also arguments that militate against the claim that we are
dealing with object clauses in (73a&b). These involve the distribution of the
anticipatory pronoun het ‘it’ and the resumptive pronoun dat ‘that’, which are
optionally used to refer to (logical) object clauses in, respectively, extraposed and
sentence-initial position; see the discussion in Subsections I to III. The examples in
this subsection reveal that the clauses in (73a&b) display a different behavior here.
The examples in (74) indicate first of all that the anticipatory object pronoun het is
obligatory, and not optional, if the verbs betreuren and waarderen are followed by
an als-clause.
(74) a. Jan betreurde (het) [dat hij niet kon komen].
Jan regretted it that he not could come
‘Jan regretted it that he couldnʼt come.’
a. Jan zou *(het) betreuren [als hij niet kon komen].
Jan would it regret if he not could come
‘Jan would regret it if he couldnʼt come.’
b. Jan waardeerde (het) [dat Els hem wou helpen].
Jan appreciated it that Els him wanted help
‘Jan appreciated it that Els was willing to help him.’
b. Jan zou *(het) waarderen [als Els hem wil helpen].
Jan would it appreciate if Els him want help
‘Jan would appreciate it if Els is willing to help him.’
 Argument and complementive clauses 667

The examples in (75) provide further support: the anticipatory pronoun het can be
replaced by the resumptive pronoun dat in sentence-initial position with °left-
dislocated dat-clauses. The primed examples, on the other hand, show that
resumptive dat is excluded with left-dislocated conditional als-clauses.
(75) a. [Dat hij niet kon komen], dat betreurde Jan zeer.
that he not could come that regretted Jan a.lot
‘That he couldnʼt come, Jan regretted very much.’
a. *[Als hij niet kan komen], dat zou Jan zeer betreuren.
if he not can come that would Jan a.lot regret
b. [Dat Els hem wou helpen], dat waardeerde Peter zeer.
that Els him wanted help that appreciated Peter a.lot
‘That Els was willing to help him, Peter appreciated very much.’
b. *[Als Els hem wil helpen], dat zou Peter zeer waarderen.
if Els him want help that would Peter a.lot appreciate
The primeless examples in (76) further show that resumptive dat is normally not
used when the dat-clause is not followed by an intonation break. The primed
examples, on the other hand, show that such constructions without dat do not arise
with als-clauses either.
(76) a. [Dat hij niet kon komen] betreurde Jan zeer.
that he not could come regretted Jan a.lot
‘That he couldnʼt come, Jan regretted very much.’
a. *[Als hij niet kan komen] zou Jan zeer betreuren.
if he not can come would Jan a.lot regret
b. [Dat Els hem wou helpen] waardeerde Peter zeer.
that Els him wanted help appreciated Peter a.lot
‘That Els was willing to help him, Peter greatly appreciated.’
b. *[Als Els hem wil helpen] zou Peter zeer waarderen.
if Els him want help would Peter a.lot appreciate
Adding an object pronoun like dat or het to the primeless examples in (76) would
make these examples ungrammatical, which may be due to the fact that the object
position is already occupied by a °trace; cf. Subsection III. Adding an object
pronoun to the primeless examples in (76), on the other hand, makes these
examples fully acceptable.
(77) a. *[Dat hij niet kon komen] betreurde Jan het/dat zeer.
that he not could come regretted Jan it/that a.lot
a. [Als hij niet kan komen] zou Jan het/dat zeer betreuren.
if he not can come would Jan it/that a.lot regret
‘If he couldnʼt come, Jan would regret it/that very much.’
b. *[Dat Els hem wou helpen] waardeerde Peter het/dat zeer.
that Els him wanted help appreciated Peter it/that a.lot
b. [Als Els hem wil helpen] zou Peter het/dat zeer waarderen.
if Els him want help would Peter it/that a.lot appreciate
‘If Els is willing to help him, Peter would greatly appreciate it/that.’
668 Syntax of Dutch: Verbs and verb phrases

The primed examples in (77) strongly suggest that conditional als-clauses and
object pronouns have different syntactic functions. This is also supported by the
fact that als-clauses in °left-dislocation constructions can be associated with the
resumptive adverbial element dan ‘then’, which also surfaces in regular conditional
constructions: cf. Als het regent, dan kom ik niet ‘If it rains, (then) I won’t come’.
Now note that the object pronoun het/dat must also be expressed when resumptive
dan is present.
(78) a. [Als hij niet kan komen], dan zou Jan *(het/dat) zeer betreuren.
if he not can come then would Jan it/that a.lot regret
‘If he canʼt come, then Jan would regret it/that very much.’
b. [Als Els hem wil helpen], dan zou Peter *(het/dat) zeer waarderen.
if Els him want help then would Peter it/that a.lot appreciate
‘If Els is willing to help him, then Peter would greatly appreciate it.’
The fact that an object pronoun must co-occur with resumptive dan conclusively
shows that object pronouns and conditional als-clauses have different (logical)
syntactic functions: object versus adverbial °adjunct. Consequently, object
pronouns cannot function as anticipatory or resumptive pronouns associated with
such als-clauses. It goes without saying that this also shows that the fact that the
conditional als-clauses in the primeless examples in (73) can apparently be replaced
by the nominal direct objects in the primed examples in (73) is not sufficient ground
for concluding that conditional als-clauses are object clauses.
The conclusion that dat- and als-clauses have different syntactic functions can
also be supported by means of the coordination facts in (79). While (79a&b) show
that two dat- and two als-clauses can easily be coordinated, (79c) shows that this is
impossible for a dat- and an als-clause. The claim that the two clause types have
different syntactic functions straightforwardly derives this.
(79) a. Jan waardeert het [[dat Marie komt] en [dat Els opbelt]].
Jan appreciates it that Marie comes and that Els prt.-calls
‘Jan appreciates it that Marie will come and that Els will ring.’
b. Jan waardeert het [[als Marie komt] en [als Els opbelt]].
Jan appreciates it if Marie comes and if Els prt.-calls
‘Jan appreciates it if Marie will come and if Els will ring.’
c. *Jan waardeert het [[als Marie komt] en [dat Els opbelt]].
Jan appreciates it if Marie comes and that Els prt.-calls
For completeness’ sake, note that the left-dislocation test can also be applied to
other cases in which one might be tempted to analyze a clause, or some other
phrase, as a direct object. For example, the phrases introduced by alsof/als in the
primeless examples in (80) resemble direct objects in that they cannot be omitted
just like that, but the fact that the left-dislocation construction does not allow the
resumptive dat but requires the manner adverb zo shows immediately that we are
dealing with adverbial phrases.
 Argument and complementive clauses 669

(80) a. Jan gedraagt zich *(alsof hij gek is).

Jan behaves REFL as.if he crazy is
‘Jan behaves as if heʼs crazy.’
a. Alsof hij gek is, zo/*dat gedraagt Jan zich.
b. Jan gedraagt zich #(als een popster)
Jan behaves REFL as a pop.star
‘Jan behaves like a pop star.’
b. Als een popster, zo/*dat gedraagt Jan zich.

5.1.2.3. Factive versus non-factive complement clause constructions

Section 5.1.2.2, sub II, has shown that finite object clauses normally do not appear
in the °middle field of the °matrix clause. The relevant examples are repeated here
as (81).
(81) a. Jan heeft gisteren beweerd [dat Els gaat emigreren].
Jan has yesterday claimed that Els goes emigrate
‘Jan said yesterday that Els is going to emigrate.’
a. *Jan heeft [dat Els gaat emigreren] gisteren beweerd.
b. Peter zal grondig onderzoeken [of het waar is].
Peter will thoroughly investigate whether it true is
‘Peter will investigate thoroughly whether it is true.’
b. *Peter zal [of het waar is] grondig onderzoeken.
There is, however, a systematic exception to this rule: the examples in (82) show
that factive verbs like onthullen ‘to reveal’ and betreuren ‘to regret’ do allow the
embedded clause to appear in the middle field. The acceptability of the primed
examples decreases when they become longer and more complex, but this simply
reflects the fact that, in general, longer constituents prefer to occur in extraposed
position.
(82) a. Jan heeft gisteren onthuld [dat Els gaat emigreren].
Jan has yesterday revealed that Els goes emigrate
‘Jan revealed yesterday that Els is going to emigrate.’
a. Jan heeft [dat Els gaat emigreren] gisteren onthuld.
b. Jan heeft nooit betreurd [dat hij taalkundige is geworden].
Jan has never regretted that he linguist has become
‘Jan has never regretted that he has become a linguist.’
b. Jan heeft [dat hij taalkundige is geworden] nooit betreurd.
The fact that factive clauses can occur in nominal argument positions was first
noticed by Kiparsky & Kiparsky (1970) and since then it has widely been assumed
that factive clauses are nominal in nature. Additional support for claiming that
factive clauses differ from argument clauses is that there are more systematic
differences between the two. The subsections below discuss some of these
differences as well as some other conspicuous properties of embedded factive
clauses.
670 Syntax of Dutch: Verbs and verb phrases

I. The truth of the embedded proposition is presupposed

The main difference between (81a) and the primeless examples in (82) is related to
the truth of the proposition expressed by the embedded clause; cf. Kiparsky &
Kiparsky (1970). Non-factive verbs are used to assert the truth of the argument
clause with varying degrees of decisiveness: by using (83a), the speaker expresses
that Jan can be held responsible for the truth of the proposition “Els is going to
emigrate”, whereas this holds only to a lesser extent when he uses (83b).
(83) a. Jan heeft beweerd [dat Els gaat emigreren]. [non-factive]
Jan has claimed that Els goes emigrate
‘Jan has said that Els is going to emigrate.’
b. Jan vermoedt [dat Els gaat emigreren]. [non-factive]
Jan suspects that Els goes emigrate
‘Jan suspects that Els is going to emigrate.’
Factive verbs, on the other hand, are used if the speaker presupposes the truth of the
proposition expressed by the embedded clause, and asserts something about it: by
using (84a), the speaker asserts about the embedded proposition “Els is going to
emigrate” that Jan revealed it and by using (84b) he asserts about the same
proposition that Peter regrets it.
(84) a. Jan heeft onthuld [dat Els gaat emigreren]. [factive]
Jan has revealed that Els goes emigrate
‘Jan has revealed that Els is going to emigrate.’
b. Peter betreurt [dat Els gaat emigreren]. [factive]
Jan regrets that Els goes emigrate
‘Jan regrets that Els is going to emigrate.’
That the speaker does not commit himself to the truth of the proposition expressed
by the argument clauses of the non-factive verbs beweren ‘to claim’ and vermoeden
‘to suspect’ in (83) is clear from the fact that he can without much ado deny that the
proposition is true. The speaker may simply think or know that the information
source is wrong, consequently, his denial of the proposition “Els is going to
emigrate” in the examples in (85) leads to a semantically coherent result.
(85) a. Jan heeft beweerd [dat Els gaat emigreren], maar dat is niet waar.
Jan has claimed that Els goes emigrate but that is not true
‘Jan has claimed that Els is going to emigrate, but that isnʼt true.’
b. Jan vermoedt [dat Els gaat emigreren], maar dat is niet waar.
Jan suspects that Els goes emigrate but that is not true
‘Jan suspects that Els is going to emigrate, but that isnʼt true.’
Things are different in sentences such as (84) with the factive verbs onthullen ‘to
reveal’ or betreuren ‘to regret’; by using these verbs the speaker expresses that he
himself considers the proposition “Els is going to emigrate” to be true, and the
denial of this proposition in the examples in (86) therefore leads to semantically
incoherent or at least surprising results.
 Argument and complementive clauses 671

(86) a. $Jan heeft onthuld [dat Els gaat emigreren], maar dat is niet waar.
Jan has revealed that Els goes emigrate but that is not true
‘Jan has revealed that Els is going to emigrate, but that isnʼt true.’
b. $Jan betreurt [dat Els gaat emigreren], maar dat is niet waar.
Jan regrets that Els goes emigrate but that is not true
‘Jan regrets that Els is going to emigrate, but that isnʼt true.’

II. Properties of factive verbs

The question as to whether a complement clause does or does not allow a factive
reading depends mainly on the meaning of the verb/predicate in the matrix clause.
In (87) we provide some examples of predicates that are typically used in factive or
non-factive contexts, as well as some predicates that can comfortably be used in
either context; see Kiparsky & Kiparsky (1970) for a similar list for English.
(87) a. Non-factive verbs: beweren ‘to claim’, concluderen ‘to conclude’,
veronderstellen ‘to suppose’, denken ‘to think’, hopen ‘to hope’, vinden
‘to consider’, volhouden ‘to maintain’, zich verbeelden ‘to imagine’
b. Factive verbs: begrijpen ‘to comprehend’, betreuren ‘to regret’, duidelijk
maken ‘to make clear’, negeren ‘to ignore’, onthullen ‘to reveal’, toegeven
‘to admit’, toejuichen ‘to applaud’, vergeten ‘to forget’, weten ‘to know’
c. Verbs that can be factive or non-factive; vertellen ‘to tell’, bekennen ‘to
admit/confess’, erkennen ‘to admit’, geloven ‘to believe’, ontkennen ‘to
deny’, vermoeden ‘to suspect’, verwachten ‘to expect’, voorspellen ‘to
predict’
Kiparsky & Kiparsky (1970) propose various tests that can be used to determine
whether or not we are dealing with a factive verb/predicate. Some of these appeal to
specific properties of English, so we will only discuss those tests that make the
desired distinction for Dutch as well. We will also discuss a number of tests
proposed in Barbiers (2000).

A. Paraphrase by het feit dat ... ‘the fact that ...’

One way of making visible that the truth of the embedded proposition is
presupposed is by making use of a paraphrase with the nominal object het feit ‘the
fact’; the contrast in the examples in (88) shows that addition of the noun phrase is
impossible if the embedded clause is non-factive, but normally acceptable (albeit
sometimes clumsy) if it is factive.
(88) a. *Jan heeft het feit beweerd [dat Els gaat emigreren]. [non-factive]
Jan has the fact claimed that Els goes emigrate
Intended reading: ‘Jan has claimed that Els is going to emigrate.’
b. Jan heeft het feit onthuld [dat Els gaat emigreren]. [factive]
Jan has the fact revealed that Els goes emigrate
‘Jan has revealed the fact that Els is going to emigrate.’
Since the direct object in (88b) is the discontinuous phrase het feit dat Els gaat
emigreren, it need not surprise us that Kiparsky & Kiparsky (1970) have proposed
that underlyingly factive clauses are noun phrases. If true, it would immediately
672 Syntax of Dutch: Verbs and verb phrases

account for the fact that factive clauses can be placed in clause-internal position,
given that the clausal complement of feit can also be placed immediately after the
noun. Observe that the complex noun phrases may either follow or precede the
adverb waarschijnlijk ‘probably’; this will become relevant later in our discussion.
(89) a. Jan heeft waarschijnlijk [het feit [dat Els gaat emigreren]] onthuld.
Jan has probably the fact that Els goes emigrate revealed
‘Jan has probably revealed (the fact) that Els is going to emigrate.’
b. Jan heeft [het feit [dat Els gaat emigreren]] waarschijnlijk onthuld.
Jan has the fact that Els goes emigrate probably revealed
‘Jan has probably revealed (the fact) that Els is going to emigrate.’

B. Negation does not affect the presupposed truth of a factive clause

Negation of the examples in (83) and (84) has different consequences for the truth
of the proposition expressed by the embedded clauses. Consider the negated
counterparts of the (a)-examples, given in (90).
(90) a. Jan heeft niet beweerd [dat Els gaat emigreren]. [non-factive]
Jan has not claimed that Els goes emigrate
‘Jan hasnʼt claimed that Els is going to emigrate.’
b. Jan heeft niet onthuld [dat Els gaat emigreren]. [factive]
Jan has not revealed that Els goes emigrate
‘Jan hasnʼt revealed that Els is going to emigrate.’
The addition of negation to the non-factive construction in (90a) has the effect that
the truth of the embedded proposition is no longer asserted. The presupposed truth
of the embedded proposition in (90b), on the other hand, is not affected; the speaker
still implies that the proposition “Els is going to emigrate” is true. Observe that the
use of negation leads to an incoherent pragmatic result with the factive verb weten
‘to know’ in simple present constructions with a first person subject: by using
example (91c) the speaker expresses that he has no knowledge of the truth of a
proposition he presupposes to be true. This problem, of course, does not arise in
(91a&b) given the speaker can readily assert that some other person/the speaker-in-
the-past was not aware of the truth of this proposition.
(91) a. Jan weet niet [dat Els gaat emigreren].
Jan knows not that Els goes emigrate
‘Jan doesnʼt know that Els is going to emigrate.’
b. Ik wist niet [dat Els gaat emigreren].
I knew not that Els goes emigrate
‘I didnʼt know that Els is going to emigrate.’
c. $Ik weet niet [dat Els gaat emigreren].
I know not that Els goes emigrate
‘I donʼt know that Els is going to emigrate.’

C. Questioning does not affect the presupposed truth of a factive clause

The formation of a yes/no-question, as in (92), reveals a similar contrast as the
addition of negation: example (92a) no longer asserts the truth of the embedded
 Argument and complementive clauses 673

proposition “Els is going to emigrate”, whereas the presupposed truth of this

proposition is not affected by question formation in (92b).
(92) a. Heeft Jan beweerd [dat Els gaat emigreren]? [non-factive]
has Jan claimed that Els goes emigrate
‘Did Jan claim that Els is going to emigrate?’
b. Heeft Jan onthuld [dat Els gaat emigreren]? [factive]
has Jan revealed that Els goes emigrate
‘Did Jan reveal that Els is going to emigrate?’
Like negation, questioning leads to an incoherent pragmatic result with the factive
verb weten ‘to know’ in simple present constructions with a first person subject: by
using example (93c) the speaker is asking whether he himself has knowledge of the
truth of a proposition he presupposes to be true. This problem, of course, does not
arise in (93a&b) since the speaker can readily ask whether some other person is or
whether the speaker-in-the-past was aware of the truth of this proposition.
(93) a. Weet Jan [dat Els gaat emigreren]?
knows Jan that Els goes emigrate
‘Does Jan know that Els is going to emigrate?’
b. Wist ik (toen) [dat Els gaat emigreren]?
knew I then that Els goes emigrate
‘Did I know then that Els is going to emigrate?’
c. $Weet ik [dat Els gaat emigreren]?
know I that Els goes emigrate
‘Do I know that Els is going to emigrate?’

D. Question-answer pairs
Consider the question-answer pairs in (94). The answers in the (a)-examples show
that non-factive verbs can be used perfectly easily when the speaker wants to
diminish his responsibility for the correctness of the answer or to attribute the
responsibility for the correctness of the answer to some other person. The (b)-
examples, on the other hand, show that factive verbs cannot be used in the syntactic
frame “subject + V + answer” at all. See Section 5.1.5, sub II, for more discussion
of question-answer pairs such as (94).
(94) Wie gaat er emigreren?
who goes there emigrate
‘Who is going to emigrate?’
a. Ik denk/vermoed Els. a. Jan zei net Els. [non-factive]
I think/suspect Els Jan said just.now Els
‘Els, I think/suspect.’ ‘Els, Jan said just now.’
b. *Ik onthul Els. b. *Jan onthulde net Els. [factive]
I reveal Els Jan revealed just.now Els
The question-answer pairs in (95) show that we find a similar contrast between non-
factive and factive verbs in the answers to yes/no-questions: whereas the non-
factive verbs in the (a)-answer can be combined with a polar phrase van niet/wel
(literally: of + negative/affirmative marker”), the factive verbs in the (b)-answers
674 Syntax of Dutch: Verbs and verb phrases

cannot. For a more extensive discussion of such polar phrases we refer to Section
5.1.2.4, sub IIIB.
(95) Gaat Els binnenkort emigreren?
goes Els soon emigrate
‘Will Els emigrate soon?’
a. Peter zegt van niet, maar ik denk van wel. [non-factive]
Peter says VAN not but I think VAN AFF
‘Peter says she wonʼt but I think she will’
b. *Jan heeft onthuld van niet/wel. [factive]
Jan has revealed VAN not/AFF
Intended reading: ‘Jan has revealed that she will (not).’
b. *Peter betreurt van niet/wel. [factive]
Peter regrets VAN not/AFF
Intended reading: ‘Peter regrets that she will (not).’

E. Wh-extraction
Non-factive and factive clauses differ in that the latter are so-called weak °islands
for wh-movement. While the primeless examples in (96) show that non-factive
clauses allow extraction of both objects and adjuncts, the primed examples show
that factive clauses allow the extraction of objects only; the °trace is used to
indicate that the wh-phrase is interpreted as part of the embedded clause. The
acceptability contrast between the two (b)-examples thus shows that factive clauses
are less transparent than non-factive clauses.
(96) a. Wati denk je [dat Peter ti gekocht heeft]? [non-factive]
what think you that Peter bought has
‘What do you think that Peter has bought?’
a. Wati betreur je [dat Peter ti gekocht heeft]? [factive]
what regret you that Peter bought has
‘What do you regret that Peter has bought?’
b. Wanneeri denk je [dat Peter ti vertrokken is]? [non-factive]
when think you that Peter left has
‘When do you think that Peter left?’
b. *Wanneeri betreur je [dat Peter ti vertrokken is]? [factive]
when regret you that Peter left has

F. Negative polarity items

That factive clauses are less transparent than non-factive clauses is also borne out
by the examples in (97). The contrast between the primeless and primed examples
shows that °negative polarity items like ook maar iets ‘anything’ or een bal (lit.: a
testicle) can be licensed by negation in the matrix clause if they are part of a non-
factive clause, but not if they are part of a factive clause. It should be noted,
however, that the strength of the argument is somewhat weakened by the fact that
this type of long-distance licensing of negative polarity items is only possible with a
limited number of non-factive verbs; see Klooster (2001:316ff.).
 Argument and complementive clauses 675

(97) a. Ik denk niet [dat Jan ook maar iets gedaan heeft]. [non-factive]
I think not that Jan OOK MAAR anything done has
‘I donʼt think that Jan has done anything.’
a. *Ik onthul niet [dat Jan ook maar iets gedaan heeft]. [factive]
I reveal not that Jan OOK MAAR anything done has
b. Ik denk niet [dat Jan (ook maar) een bal gedaan heeft]. [non-factive]
I think not that Jan OOK MAAR a testicle done has
‘I donʼt think that Jan has lifted so much as a finger.’
b. *Ik onthul niet [dat Jan (ook maar) een bal gedaan heeft]. [factive]
I reveal not that Jan OOK MAAR a testicle done has

III. Factors affecting factivity

The discussion in Subsection II may have suggested that the verb/predicate of the
matrix clause fully determines whether the embedded proposition can be construed
as factive or not. However, it seems that there are a number of additional factors
that may affect a verb’s ability to take a factive complement; in fact, Barbiers
(2000:193) claims that a factive reading can be forced upon the clausal complement
of most verbs in (87a).

A. Adverbial phrases
It is frequently not immediately obvious whether we can classify a specific verb as
factive or non-factive. For example, Kiparsky & Kiparsky (1970) take a verb such
as geloven ‘to believe’ in (98) to be non-factive, which at first sight seems to be
confirmed by the fact that placing the dependent clause in the middle field of the
matrix clause gives rise to a degraded result.
(98) a. dat Marie gelooft [dat Els gaat emigreren].
that Marie believes that Els goes emigrate
‘that Marie believes that Els is going to emigrate.’
b. *dat Marie [dat Els gaat emigreren] gelooft.
However, when we add an adverb like eindelijk ‘finally’ or nooit ‘never’, as in (99),
placement of the dependent clause in the middle field of the matrix clause becomes
much more acceptable. This indicates that it is not just the verb which determines
whether the construction is factive or not, but that the wider syntactic context also
plays a role.
(99) a. dat Marie eindelijk/nooit gelooft [dat Els gaat emigreren].
that Marie finally/never believes that Els goes emigrate
‘that Marie finally/never believes that Els is going to emigrate.’
b. dat Marie [dat Els gaat emigreren] eindelijk/nooit gelooft.

B. The anticipatory pronoun het ‘it’

Addition of the °anticipatory pronoun het may also favor a factive reading of an
embedded proposition; cf. Kiparsky & Kiparsky (1970: 165). This is very clear with
a verb such as verwachten ‘to expect’: whereas examples such as (100a) without the
anticipatory pronoun are normally used when the expectation is not borne out,
676 Syntax of Dutch: Verbs and verb phrases

examples such as (100b) with the anticipatory pronoun het are regularly used when
the expectation is fulfilled.
(100) a. Ik had verwacht [dat Els zou emigreren].
I had expected that Els would emigrate
‘Iʼd expected that Els would emigrate (but I was wrong).’
b. Ik had het verwacht [dat Els zou emigreren].
I had it expected that Els would emigrate
‘I had expected it that Els would emigrate (and you can see that I was right).’
Application of this test is not always easy, however. For example, it is not true that
factive clauses must be introduced by the anticipatory pronoun; many factive verbs
can occur without it, as will be clear from inspecting the factive constructions
discussed so far. It will also be clear from the fact that a factive reading of example
(100a) is greatly favored when we add the adverb al ‘already’, as in (101a). For
completeness’ sake, (101b) shows that al can also be added to (100b).
(101) a. Ik had al verwacht [dat Els zou emigreren].
I had already expected that Els would emigrate
‘Iʼd already expected that Els would emigrate.’
b. Ik had het al verwacht [dat Els zou emigreren].
I had it already expected that Els would emigrate
‘I had already expected it that Els would emigrate.’
Complications also arise in examples containing the pronoun het. Consider the
examples in (102) with the verb vertellen ‘to tell’, which can also be used either as
a non-factive or as a factive verb. The former is clear from (102a), which shows
that the speaker has no trouble in denying the truth of the proposition expressed by
the complement clause in the first conjunct by means of the second conjunct. The
continuation in (102b) is of course compatible with a factive interpretation.
(102) Jan heeft me verteld [dat hij decaan wordt] ...
Jan has me told that he dean becomes
‘Jan has told me that heʼll become dean of the faculty ... ’
a. ... maar dat was maar een geintje. [non-factive]
but that was just a joke
‘... but that was just a joke.’
b. ... maar dat wist ik al. [factive]
but that knew I already
‘... but I knew that already.’
Example (103) seems to support the claim that adding the anticipatory pronoun het
‘it’ to the first conjunct in (102) favors a factive reading: the continuation in (103a)
seems marked because it suggests that the speaker is contradicting himself by
denying the presupposed truth of the complement clause in the first conjunct.
 Argument and complementive clauses 677

(103) Jan heeft het me verteld [dat hij decaan wordt] ...
Jan has it me told that he dean becomes
‘Jan has told me that heʼll become dean of the faculty ... ’
a. #... maar dat was maar een geintje. [non-factive]
but that was just a joke
‘... but that was just a joke.’
b. ... maar dat wist ik al. [factive]
but that knew I already
‘... but I knew that already.’
However, giving a reliable judgment on the acceptability of (103a) is hampered by
the fact that het ‘it’ need not be interpreted as an anticipatory pronoun but can also
be used as a regular pronoun referring to some previous proposition, in which case
the postverbal clause simply repeats the contents of that proposition as some kind of
afterthought. This interpretation is especially clear when the clause is preceded by
an intonation break. The fact that this reading is possible is indicated by the number
sign #.

C. Passivization
If the presence of the anticipatory pronoun het ‘it’ really does trigger a factive
reading of the complement clause, this would be in line with the observation in
Haeseryn et al. (1997:1138) that passive constructions with factive verbs normally
take the anticipatory pronoun het ‘it’ as their subject, while passive constructions
with non-factive verbs are normally impersonal, that is, involve the °expletive er
‘there’. As English has no impersonal passive, this effect cannot be replicated in the
translations; English uses it throughout.
(104) a. Er/#Het wordt algemeen beweerd [dat Jan decaan wordt]. [non-factive]
there/it is generally claimed that Jan dean becomes
‘It is generally claimed that Jan will become dean.’
b. Het/??Er wordt algemeen toegejuicht [dat Jan decaan wordt]. [factive]
it/there is generally applauded that Jan dean becomes
‘It is generally applauded that Jan will become dean.’
Haeseryn et al. (1997) also note that the use of the pronoun het becomes fully
acceptable in (104a) if the embedded clause is preceded by an intonation break: this
triggers the regular pronominal interpretation already mentioned in connection with
(103a) where the pronoun refers to some previously given proposition, repeated by
the embedded clause as an afterthought. This is again indicated by the number sign.
Applying the passivization test to the examples in (102) and (103) and using
the continuation ... maar dat was een geintje ‘... but that was a joke’, we get the
results in the (a)-examples in (105). The use of the impersonal passive in the
primeless example gives rise to a fully coherent result but the use of the personal
passive in the primed example again has the feeling of a contradiction. But example
(105a) becomes acceptable again if the pronoun het is taken to refer to some
previous proposition, in which case the clause is preferably preceded by an
intonation break. For completeness’ sake, the (b)-examples show that the
678 Syntax of Dutch: Verbs and verb phrases

continuation with ... maar dat wist ik al ‘... but I knew that already’ is compatible
with both the impersonal and the personal passive.
(105) a. Er werd me verteld [dat hij decaan wordt], maar dat was een geintje.
there was me told that he dean becomes but that was a joke
‘I was told that heʼll become dean of the faculty but that was just a joke ’
a. #Het werd me verteld [dat hij decaan wordt], maar dat was een geintje.
it was me told that he dean becomes but that was a joke
b. Er werd me verteld [dat hij decaan wordt], maar dat wist ik al.
there was me told that he dean becomes but that knew I already
‘I was told that heʼll become dean of the faculty but I knew that already.’
b. Het werd me verteld [dat hij decaan wordt], maar dat wist ik al.
it was me told that he dean becomes but that knew I already
‘It was told to me that heʼll become dean but I knew that already.’

D. Placement of the dependent clause in the middle field of the matrix clause
The examples in (106) show that placement of the object clause in the middle field
blocks the non-factive reading; the continuation in (106a) give rise to an incoherent
reading. This shows that word order may disambiguate examples such as (102).
(106) Jan heeft me [dat hij decaan wordt] gisteren verteld ...
Jan has me that he dean becomes yesterday told
‘Jan told me yesterday that heʼll become dean of the faculty ...’
a. $... maar hij maakte maar een geintje. [non-factive]
but he made just a joke
‘... but he just made a joke.’
b. ... maar dat wist ik al. [factive]
but that knew I already
‘... but I knew that already.’

IV. The position of the factive clause in the middle field

Factive clauses occupying a position in the middle field of the matrix clause may be
separated from the verbs in clause-final position by one or more adverbs (if
present). This is illustrated in (107) by means of the modal adverb waarschijnlijk
‘probably’.
(107) a. dat Jan waarschijnlijk betreurt [dat hij taalkundige is geworden].
that Jan probably regrets that he linguist has become
‘that Jan probably regrets that he has become a linguist.’
b. *dat Jan waarschijnlijk [dat hij taalkundige is geworden] betreurt.
c. dat Jan [dat hij taalkundige is geworden] waarschijnlijk betreurt.
It should be noted that the pattern in (107) differs from the pattern that we find with
the noun phrase het feit dat ... ‘the fact that ...’ in (108). As (107b) and (108b) differ
in acceptability, this can be taken as a potential problem for the hypothesis in
Kiparsky & Kiparsky (1970) that factive clauses are reduced noun phrases.
 Argument and complementive clauses 679

(108) a. dat Jan waarschijnlijk het feit betreurt [dat hij taalkundige is geworden].
that Jan probably the fact regrets that he linguist has become
‘that Jan probably regrets the fact that he has become a linguist.’
b. dat Jan waarschijnlijk het feit [dat hij taalkundige is geworden] betreurt.
c. dat Jan het feit [dat hij taalkundige is geworden] waarschijnlijk betreurt.
One way to approach this problem for Kiparsky & Kiparsky’s hypothesis might be
to claim that the word order difference between (107a) and (107c) suffices to make
the information-structural distinction between °focus (“discourse-new information”)
and presupposition (“discourse-old information”), whereas in (108) this distinction
rather relies on the position on the nominal part het feit; see Section N8.1.3 for
discussion. It remains to be seen, however, whether this line of thinking would lead
to a fully satisfactory account of the contrast between (107) and (108).

V. Wh-extraction from factive clauses

If we accept the suggestion from Section N8.1.3 that the word order in (108c) is
derived by leftward movement of the nominal object, it seems rather attractive to
assume that the order in (107c) is derived by leftward movement of the factive
clause. An empirical argument in favor is that we may now appeal to the °freezing
effect: the factive clause is a strong °island for wh-extraction if part of the middle
field of the matrix clause, but not if it follows the verbs in clause-final position.
(109) a. Welki boek heeft Jan altijd betreurd [dat hij ti niet gekocht heeft]?
which book has Jan always regretted that he not bought has
‘Which book has Jan always regretted that he hasnʼt bought?’
b. *Welki boek heeft Jan [dat hij ti niet gekocht heeft] altijd betreurd?
which book has Jan that he not bought has always regretted
Recall from Subsection IIE, however, that factive clauses are weak islands in the
sense that wh-extraction is restricted to nominal objects; wh-extraction of, e.g.,
adverbial phrases is excluded irrespective of the position of the factive clause; this
is illustrated again in (110).
(110) a. *Waari heeft Jan altijd betreurd [dat hij ti zijn boek gepubliceerd heeft]?
where has Jan always regretted that he his book published has
b. *Waari heeft Jan [dat hij ti zijn boek gepubliceerd heeft] altijd betreurd?
where has Jan that he his book published has always regretted
The observation that factive clauses exhibit the behavior of weak islands is actually
another problem for Kiparsky & Kiparsky’s hypothesis that factive clauses are
reduced noun phrases; complex noun phrases are generally strong islands in the
sense that they also block extraction of nominal objects from their clausal
complement. The examples in (111) show that this holds irrespective of whether the
clause precedes or follows the verbs in clause-final position.
(111) a. *Welki boek heeft Jan altijd het feit betreurd [dat hij ti niet gekocht heeft]?
which book has Jan always the fact regretted that he not bought has
b. *Welki boek heeft Jan altijd het feit [dat hij ti niet gekocht heeft] betreurd?
which book has Jan always the fact that he not bought has regretted
680 Syntax of Dutch: Verbs and verb phrases

VI. The syntactic status of factive clauses

So far, we have more or less adopted Kiparsky & Kiparsky’s hypothesis that factive
clauses are reduced noun phrases, but Subsections IV and V have discussed a
number of potential problems for this hypothesis. So, it might be advisable to look
for another analysis to account for the differences in behavior between non-factive
and factive clauses. One such analysis is provided in Barbiers (2000), who argues
that while non-factive clauses are complements of the verb, factive clauses are
°adjuncts. This proposal is interesting because it would immediately account for the
fact that factive clauses can occur in the middle field of the clause, given that this is
generally possible with adjunct clauses, as is shown by the examples in (112).
(112) a. dat Peter [nadat hij afscheid genomen had] snel vertrok.
that Peter after he leave taken had quickly left
‘that Peter left quickly after heʼd said good-bye.’
a. dat Peter snel vertrok [nadat hij afscheid genomen had].
b. dat Jan [omdat hij ziek was] niet kon komen.
that Jan because he ill was not could come
‘that Jan couldnʼt come because he was ill.’
b. dat Jan niet kon komen [omdat hij ziek was].
If factive clauses are indeed adjuncts, we expect them to entertain a looser relation
to the matrix verb than non-factive verbs. Barbiers claims that this expectation is
indeed borne out and he demonstrates this by pointing to the fact that non-factive
clauses must be pied-piped under °VP-topicalization, whereas factive clauses can
be stranded.
(113) a. Jan zal niet vinden [dat het probleem nu opgelost is].
Jan will not find that the problem now solved is
‘Jan wonʼt think that the problem has been solved now.’
a. *Vinden zal Jan niet [dat het probleem nu opgelost is].
b. Jan zal niet toegeven [dat het probleem nu opgelost is].
Jan will not admit that the problem now solved is
‘Jan wonʼt admit that the problem has been solved now.’
b. Toegeven zal Jan niet [dat het probleem nu opgelost is].
Another observation provided by Barbiers that may point in the same direction is
that °stranding of the clause may disambiguate examples such as (114a): whereas
(114a) can be factive (the speaker knows that Jan has been ill) or non-factive (the
speaker expects that Jan will tell a lie, e.g., to excuse his absence), example (114b)
can only have the former reading.
(114) a. Jan zal wel vertellen [dat hij ziek was]. [non-factive or factive]
Jan will PRT tell that he ill was
‘Jan will probably say that he was ill.’
b. Vertellen zal Jan wel [dat hij ziek was]. [factive only]
However, there are at least three potential problems with Barbiers’ proposal. First,
the judgments in (113) and (114) are somewhat delicate and not all speakers are
able to produce the same results. Second, as was pointed out by Barbiers himself,
 Argument and complementive clauses 681

the hypothesis does not account for the fact that factive clauses are weak (and not
strong) islands, given that adjunct clauses normally block wh-extraction of nominal
objects as well. Third, assigning adjunct status to factive clauses would lead to the
expectation that factive clauses can be omitted (which adjunct clauses generally
can), which is not borne out: *Jan betreurde. We therefore leave the question as to
whether Barbiers’ hypothesis is tenable to future research.

VII. Factive interrogative clauses

The term factivity is mostly restricted to verbs selecting declarative clauses, due to
the fact that it is defined in terms of the truth value of the proposition expressed by
sentential complements. A typical example of such a definition is found in Crystal
(1991): the term factivity is “used in the classification of verbs, referring to a verb
which takes a complement clause, and where the speaker presupposes the truth of
the proposition expressed in that clause”. The application of this definition is
illustrated again in the examples in (115), in which S1  S2 stands for “by uttering
sentence S1 the speaker presupposes that the proposition P expressed by S2 is true”.
(115) a. Jan denkt dat Els morgen vertrekt.   Els vertrekt morgen. [non-factive]
Jan thinks that Els tomorrow leaves Els leaves tomorrow
‘Jan thinks that Els is leaving tomorrow   Els is leaving tomorrow.’
b. Jan betreurt dat Els morgen vertrekt.  Els vertrekt morgen. [factive]
Jan regrets that Els tomorrow leaves Els leaves tomorrow
‘Jan regrets that Els is leaving tomorrow.  Els is leaving tomorrow.’
Definitions of this sort exclude the existence of factive verbs selecting an
interrogative complement clause: interrogative clauses differ from declaratives in
that they do not express full propositions as they are characterized by indeterminacy
in the value of some variable represented by the yes/no-operator or wh-phrase; cf.
Grimshaw (1979). Whether or not this exclusion is justified can be tested by
investigating factive verbs like vergeten ‘to forget’ and weten ‘to know’, both of
which may also take an interrogative complement clause. First, consider the
examples in (116).
(116) a. Jan weet dat Els morgen vertrekt.  Els vertrekt morgen.
Jan knows that Els tomorrow leaves Els leaves tomorrow
‘Jan knows that Els is leaving tomorrow.  Els is leaving tomorrow.’
b. Jan weet of Els morgen vertrekt.   Els vertrekt morgen.
Jan knows whether Els tomorrow leaves Els leaves tomorrow
‘Jan knows whether Els is leaving tomorrow.   Els is leaving tomorrow.’
This sentence pair indeed suggests that verbs taking an interrogative argument
clause are non-factive: by uttering sentence (116b), the speaker does not commit
himself to the truth of the proposition expressed by the sentence on the right-hand
side of the arrow. This is not surprising, of course: the speaker’s reference to Jan as
a source of more information about the truth of the proposition only makes sense if
the speaker does not know the answer to the embedded question himself.
Things seem to be different, however, with embedded wh-questions. Consider
the contrast between the examples in (117). By uttering the sentence in (117a) the
682 Syntax of Dutch: Verbs and verb phrases

speaker does not entail that the proposition “Els is leaving” is true, whereas the
speaker does entail this by uttering the sentence in (117b).
(117) a. Jan vroeg wanneer Els vertrekt.   Els vertrekt.
Jan asked when Els leaves Els leaves
‘Jan asked when Els is leaving. 
 Els is leaving.’
b. Jan weet wanneer Els vertrekt.  Els vertrekt.
Jan knows when Els leaves Els leaves
‘Jan knows when Els is leaving.  Els is leaving.’
The verbs vragen ‘to ask’ and weten ‘to know’ thus differ in that the first is clearly
non-factive, but that the second is factive in the slightly more restricted sense that
the truth of the proposition expressed by the non-wh part of the complement clause
is presupposed by the speaker. The examples in (118) show that this difference
between vragen and weten not only holds in cases in which the wh-phrase is an
adjunct of the embedded clause, but also if it is an argument.
(118) a. Jan vroeg wie er vertrekt.  Er vertrekt iemand.
Jan asked who there leaves there leaves someone
‘Jan asked who is leaving.  someone is leaving.’
b. Jan weet wie er vertrekt.  Er vertrekt iemand.
Jan knows who there leaves there leaves someone
‘Jan knows who is leaving.  someone is leaving.’
As we have seen in Subsection II, factive verbs have the property that negating or
questioning the clause they are heading does not affect the entailment, that is, the
examples in (119) have the same entailment as example (116a).
(119) a. Jan weet niet dat Els morgen vertrekt.  Els vertrekt morgen.
Jan knows not that Els tomorrow leaves Els leaves tomorrow
‘Jan doesnʼt know that Els is leaving tomorrow.  Els is leaving tomorrow.’
b. Weet Jan dat Els morgen vertrekt?  Els vertrekt morgen.
knows Jan that Els tomorrow leaves Els leaves tomorrow
‘Does Jan know that Els is leaving tomorrow?  Els is leaving tomorrow.’
The examples in (120) show that the (b)-examples in (117) and (118) likewise pass
this litmus test for factivity; by uttering the sentences on the left-hand side of the
arrow the speaker entails that the propositions expressed by the sentences on the
right-hand side of the arrows are true.
(120) a. Jan weet niet wanneer Els vertrekt.  Els vertrekt.
Jan knows not when Els leaves Els leaves
‘Jan doesnʼt know when Els is leaving.  Els is leaving.’
a. Weet Jan wanneer Els vertrekt?  Els vertrekt.
knows Jan when Els leaves Els leaves
‘Does Jan know when Els is leaving?  Els is leaving.’
 Argument and complementive clauses 683

b. Jan weet niet wie er vertrekt.  Er vertrekt iemand.

Jan knows not who there leaves there leaves someone
‘Jan doesnʼt know who is leaving.  Someone is leaving.’
b. Weet Jan wie er vertrekt?  Er vertrekt iemand.
knows Jan who there leaves there leaves someone
‘Does Jan know who is leaving?  Someone is leaving.’
The syntactic tests for factivity yield slightly equivocal results. Like the factive
declarative clause in (121a), the factive interrogative clauses in (121b&c) can be
introduced by the anticipatory pronoun het ‘it’.
(121) a. Jan weet het dat Els morgen vertrekt.
Jan knows it that Els tomorrow leaves
‘Jan knows it that Els is leaving tomorrow.’
b. Jan weet het wanneer Els vertrekt.
Jan knows it when Els leaves
‘Jan knows it when Els is leaving.’
c. Jan weet het wie er vertrekt.
Jan knows it who there leaves
‘Jan knows it who is leaving.’
However, it seems that placement of a factive complement in the middle field of the
matrix clause gives rise to a less felicitous result if the complement clause is
interrogative than if it is declarative; whereas (122a) is merely stylistically marked,
the examples in (122b&c) seem degraded (although they may improve a little with
a contrastive focus accent on the wh-word).
(122) a. dat Jan [dat Els morgen vertrekt] nog niet weet.
that Jan that Els tomorrow leaves not yet knows
‘that Jan doesnʼt yet know that Els is leaving tomorrow.’
b. ??dat Jan [wanneer Els vertrekt] nog niet weet.
that Jan when Els leaves not yet knows
‘that Jan doesnʼt yet know when Els will be leaving.’
c. ??dat Jan [wie er vertrekt] nog niet weet.
that Jan who there leaves not yet knows
‘that Jan doesnʼt yet know who is leaving.’
Note that the distinction between two types of wh-questions is not new and dates
back at least to Groenendijk & Stokhof (1984:91ff.), who phrase the distinction in
terms of pragmatic implicatures instead of factivity, that is, the speaker’s
presupposition. Since a detailed study of the syntactic behavior of factive
interrogative constructions is not yet available as far as we know, we will leave this
to future research.

VIII. Conclusion
The previous subsections have shown that there are a large number of systematic
differences between non-factive and factive clauses, which suggests that they must
also receive a different syntactic analysis. Kiparsky & Kiparsky (1970) argued that
684 Syntax of Dutch: Verbs and verb phrases

the difference should be expressed by assuming a difference in categorial status:

non-factive clauses are clausal complements and factive clauses are reduced
nominal complements. Barbiers (2000) argued that the distinction is related to
syntactic function: non-factive clauses are complements of the verb, whereas
factive clauses are adjuncts. The two proposals are embedded in a larger set of
theoretical assumptions and we have seen that they each have their own problems.
The discussion has revealed at any rate that the fact that factive clauses can occur in
the middle field of the matrix clause is not just some isolated fact but that it is part
of a wider set of facts that still needs to receive an explanation.

5.1.2.4. Reported speech

The complement clauses discussed in the preceding sections all have the form of
finite embedded clauses, that is, they are introduced by a complementizer (dat ‘that’
or of ‘whether’) or a wh-phrase, and have the finite verb in clause-final position.
Complement clauses of this kind are also found in sentences such as (123a), in
which the speaker reports what someone else has said, thought, etc. The sentences
in (123b&c) show, however, that there are also alternative ways.
(123) a. Jan zei/dacht dat hij ziek was. [indirect reported speech/quote]
Jan said/thought that he ill was
‘Jan said/thought that he was ill.’
b. Jan zei/dacht: “Ik ben ziek”. [direct reported speech/quote]
Jan said/thought I am ill
‘Jan said/thought: “Ik ben ziek”.’
c. Jan zei/dacht hij was ziek. [semi-direct reported speech/quote]
Jan said/thought he was ill
Although the examples in (123) show that constructions like these are not strictly
limited to speech proper but may also pertain to thoughts, they are normally said to
involve REPORTED SPEECH. We will therefore refer to the whole set of constructions
as reported speech constructions, and to the parts in italics, which express the
reported parts, as QUOTES. Although quotes are often analyzed as direct object
clauses (see, e.g., Haeseryn et al. 1997), we will see that this is not entirely correct
for all cases: see also Corver (1994), Corver & Thiersch (2003), and De Vries
(2006). For this reason we will refer to the clauses headed by the verb of
saying/thinking not as °matrix clauses but, more neutrally, as SAY-CLAUSES.
The way of reporting speech in (123a) is normally referred to as INDIRECT
REPORTED SPEECH. An important property of this construction is that the embedded
clause(s) does/do not necessarily correspond in a one-to-one fashion to the actual
utterance(s) of the source indicated: for example, if Jan is a very talkative person,
the embedded clause may simply summarize a story that took 30 minutes to tell,
that is, example (123a) does not imply that Jan literally said: “Ik ben ziek”. This
distinguishes indirect from DIRECT REPORTED SPEECH; example (123b) is only true
if Jan pronounced the sentence Ik ben ziek, for which reason we repeated this
sentence literally in the translation of (123b). Another difference, which is
illustrated in (124), is that direct quotes can consist of a sequence of independent
 Argument and complementive clauses 685

sentences, whereas in indirect reported speech constructions each assertion must be

realized as a separate dependent clause.
(124) a. Jan zei/dacht [[dat hij ziek was] en [dat hij thuis bleef]]. [indirect]
Jan said/thought that he ill was and that he at.home stayed
‘Jan said that he was ill and that he would stay at home.’
b. Jan zei/dacht: “Ik ben ziek. Ik blijf thuis”. [direct]
Jan said/thought I am ill I stay at.home
‘Jan said: “Ik ben ziek. Ik blijf thuis”.’
In example (123c), we are dealing with SEMI-DIRECT REPORTED SPEECH (also
known as ERLEBTE REDE), which constitutes a kind of in-between category. It
differs from indirect reported speech in that the quote has the form of a main clause.
This is clear from the position of the finite verb: if we are dealing with indirect
reported speech, the finite verb should occupy the clause-final position, whereas it
should be in second position in semi-direct reported speech. The placement of the
finite verb is clearly related to the distribution of the complementizer: the examples
in (125) show that the complementizer is obligatory in indirect reported speech
constructions with declarative quotes, but that it cannot appear in semi-direct
reported speech constructions. This also shows that semi-direct reported speech
constructions such as (125b) cannot be derived from direct reported speech
constructions such as (125a) by deletion of the complementizer dat, but that they
constitute a construction type in their own right.
(125) a. Jan zei *(dat) hij ziek was. [indirect]
Jan said that he ill was
‘Jan said that he was ill.’
b. Jan zei (*dat) hij was ziek. [semi-direct]
Jan said that he was ill
Although semi-direct reported speech does not involve a literal quote, it differs
from indirect reported speech in that the relation with what was actually said is
much tighter. Example (123c), for instance, suggests that Jan said something like Ik
ben ziek. Semi-direct quotes differ from direct quotes mainly in that first and second
person pronouns are replaced by third person pronouns and that the present tense of
the reported sentence is adapted to conform to the past tense of the verb zeggen ‘to
say’; cf. Lodewick (1975:169-70). The semi-direct equivalent of the direct reported
speech construction in (126a) would then be as in (126b).
(126) a. Jan dacht: “Ik haat je uit de grond van mijn hart”. [direct]
Jan thought I hate you from the bottom of my heart
b. Jan dacht, hij haatte hem uit de grond van zijn hart. [semi-direct]
Jan said he hated him from the bottom of his heart
Semi-direct reported speech is not often used in colloquial speech but is
regularly found as a stylistic device in modern literature, especially for expressing
the internal thoughts of the protagonist(s) of a story (the so-called interior
monologue); Lodewick in fact claims that it is a characteristic feature of
impressionistic writings from around 1900. The use of semi-direct reported speech
686 Syntax of Dutch: Verbs and verb phrases

constructions implies that, like direct quotes, semi-direct quotes may involve
sequences of sentences; this expectation is borne out, as is illustrated in (127b) by
means of the semi-direct counterpart of the direct reported speech construction in
(124b), repeated here as (127b).
(127) a. Jan zei/dacht: “Ik ben ziek. Ik blijf thuis”. [direct]
Jan said/thought I am ill I stay at.home
‘Jan said: “Ik ben ziek. Ik blijf thuis”.’
b. Jan zei/dacht, hij was ziek, hij bleef thuis. [semi-direct]
Jan said/thought he was ill he stayed at.home
Embedded clauses in indirect reported speech constructions such as (123a) can
be pronominalized (Jan zei het ‘Jan said it’), which suggests that they function as
direct object clauses. It is often assumed without much argumentation that direct
and semi-direct reported speech constructions like (123b&c) also involve direct
object clauses; see Haeseryn et al. (1997:1100). This is, however, far from obvious:
the quotes in the two examples in (127) consist of series of sentences, and this
makes it is very unlikely that quotes have the function of direct object. In fact, it
may even be the case that we are dealing with a relation of an entirely different sort
given that the part Jan zei need only be used in examples like (123b&c) when the
context leaves open what the source of the quote is; if the source is known, it can
readily be omitted. This is illustrated in the little scene in (128), which might be
used as the start of a story. See also the discussion of what Huddleston and Pullum
(2002: 1029) call free indirect/direct speech.
(128) Jan kwam in zijn pyjama de kamer binnen.
Jan came in his pajamas the room inside
‘Jan entered the room in his pajamas.’
a. (Hij dacht:) “Ik ben ziek. Ik blijf thuis”. [direct]
he thought I am ill I stay at.home
‘(He thought:) “Ik ben ziek. Ik blijf thuis”.’
b. (Hij dacht,) hij was ziek; hij bleef thuis. [semi-direct]
he thought he was ill he stayed at.home
‘(He thought,) he was ill. He would stay in.’
Matrix clauses in indirect reported speech constructions, on the other hand, can only
be left unexpressed under very special circumstances. Sentence (129b), for
example, cannot replace the continuations of the story in (128a&b), but is only
acceptable as an answer to a question such as (129a)—we are dealing with some
kind of ellipsis; the part of the answer that can be recovered from the original
question (here: the matrix clause) is simply omitted.
(129) a. Wat zei Jan?
what said Jan
‘What did Jan say?’
b. Dat hij ziek was en dat hij thuis bleef. [indirect]
that he ill was and that he at.home stayed
‘That he was ill and that he would stay in.’
 Argument and complementive clauses 687

We will see in Subsection II that this difference is reflected in several other ways,
and that there are reasons for assuming that in many cases direct and semi-direct
reported speech constructions are not regular transitive constructions. Instead, the
quotes function as full-fledged sentences with parenthetical say-clauses.
We already mentioned that semi-direct reported speech is normally used in
written language and cannot be found in colloquial speech so frequently. Subsection
III will show, however, that there is also a reported speech construction that is
normally avoided in writing but which is highly frequent in speech; cf. Verkuyl
(1977) and Romein (1999). This construction, which is illustrated in (130), involves
the QUOTATIVE PREPOSITION van followed by an intonation break, which may
optionally be preceded by a hesitation marker like eh ‘er’, and a quote. The quote
can be either direct or, less frequently, indirect; cf. Verkuyl (1977).
(130) a. Marie dacht van (eh) ... Hij komt straks wel weer terug.
Marie thought VAN er he comes later PRT again back
‘Marie thought something like: “Heʼll probably return later again”.’
b. Marie dacht van (eh) ... dat hij straks wel weer terug komt.
Marie thought VAN er that he later PRT again back comes
The three types of reported speech constructions introduced above will be discussed
in separate subsections. Subsection I discusses indirect reported speech and shows
that the indirect quote functions as a regular argument clause. Subsection II
continues with a discussion of (semi-)direct reported speech and argues that the say-
clause in such constructions is often (but not always) parenthetical. Subsection III
concludes with a discussion of the colloquial quotative van-construction in (130).

I. Indirect reported speech

Quotes in indirect reported speech constructions behave in many respects like other
types of direct object clauses. The following subsections will show this for a
number of properties of object clauses, which are discussed more extensively in
Sections 5.1.2.1 to 5.1.2.3. We will also discuss some facts not mentioned there,
which can be used to provide support for the claim that indirect quotes are regular
object clauses.

A. Selection restrictions on the embedded clause

The form of indirect quotes is determined to a large extent by the main verb: verbs
like zeggen ‘to say’ and denken ‘to think’ select declarative clauses, whereas verbs
like vragen ‘to ask’ select interrogative clauses. See Section 5.1.2.1 for a more
extensive discussion of the selection restrictions on declarative and interrogative
object clauses.
(131) a. dat Peter zei/dacht [dat Jan ziek was].
that Peter said/thought that Jan ill was
‘that Peter said/thought that Jan was ill.’
b. dat Marie vroeg [of Jan ziek was].
that Marie asked whether Jan ill was
‘that Marie asked whether Jan was ill.’
688 Syntax of Dutch: Verbs and verb phrases

B. Position of the embedded clause

The examples in (131) show that indirect quotes normally follow the verb(s) in
clause-final position—placing such quotes in the °middle field is normally marked
and triggers a factive reading; cf. Section 5.1.2.3. Topicalization of indirect quotes
is possible, in which case they are optionally followed by the resumptive pronoun
dat ‘that’; see Section 5.1.2.2 for a more extensive discussion of the placement of
direct object clauses.
(132) a. [Dat Peter ziek is] (dat) zei/dacht Jan.
that Peter ill is that said/thought Jan
b. [Of Jan ziek was] (dat) vroeg Marie.
whether Jan ill was that asked Marie

C. The use of an anticipatory pronoun

The use of an °anticipatory pronoun seems possible but marked; the examples in
(133) are more likely to be construed with a regular, discourse-related interpretation
of the pronoun, which again favors a factive reading of the embedded clause; see
Section 5.1.2.3, sub IIIB.
(133) a. dat Peter het zei/dacht [dat Jan ziek was].
that Peter it said/thought that Jan ill was
‘that Peter said/thought it that Jan was ill.’
b. dat Marie het vroeg [of Jan ziek was].
that Marie it asked whether Jan ill was
‘that Marie asked it whether Jan was ill.’

D. Wh-extraction
Embedded declarative clauses are fully transparent for wh-extraction in the sense
that both arguments and adjuncts can be extracted. See Section 5.1.1, sub III, for
discussion of the fact that wh-extraction becomes unacceptable if an anticipatory or
deictic pronoun is added.
(134) a. Wiei zei/dacht je [dat ti dat boek gekocht had]? [subject]
who said/thought you that that book bought has
‘Who did you say/think had bought that book.’
b. Wati zei/dacht je [dat Peter ti gekocht heeft]? [object]
what said/thought you that Peter bought has
‘What did you say/think that Peter has bought?’
c. Wanneeri zei/dacht je [dat Peter ti vertrokken was]? [adjunct]
when said/thought you that Peter left had
‘When did you say/think that Peter had left?’
Wh-extraction is not possible from embedded interrogative clauses. The standard
analysis in generative grammar is that this is due to the fact that wh-extraction
cannot apply in one fell swoop but must proceed via the clause-initial position of
the object clause; this position is available in declarative examples such as (134),
but occupied by a wh-phrase in embedded wh-questions such as (135) or a
phonetically empty question °operator in embedded yes/no-questions.
 Argument and complementive clauses 689

(135) a. *Wiei vroeg je [watj ti tj gekocht heeft]? [subject]

who asked you what bought has
Compare: ‘*Who did you ask what has bought?’
b. *Watj vroeg je [wiei ti tj gekocht heeft]? [object]
what asked you who bought has
Compare: ‘*What did you ask who has bought?’
c. *Wanneerj vroeg je [wiei ti tj vertrokken was]? [adjunct]
when asked you who left had
Compare: ‘*When did you ask who had left?’
Note in passing that, contrary to what has been reported for English, wh-extraction
of the subject in (134a) is acceptable despite the presence of a complementizer and
that most Dutch speakers find the three examples in (135) equally unacceptable. We
will not digress on these issues here but refer the reader instead to Section 11.3.1
for an extensive discussion of the restrictions on wh-extraction. Note also that
example (135c) is fully acceptable if wanneer ‘when’ is construed as a modifier of
the matrix clause but this is, of course, not the reading intended here (as is indicated
by the placement of the °trace tj within the embedded clause).

E. Binding
Referential personal pronouns as part of an indirect quote can be bound by an
antecedent in the say-clause; see Section N5.2.1.5, for an extensive discussion of
°binding of such pronouns. Since such pronouns can also co-refer with some
referential expression as a result of °accidental coreference, we have to appeal to
examples in which the antecedent is a quantified expression like iedereen
‘everyone’ or niemand ‘nobody’ in order to show this. Example (136a) first shows
that the pronoun hij cannot be used as referentially dependent on a
universally/negatively quantified expression if the latter is part of some other
sentence; in such cases, the pronoun must refer to some known entity in the domain
of discourse. The fact that the pronoun can have a °bound variable reading, that is,
can be interpreted as referentially dependent on the quantifiers in (136b) shows that
we are not dealing with accidental coreferentiality but with binding. Italics indicate
the intended binding relation.
(136) a. *Iedereen/Niemand bleef thuis. Hij was ziek.
everybody/nobody stayed at.home he was ill
b. Iedereen/Niemand zei [dat hij ziek was].
everybody/nobody said that he ill was
‘Everybody/Nobody said that he was ill.’
The acceptability of the bound variable reading in (136b) unambiguously shows
that we are dealing with an object clause; if the indirect quote were not the object of
the verb zeggen ‘to say’, there would be no °c-command relation between the
subject of the say-clause and the pronoun and, consequently, binding would be
wrongly predicted to be impossible, just as in (136a).
690 Syntax of Dutch: Verbs and verb phrases

F. Licensing of negative polarity items

That indirect quotes are object clauses is also shown by the fact that °negative
polarity items (NPIs) like ook maar iets ‘anything’ as part of an indirect quote can
be licensed by some negative element in the say-clause. The reason is that, like
binding, NPI licensing requires c-command between the NPI and its licenser. NPI-
licensing is excluded in (137a) since the NPI and its potential licenser niemand
‘nobody’ are not in the same sentence and there is consequently no c-command
relation between them; NPI-licensing is possible in (137b) since the subject of the
matrix clause does c-command the NPI in the embedded object clause. Italics
indicate the relation between the NPI and its intended licenser.
(137) a. Niemand bleef thuis. *Hij had daar ook maar iets te doen.
nobody stayed at.home he had there anything to do
b. Niemand dacht dat hij thuis ook maar iets te doen had.
nobody thought that he at.home anything to do had
‘Nobody thought that he had anything to do at home.’

G. Conclusion
The previous subsections have shown that quotes in indirect reported speech
constructions are direct object clauses. They exhibit the behavior of regular object
clauses, which was discussed more extensively in Sections 5.1.2.1 to 5.1.2.3.
Additionally, the discussion of binding and NPI-licensing has established that
subjects of say-clauses c-command the constituents in indirect quotes, which lends
credence to the claim that such quotes are regular direct object clauses.

II. Direct and semi-direct reported speech

This subsection discusses the question as to whether (semi-)direct quotes should be
considered direct object clauses. Subsections A and B show that the evidence is
rather varied, from which we will conclude that (semi-)direct reported speech
constructions are often structurally ambiguous. Subsection C provides some
additional support for this conclusion, and Subsection D concludes with a brief note
on the internal structure of the relevant constructions. Since (semi-)direct reported
speech constructions have not yet been studied extensively from a syntactic point of
view, much of what follows is tentative in nature and should therefore be taken with
care.

A. Direct reported speech

Direct reported speech constructions are often ambiguous. We will argue that such
constructions allow not only an analysis as regular transitive constructions in which
the quote functions as a direct object, but also an analysis in which the quote can
function as a main clause with an embedded parenthetical say-clause; cf. De Vries
(2006).

1. Are direct quotes direct objects?

Example (138a) strongly suggests that the direct quote in (138b) functions as the
direct object of the verb zeggen ‘to say’. The fact that the pronoun in (138a) cannot
be omitted shows that zeggen is a transitive verb that cannot be used pseudo-
 Argument and complementive clauses 691

intransitively. The fact that the direct quote is the only candidate that could function
as direct object in (138b) therefore seems to leave us no other option than to
conclude that it must have this syntactic function.
(138) a. Jan zei *(het).
Jan said it
b. Jan zei: “Ik ben ziek”.
Jan said I am ill
Although this line of argumentation seems quite convincing, there are various
reasons to reject the conclusion that direct quotes always function as object clauses.
First, it seems that introducing the direct quote with an anticipatory/deictic pronoun
het ‘it’ is not normally possible. Although example (139a) is fully acceptable, the
pronoun het does not seem to refer to the direct quote but to some other proposition.
This is evident from the fact illustrated in (139b) that the pronoun can be replaced
by an indirect quote such as the one in square brackets. Besides, example (139c)
shows that we would rather use phrases like als volgt ‘as follows’ or the manner
adverb zo ‘thus’ if we want to anticipate the direct quote.
(139) a. Jan vroeg het haar eindelijk: “Als ik je zie begint mijn hart te bonken: boem,
boem, boem ... Ik kan niet langer zonder jou!”
‘Jan finally asked her it: “Whenever I see you my heart starts pounding
boom, boom, boom ... I can no longer live without you!”’
b. Jan vroeg haar eindelijk [of ze met hem wilde trouwen]: “Als ik je zie begint
mijn hart te bonken: boem, boem, boem ... Ik kan niet langer zonder jou!”
‘Jan finally asked her whether she would marry him: “Whenever I see you
my heart start pounding boom, boom, boom ... I can no longer live without
you!”’
c. Jan vroeg het haar als volgt/zo: “Als ik je zie begint mijn hart te bonken:
boem, boem, boem .... Ik kan niet langer zonder jou!”
‘Jan finally asked her it as follows/thus: “Whenever I see you my heart starts
pounding boom, boom, boom ... I can no longer live without you!”’
From the discussion of the examples in (139) we are forced to conclude that the
direct quote does not function as a direct object in the examples in (139). Barbiers
(2000:190) even suggests that postverbal direct quotes are not even part of the
preceding say-clauses given that their intonational contour is entirely independent;
they are always preceded by a distinct intonation break. He suggests that this makes
it more likely that postverbal direct quotes function as some kind of afterthought
since afterthoughts exhibit the same prosodic effect. Barbiers does not claim that
direct quotes are never direct objects, but he asserts that they can only have this
function if they occur in the middle field of the clause, as in (140), in which case
they have the same distribution as nominal objects. Note in passing that examples
such as (140) quickly degrade when the quote gets longer.
692 Syntax of Dutch: Verbs and verb phrases

(140) a. Jan heeft “hallo” tegen de leraar gezegd.

Jan has hello to the teacher said
‘Jan has said “hallo” to the teacher.’
b. Jan heeft “ik ben ziek” tegen de leraar gezegd.
Jan has I am ill to the teacher said
‘Jan has said “Ik ben ziek” to the teacher.’
In (140) it is not entirely clear whether we are really dealing in (140) with reported
speech in the sense intended here. It may also be the case that we simply have to do
with an autonomous use of the word/phrase in question. That this may be the case is
strongly suggested by the fact that an utterance such as (140a) can quite naturally be
followed by something such as (141a). De Vries (2006) provides a similar example
and adds that the quote can also be in a language other than Dutch. This again
suggests that quotes may involve the autonomous use of the word/phrase in
question, and that this is the reason why they behave syntactically as nominal
arguments of the verb. In the discussion below we will ignore the autonomous use
of quotes in the middle field of the clause.
(141) a. Dat is onbeleefd: hij had “goedemorgen” moeten zeggen.
that is rude he should good.morning have said
‘That is rude: he should have said “goedemorgen”.’
b. John heeft “I am ill” tegen de leraar gezegd.
John has I am ill to the teacher said
‘John has said “Iʼm ill” to the teacher.’
Barbiers does not discuss direct quotes in the left periphery of the utterance, as
in (142), but it seems that such constructions show that direct quotes have an
ambiguous syntactic status. Although the construction in (142a) is the one
commonly used, the examples in (142b&c) show that it is also possible to add the
demonstrative pronoun dat or the manner adverb zo as a resumptive element.
(142) a. “Ik ben ziek”, zei Jan.
I am ill said Jan
‘“Ik ben ziek”, Jan said.’
b. “Ik ben ziek”, dat zei Jan.
I am ill that said Jan
c. “Ik ben ziek”, zo zei Jan.
I am ill thus said Jan
Subsection B will show that example (142b) can be analyzed as a °left-dislocation
construction. This example would then receive a similar analysis as example (143a)
in which the resumptive pronoun dat has a neuter singular antecedent functioning as
the logical direct object of the sentence. Example (143b) is added to show that other
resumptive pro-forms are used when the left-dislocated element has some other
logical function: the resumptive pro-form dan, for example, is used when the left-
dislocated element is the temporal adverb morgen ‘tomorrow’.
 Argument and complementive clauses 693

(143) a. Dat boek, dat heb ik al gelezen.

that book that have I already read
‘That book, Iʼve already read it.’
b. Morgen, dan ga ik naar Groningen.
tomorrow then go I to Groningen
‘Tomorrow, Iʼll be going to Groningen then.’
Although Subsection B will argue that (142c) is not a left-dislocation construction,
the fact that the manner adverb zo is used in a similar resumptive function
immediately suggests that the direct quote does not function as the logical direct
object of the say-clause. This conclusion receives further support from (144).
Example (144a) shows that the left-dislocation construction with the resumptive
pronoun dat does not allow the addition of the object pronoun het, which is to be
expected given that the resumptive pronoun already performs this function.
Example (144b), on the other hand, shows that, in the right context, the addition of
the object pronoun het is admissible in the construction with zo, which proves that
the direct quote does not function as the logical direct object of the say-clause in
this case.
(144) a. “Ik ben ziek”, dat zei Jan *(het).
I am ill that said Jan it
b. “Ik ben ziek”, zo zei Jan (het).
I am ill thus said Jan it
The fact that direct quotes need not function as (logical) direct objects of the say-
clause, established by the examples in (139) and (142) to (144), shows that our
earlier conclusion on the basis of example (138a) that the verb zeggen ‘to say’ may
not occur without a direct object is wrong; if a direct quote is present with some
other function than (logical) direct object of the say-clause, the direct object of the
verb zeggen can apparently remain unexpressed.
To sum up, this subsection has provided evidence that direct quotes preceded
by a say-clause do not function as the (logical) direct object of this say-clause. The
situation is different when the say-clause follows the quote; the quote may then
have the function of (logical) direct object, in which case the resumptive pronoun
dat can be inserted between the quote and the finite verb, or it may have an
adverbial function, in which case the resumptive pro-form surfaces as the manner
adverb zo. Observe that this conclusion raises the question as to how the selection
restrictions imposed by the matrix verbs on the direct quote can be accounted for if
the latter functions as an adjunct. Given that this cannot be accounted for by
normally assumed syntactic means (that is, subcategorization), a pragmatic account
seems to be called for. We leave this for future research.
(145) a. Jan zei/*vroeg: “Els wil vast wel een ijsje”.
Jan said/asked Els wants PRT PRT an ice.cream
‘Jan said: “Iʼm sure Els would like to have an ice cream”.’
b. Jan vroeg/*zei: “Wie wil er een ijsje?”.
Jan asked/said who wants there an ice.cream
‘Jan asked: “Who would like to have an ice cream?”.’
694 Syntax of Dutch: Verbs and verb phrases

2. Direct quotes and parenthetical clauses

The previous subsection has shown that direct quotes can but need not function as
direct object clauses of verbs of saying/thinking when they precede the say-clause.
The following question now arises: what is the structure of those constructions in
which the quote does not function as direct object? This subsection argues that
direct quotes are regular main clauses in such cases, which contain a parenthetical
say-clause. A first step in the argument involves the possible word orders in the
three constructions in (146).
(146) a. “Peter zal het boek morgen brengen”, zei Marie.
Peter will the book tomorrow bring said Marie
‘“Peter zal het boek morgen brengen”, Marie said.’
b. “Peter zal het boek morgen brengen”, dat zei Marie.
Peter will the book tomorrow bring that said Marie
c. “Peter zal het boek morgen brengen”, zo zei Marie.
Peter will the book tomorrow bring thus said Marie
We begin our discussion with example (146b), which we analyze as a left-
dislocation construction. Example (147a) shows that the direct quote need not
precede the say-clause but can also be right-dislocated, in which case the
resumptive pronoun dat will be replaced by the proximate demonstrative pronoun
dit ‘this’. The example which is crucial for our discussion is (147b), which shows
that the direct quote cannot be split by the say-clause.
(147) a. Marie zei dit: “Peter zal het boek morgen brengen”.
Marie said this: Peter will the book tomorrow bring
‘Marie said the following: “Peter zal het boek morgen brengen”’.
b. *“Peter”, dat/dit zei Marie, “zal het boek morgen brengen”.
Peter that/this said Marie will the book tomorrow bring
We should keep in mind, however, that reliable judgments on examples such as
(147b) are sometimes hampered by the fact that the same string is acceptable with a
non-quote interpretation: the speaker then simply provides a statement of his own
and uses a parenthetical clause to point at Marie as his source of information. This
is brought out in example (148a), in which the adverb tenminste ‘at least’ forces the
intended non-quote reading. Example (148b) shows that the parenthetical clause
cannot appear in a position preceding the constituent in sentence-initial position
(here: Peter).
(148) a. “Peter”, dat zei Marie tenminste, “zal het boek morgen brengen”.
Peter that said Marie at.least will the book tomorrow bring
‘According to Marie at any rate, Peter will bring the book tomorrow.’
b. *Marie zei dat/dit tenminste, “Peter zal het boek morgen brengen”.
Marie said that/this at.least Peter will the book tomorrow bring
Putting aside the non-quote reading, we are forced to conclude that the construction
in (146b) with resumptive dat differs sharply from the construction in (146a)
without a resumptive pronoun. The examples in (149) bear out that in the latter case
 Argument and complementive clauses 695

the direct quote can be split in various places by the say-clause. The examples in
(150) show that the same thing holds for construction (146c) with zo.
(149) a. “Peter zal het boek morgen brengen”, zei Marie.
Peter will the book tomorrow bring said Marie
‘Peter will bring the book tomorrow, Marie said.’
b. “Peter”, zei Marie, “zal het boek morgen brengen”.
c. “Peter zal”, zei Marie, “het boek morgen brengen”.
d. “Peter zal het boek ”, zei Marie, “morgen brengen”.
?
e. “Peter zal het boek morgen”, zei Marie, “brengen”.
(150) a. “Peter zal het boek morgen brengen”, zo zei Marie.
Peter will the book tomorrow bring thus said Marie
b. “Peter”, zo zei Marie, “zal het boek morgen brengen”.
c. “Peter zal”, zo zei Marie, “het boek morgen brengen”.
d. “Peter zal het boek ”, zo zei Marie, “morgen brengen”.
?
e. “Peter zal het boek morgen”, zo zei Marie, “brengen”.
The fact that the direct quotes can be split in (149) and (150) suggests that we are
dealing with parenthetical constructions. A potential problem is that example (151a)
shows that the presumed parenthetical say-clause in (149) may also precede the
quote; this is unexpected as example (148b) has shown that parenthetical clauses
cannot do so. However, there seems to be more to this than meets the eye given that
the say-clause in (150) behaves as expected and is indeed unable to precede the
quote: example (151b) is only acceptable if the sentence contains an object pronoun
like het.
(151) a. Marie zei: “Peter zal het boek morgen komen brengen”.
Marie said Peter will the book tomorrow come bring
b. Marie zei *(het) zo: “Peter zal het boek morgen komen brengen”.
Marie said it thus Peter will the book tomorrow come bring
The fact that the addition of het to the examples in (150) is unusual, to say the least,
suggests that (149) and (150) involve constructions entirely different from (151);
whereas the former involve parenthetical say-clauses, the say-clauses in the latter
may be regular transitive main clauses.
If we are indeed concerned with parenthetical clauses in (149) and (150), we
expect to find a wider range of examples that do not involve verbs of
saying/thinking. This expectation is borne out; in fact, writers have created an
infinite number of variations on this theme. A number of rather conventional
examples are given in (152). Note that the quotes cannot be analyzed as arguments
of the verbs beginnen ‘to start’, vervolgen ‘to continue’, and besluiten ‘to conclude’
in these examples: these verbs already have a direct object, zijn verhaal ‘his story’;
see De Vries (2006) for a number of less conventional examples.
(152) a. “De wind”, (zo) begon hij zijn verhaal, “was stormachtig”.
the wind thus started he his story was tempestuous
‘”The wind”, (thus) he started his story, “was tempestuous”.’
696 Syntax of Dutch: Verbs and verb phrases

b. “De boot” (zo) vervolgde hij zijn verhaal, “was in gevaar”.

the boat thus continued he his story was in danger
‘“The boat”, (thus) he continued his story, “was in danger”.’
c. “De schipper”, (zo) besloot hij zijn verhaal, “spoelde dood aan”.
the skipper thus concluded he his story washed dead ashore
‘“The skipper”, (thus) he concluded his story, “washed ashore dead”.’
In order to give an impression of the semantic verb types that can be used in
parenthetical say-clauses, we provide a small sample in (153), adapted from De
Vries (2006). Note that this list includes a number of intransitive verbs like
schreeuwen ‘to shout’, which provides further support for the claim that a direct
quote does not function as an argument of the main verb in parenthetical say-
clauses.
(153) a. Saying, thinking and writing: antwoorden ‘to answer’, denken ‘to think’,
prediken ‘to preach’, schrijven ‘to write’, vertellen ‘to tell’, vragen ‘to ask’,
zeggen ‘to say’
b. Manner of speech and sound emission, schreeuwen ‘to shout’, vloeken ‘to
curse’, zuchten ‘to sigh’, giechelen ‘to giggle’, schateren ‘to roar’,
trompetteren ‘to trumpet’, sissen ‘to hiss’, zingen ‘to sing’
c. Thinking, observation and explanation: concluderen ‘to conclude’, denken
‘to think’, fantaseren ‘to fantasize’, opmerken ‘to observe’, peinzen ‘to
contemplate’, verduidelijken ‘to clarify’

B. Semi-direct reported speech

Semi-direct reported speech constructions exhibit more or less the same syntactic
behavior as their direct counterparts. The direct reported speech constructions in
(139), for instance, can easily be transformed into the semi-direct reported speech
constructions in (154). It shows that, like direct quotes, semi-direct quotes need not
function as direct objects of the verb zeggen ‘to say’. It may therefore be the case
that in examples such as (154) the direct quote is actually not part of the first
sentence, but consists of a series of independent sentences.
(154) a. Jan vroeg het haar eindelijk. Als hij haar zag, begon zijn hart te bonken:
boem, boem, boem ... Hij kon niet langer zonder haar!
‘Jan finally asked her it. Whenever he saw her his heart started pounding
boom, boom, boom ... He could no longer live without her!’
b. Jan vroeg haar eindelijk of ze met hem wilde trouwen. Als hij haar zag, begon
zijn hart te bonken: boem, boem, boem ... Hij kon niet langer zonder haar!
‘Jan finally asked her whether she would marry him. Whenever he saw her
his heart started pounding boom, boom, boom ... He could no longer live
without her!’
c. Jan vroeg het haar als volgt/zo. Als hij haar zag, begon zijn hart te bonken:
boem, boem, boem ... Hij kon niet langer zonder haar!
‘Jan finally asked her it as follows/thus. Whenever he saw her his heart
started pounding boom, boom, boom ... He could no longer live without her!’
 Argument and complementive clauses 697

The hypothesis that semi-direct quotes are independent sentences may also account
for the fact that semi-direct quotes cannot be embedded whereas direct reported
speech constructions can. The acceptability contrast between the two primed
examples in (155) illustrates this. However, it may not be a syntactic issue after all:
embedding semi-direct speech constructions may simply be inconsistent with the
fact that a semi-direct quote is a stylistic means used for expressing the internal
thoughts of the protagonist(s) of a story; see the discussion in the introduction to
Section 5.1.2.4. We leave this issue for future research.
(155) a. Jan dacht: “Ik ben ziek”.
Jan thought I am ill
a. Ik weet zeker dat Jan dacht: “Ik ben ziek”.
I know for.sure that Jan thought I am ill
b. Jan dacht: hij was ziek.
Jan thought he was ill
b. $Ik weet zeker dat Jan dacht hij was ziek.
I know for.sure that Jan thought he was ill
Like direct quotes, semi-direct quotes seem to have an ambiguous syntactic status,
as is clear from the fact that the direct quotes in (142) can be replaced by semi-
direct quotes without any difficulty.
(156) a. Hij was ziek, zei Jan.
he was ill said Jan
‘He was ill, Jan said.’
b. Hij was ziek, dat zei Jan.
he was ill that said Jan
c. Hij was ziek, zo zei Jan.
he was ill thus said Jan
The acceptability of (156b&c) suggests that semi-direct quotes may function as
independent sentences with parenthetical say-clauses. This is confirmed by the
examples in (157), which show that the say-clauses may split the quotes.
(157) a. Hij, zei Jan, was ziek.
he said Jan was ill
b. Hij, zo zei Jan, was ziek.
he thus said Jan was ill
The acceptability of (156b) suggests that semi-direct quotes may also function as
direct objects, which seems to be confirmed by the fact illustrated in (158a) that the
say-clause with resumptive dat cannot split the quote. Note, however, that this
sentence is fully acceptable if it is interpreted as an assertion made by the speaker
himself, who simply points to Jan as his source of information by means of a
parenthetical clause, as in (158b).
(158) a. *Hij, dat zei Jan, was ziek. [semi-direct quote]
he, that said Jan, was ill
b. Hij, dat zei Jan tenminste, was ziek. [non-quote]
he, that said Jan at least, was ill
698 Syntax of Dutch: Verbs and verb phrases

The difference between the semi-direct reported speech constructions in (157) and
the non-quote construction in (158b) can be brought out even more clearly by
taking coreferentiality into account: whereas the pronouns in the quotes in (157) can
be interpreted as coreferential with the subject Jan of the say-clauses, the pronoun
in (158b) does not allow this so easily. We illustrate this in the primeless examples
of (159) by means of slightly more elaborate examples. Example (159b) shows that
the proper name in (159b) must be replaced by a pronoun in order to allow the
intended coreferentiality reading, and even then this reading is often emphasized in
speech by addition of the emphatic marker zelf ‘himself’. Note that some of our
informants do allow the intended coreference relation indicated in (159b); this is
indicated by the percentage sign.
(159) a. Morgen zou hij, (zo) zei Jan, vroeg vertrekken.
tomorrow would he, thus said Jan early leave
‘He, said Jan, would leave early tomorrow.’
b. %Morgen zou hij, dat zei Jan tenminste, vroeg vertrekken.
tomorrow would he, thus said Jan at.least early leave
b. Morgen zou hij, dat zei hij tenminste (zelf), vroeg vertrekken.
tomorrow would he, thus said he at.least himself early leave
Although the discussion above suggests that semi-direct quotes may function not
only as independent sentences but also as direct objects of the verb of saying, it
should be noted that they never occur in the middle field of the say-clause: the
direct reported speech construction in (140b), repeated here as (160a), does not
have a semi-direct counterpart: example (160b) is unacceptable under the intended
reading and can at best be interpreted as a direct quote, that is, with the
interpretation that Jan literally said “Hij was ziek”.
(160) a. Jan heeft “ik ben ziek” tegen de leraar gezegd.
Jan has I am ill to the teacher said
‘Jan has said “Iʼm ill” to the teacher.’
b. #Jan heeft hij was ziek tegen de leraar gezegd.
Jan has he was ill to the teacher said
In addition, the examples in (161) show that it is impossible to use semi-direct
quotes as the complement of a noun.
(161) a. Jan beweerde: “ik ben ziek”.
Jan claimed I am ill
a. Jans bewering “Ik ben ziek” kwam als een nare verrassing.
Janʼs assertion I am ill came as a nasty surprise
b. Jan beweerde hij was ziek.
Jan claimed he was ill
b. *Jans bewering hij was ziek kwam als een vervelende verrassing.
Janʼs assertion he was ill came as a nasty surprise
If say-clauses without resumptive dat are indeed parenthetical clauses, we
expect that, just as in the case of direct quotes, we should find a wider range of
examples that do not involve verbs of saying/thinking. This expectation is again
 Argument and complementive clauses 699

borne out. A number of rather conventional examples are given in (162). As in

(152), the quote cannot be analyzed as an argument of the verbs beginnen ‘to start’,
vervolgen ‘to continue’, and besluiten ‘to conclude’ since these already have a
direct object, zijn verhaal ‘his story’. Analyzing the say-clauses as parentheticals
seems the only option therefore.
(162) a. Hij moest, (zo) begon de schipper zijn verhaal, bij storm uitvaren.
he had.to thus started he skipper his story during storm out sail
‘He had to set sail, (thus) the skipper started his story, during a gale.’
b. Zijn boot (zo) vervolgde de schipper zijn verhaal, was in gevaar.
his boat thus continued the skipper his story was in danger
‘His boat, (thus) the skipper continued his story, was in danger.’
c. Dit, (zo) besloot de schipper zijn verhaal, redde hem van de dood.
this thus concluded the skipper his story saved him from the death
‘This, thus the skipper concluded his story, saved him from death.’

C. Additional evidence for structural ambiguity

This subsection discusses a number of additional arguments, mainly taken from
Corver (1994) and Corver & Thiersch (2003), in favor of the conclusion reached in
the previous subsections that (semi-)direct reported speech constructions can be
structurally ambiguous. Their point of departure is the observation in Reinhart
(1983) that (semi-)direct reported speech constructions can also be semantically
ambiguous; the say-clause may be either subject- or speaker-oriented. We will see
that this semantic ambiguity correlates with the structural ambiguity discussed in
the previous subsections.

1. Subject-oriented reading
The subject-oriented reading is triggered by questions such as (163a). The
interrogative clause is transitive and directed towards the subject matter of the
addressee’s thoughts; the answer in (163b) therefore plausibly involves a transitive
structure as well. The fact that the direct quote in (163b) may be replaced by the
indirect quote in (163b) provides additional support for this conclusion, given that
Subsection I has established that an indirect quote also has the function of direct
object.
(163) a. Wat denk je?
what think you
‘What do you think?’
b. Ik vertrek om zeven uur, denk ik.
I leave at seven oʼclock think I
b. [Dat ik om zeven uur vertrek] denk ik.
that I at seven oʼclock leave think I

2. Speaker-oriented reading
The speaker-oriented reading is triggered by questions such as (164a). The person
asking the question is not interested in the addressee’s thoughts but in information
about a specific state of affairs. The person answering the question simply adds a
700 Syntax of Dutch: Verbs and verb phrases

parenthetical say-clause as a warning; he is not completely sure that his answer

is/will come true. The suggestion that we are dealing with a parenthetical say-clause
implies that the direct quote in (164b) is a main clause, and this correctly predicts
that it cannot be replaced by an indirect quote, as the latter functions as an
embedded clause; example (164b) is not a felicitous answer to question (164a).
(164) a. Hoe laat vertrek je?
how late leave you
‘When will you leave?’
b. Ik vertrek om zeven uur, denk ik.
I leave at seven oclock, think I
b. #[dat ik om zeven uur vertrek], denk ik.
that I at seven oclock leave think I
Note in passing, however, that the transitive construction Ik denk dat ik om zeven
uur vertrek ‘I think that I will leave at seven o’clock’ would be a felicitous answer
to (164). It is not entirely clear why this is possible and why a similar transitive
reading of the say-clause is blocked for example (164b).

3. Differences between the subject- and the speaker-oriented reading

Our suggestion that the subject- and speaker-oriented readings are associated with,
respectively, transitive and parenthetical structures is supported by a number of
additional facts, although it should be noted that judgments of the relevant
examples are sometimes subtle. First, since we have seen that (semi-)direct quotes
can only be split by parenthetical clauses, we predict that an example such as (165)
cannot be used as an answer to the question Wat denk je? ‘What do you think?’ in
(163a); it does seem that it can only be used as an answer to the question Hoe laat
vertrek je? ‘At what time will you leave?’ in (164a).
(165) Ik vertrek, denk ik, om zeven uur.
I leave think I at seven oʼclock
Secondly, we predict that the direct object clause will be semantically reconstructed
in its original object position in the subject-oriented construction. This means that a
referential expression like Jan that is embedded in the quote cannot be bound by,
e.g., the subject of the say-clause because this would violate °binding condition C
on referential expressions. This also seems to be the case: whereas the pronoun hij
can readily be interpreted as coreferential with Jan in (166b), this seems excluded
in example (166b); see Reinhart (1983) for similar judgments on English. Of
course, the intended interpretation is that Jan is the brother of the person answering
the question.
(166) a. Wat zei je broer?
what said your brother
‘What did your brother say?’
b. Hij was ziek, zei Jan.
he was ill said Jan
b. *Jan was ziek, zei hij.
Jan was ill said he
 Argument and complementive clauses 701

Reconstruction is not relevant in the case of the speaker-oriented reading, and we

therefore correctly predict that the question in (167a) can be answered by the
primed (b)-example. Corver & Thiersch claim that the primeless (b)-example is
marked as an answer to (167a). If so, this may follow from the fact that, apart from
cases of reconstruction, referential expressions tend to precede pronouns that they
are coreferential with.
(167) a. Waarom bibbert je broer zo?
why shivers your brother like.that
‘Why is your brother shivering like that?’
b. %Hij was ziek, zei Jan.
he was ill said Jan
b. Jan was ziek, zei hij.
Jan was ill said he
Corver & Thiersch provide similar evidence based on °bound variable readings of
pronouns. Our prediction is that such readings are only possible if the say-clause is
transitive, that is, if it has a subject-oriented reading. Corver & Thiersch claim that
this is indeed borne out, but some speakers have difficulty with getting a bound
variable reading in both cases. For this reason we have added a percentage sign to
example (168a).
(168) a. Wat zei iedereen?
what said everyone
‘What did everyone say?’
a. %Hij zou staken, zei iedereen.
he would go.on.strike said everyone
b. Waarom loopt iedereen weg?
why walks everyone away
‘Why is everyone walking away?’
b. *Hij zou staken, zei iedereen.
he would go.on.strike said everyone
A final piece of evidence supporting the claim that subject- and speaker-oriented
readings are associated with, respectively, the transitive and the parenthetical
structure is provided in (169). The tag question ... of toch niet? is used as an
afterthought expressing doubt on the part of the speaker about the preceding
assertion: De auto is kapot, ... of toch niet ‘The car is broken, ... or maybe not?’ The
(a)-examples in (169) show that the tag question can have scope over the entire
preceding clause when we are dealing with the transitive, subject-oriented
construction: the speaker expresses doubt about whether Jan did indeed say that the
car was broken. In the (b)-examples, on the other hand, the tag question has scope
over the quote only; the speaker expresses his doubt about whether the car was
broken at all, not about whether Jan was his source of information. This contrast
will follow from our proposal if parenthetical clauses are not part of the core
information. The scope of the tag questions in (169) is indicated by italics.
702 Syntax of Dutch: Verbs and verb phrases

(169) a. Wat zei Jan over die auto?

what said Jan about that car
‘What did Jan say about that car?’
a. Hij was kapot, zei Jan, ... of toch niet?
he was broken said Jan, or PRT not
‘It was broken, said Jan, ... or did he not?’
b. Wat is er met die auto?
what is there with that car
‘What is the matter with that car?’
b. Hij was kapot, zei Jan, ... of toch niet?
he was broken said Jan, or PRT not
‘It was broken, said Jan, ... or was it not?’

D. The structure of parenthetical (semi-)direct reported speech constructions

The discussion in the previous subsections has shown that analyzing the say-clauses
such as (170) as parentheticals seems a feasible option. It is, however, far from clear
what the internal structure of these parenthetical clauses is. Due to the fairly recent
discovery that (semi-)direct reported speech constructions may involve
parenthetical clauses, the issue has not received much attention in the literature so
far.
(170) “Peter zal het boek morgen brengen”, (zo) zei Marie.
Peter will the book tomorrow bring thus said Marie
‘“Peter zal het boek morgen brengen”, Marie said.’
An analysis was proposed for English in Branigan & Collins (1993), Collins &
Branigan (1997) and Collins (1997); it involves movement of a phonetically empty
°operator that functions as the object of the verb of saying into the clause-initial
position of the parenthetical say-clause. The desired interpretation is derived by
assuming that the empty operator is coindexed with the quote. Applied to the Dutch
cases, this would correctly account for the fact that the verb of saying is the first
visible element in the parenthetical clause in (146a) as the first position of the
clause is filled by a phonetically empty element.
(171) a. [Ik ben ziek]i , [OPi zei Jan ti].
b. [Hij was ziek]i , [OPi zei Jan ti].
However, the proposal in (171) does not account for the fact established in the
previous subsections that the overt counterpart of the postulated empty operator is
zo ‘so’, not dat ‘that’ (which in fact also holds for English). A proposal that would
solve this problem can be found in Corver (1994) and Corver & Thiersch (2003),
who assume that zo is phrasal and, in fact, contains a phonetically empty
pronominal element pro functioning as a direct object; see De Vries (2006) for a
similar intuition.
(172) a. [Ik ben ziek]i , [[pro-zo]i zei Jan ti].
b. [Hij was ziek]i , [[pro-zo]i zei Jan ti].
 Argument and complementive clauses 703

Although this proposal may raise all kinds of technical issues (like the fact that
Dutch normally does not allow pro objects and that pro is not directly related by
movement to the object position of the parenthetical clause), it would account for
the fact that het cannot be present in parenthetical say-clauses. In structures such as
(173) the direct object is expressed twice; once by pro and once by the pronoun het.
(173) a. *[Ik ben ziek]i , [[pro-zo]i zei Jan het ti].
b. *[Hij was ziek]i , [[pro-zo]i zei Jan het ti].
A potential problem for this proposal is that zo can sometimes be combined with a
direct object; this was shown in (152) and (162) above where the (optional) noun
phrase zijn verhaal ‘his story’ clearly functions as a direct object. In examples such
as (151b), repeated below as (174b), a direct object is even obligatory: whereas the
pronoun het can be left out in the (transitive) direct reported speech construction in
(174a), it must be present in the construction with zo in (174b). If zo indeed
contained an empty pronominal element, this would be surprising.
(174) a. Marie zei: “Peter zal het boek morgen komen brengen”.
Marie said Peter will the book tomorrow come bring
b. Marie zei *(het) zo: “Peter zal het boek morgen komen brengen”.
Marie said it thus Peter will the book tomorrow come bring
For completeness’ sake, it should be noted that there are also (transitive) direct
reported speech constructions with zo in which the pronoun can be left out. This is
the case in constructions such as (175) containing discourse particles like nog and
maar.
(175) a. Ik zei (het) nog zo; “je moet opletten voor die auto”.
I said it PRT. so you must take.heed of that car
b. Ik zeg (het) maar zo: “morgen is er weer een dag”.
I say it PRT. so tomorrow is there again a day

The discussion above seems to lead to the rather ad hoc assumption that in certain
constructions zo obligatorily contains the empty pronoun pro, whereas in other
constructions it cannot or only optionally do so. However, if we reject Corver &
Thiersch’ proposal for this reason, we have to conclude that we may leave the direct
object of the verb zeggen unexpressed in examples like (149) and (150), despite the
fact that example (138a) has shown that the verb zeggen normally cannot occur
without a direct object. This position would be equally ad hoc. It shows that we do
not yet have a fully satisfactory analysis for parenthetical say-clauses at our
disposal. Since we have nothing more enlightening to say about this issue at the
moment, we leave it for future research and simply conclude that direct and semi-
direct reported speech constructions can be ambiguous.

III. Quotative and polar van-constructions

We conclude the discussion of reported speech with a look at quotative van-
constructions such as (176), which are typically (but not exclusively) found in
colloquial speech and informal writing; see Verkuyl (1977), Romein (1999), and,
especially, Foolen et al. (2006). The literature often claims that the quotative
704 Syntax of Dutch: Verbs and verb phrases

preposition van is of a similar kind as the preposition van that we find in polar van-
constructions such as (176b). For this reason, we will also discuss the latter
construction in this subsection.
(176) a. Jan zei van ... kom morgen maar even langs.
Jan said VAN come tomorrow PRT along come
‘Jan said something like: “Drop in tomorrow, if you like”.’
b. Jan zei van niet/wel.
Jan said VAN not/AFF
‘Jan denied/affirmed it.’
Observe that we do not use quotation marks in the quotative van-construction since
we will see that it differs from the direct reported speech construction discussed in
Subsection II in that it is not used to reproduce utterances literally.

A. Quotative van-constructions
This subsection investigates the quotative van-construction and is organized as
follows. Subsection 1 discusses the internal make-up of the quotative van-phrase,
Subsection 2 the meaning of the quotative van-construction as a whole, and
Subsection 3 some of its syntactic properties. As we go along, we will point out a
number of differences between quotative van-constructions and reported speech
constructions without van.

1. The quotative van-phrase

Quotative van-constructions involve the QUOTATIVE PREPOSITION van, which is
followed by an optional hesitation marker like eh ‘er’, an intonation break, and a
quote. The examples in (177) show that the quote can be declarative or interrogative
in nature; the reader can find many more interrogative examples on the internet by
doing a Google search on, e.g., the strings [vroeg van kan] (... asked VAN be able ...)
or [vroeg van hoe] (... asked VAN how ...).
(177) a. Marie dacht van (eh) ... hij komt straks wel weer terug.
Marie thought VAN er he comes later PRT again back
‘Marie thought something like: “Heʼll probably come back later”.’
b. Marie vroeg van (eh) ... kan je me even helpen?
Marie asked VAN er can you me for.a.moment help
‘Marie asked something like: “Can you help me a moment?”.’
b. Marie vroeg van (eh) ... wie leest zoʼn boek nou?
Marie asked VAN er who reads such.a book now
‘Marie asked something like: “Who on earth reads a book like that?”.’
The examples in (177) involve direct quotes but it is also possible to have indirect
quotes. The examples in (178) show that the quotes can again be declarative or
interrogative in nature; the reader can find many more interrogative examples on
the internet by a Google search on, e.g., the strings [vroeg van of] (... asked VAN
whether ...) or [vroeg van hoe] (... asked VAN how ...).
 Argument and complementive clauses 705

(178) a. Marie dacht van (eh) ... dat hij straks wel weer terug komt.
Marie thought VAN er that he later PRT again back comes
b. Marie vroeg van (eh) ... of ik eventjes kan helpen.
Marie asked VAN er whether I for.a.moment can help
b. Marie vroeg van (eh) ... wie zoʼn boek nou leest?
Marie asked VAN er who such.a book now reads
Quotative van-constructions frequently occur with verbs normally taking a direct
quote in writing. Romein (1999) even suggests that the preposition van has a similar
function as the colon in written language. It should be noted, however, that the
quotative van-phrase may also be used as modifier/complement of non-verbal
phrases that cannot be used in direct reported speech constructions without van; cf.
Foolen et al. (2006). This will be clear from the difference in acceptability between
the primeless and primed examples in (179).
(179) a. Hij zit daar met een gezicht van ... ik heb niets verkeerds gedaan.
he sits there with a face VAN I have nothing wrong done
‘He sits there with a face expressing: “I havenʼt done anything wrong”.’
a. *Hij zit daar met een gezicht: “ik heb niets verkeerds gedaan”.
he sits there with a face I have nothing wrong done
b. Hij had het idee van ... nu eventjes doorbijten!
he had the idea VAN now for.a.while keep.trying
‘He had the idea: “Just grin and bear it for a while!”.’
b. *Hij had het idee: “Nu eventjes doorbijten!”.
he had the idea now for.a.while keep.trying

2. Meaning aspects of the quotative van-construction

Direct reported speech constructions without van differ in another respect: the quote
following quotative van need not be identical to the reported utterance or thought,
but is presented as an approximation at best; cf. Van Craenenbroeck (2002) and
Foolen et al. (2006: 142-3). This is clear from the fact illustrated in (180) that
quotative van-phrases often co-occur with the indefinite pronoun iets/zoiets
‘something’. Note that in (180a) the preposition van can be replaced by als ‘like’ if
the construction contains a verb of speaking or thinking; this seems less felicitous in
cases such as (180b), in which the quotative van-phrase functions as a
modifier/complement of a nominal phrase.
(180) a. Hij dacht (zo)iets van/als ... dat vertik ik!
he thought something VAN/like that refuse.flatly I
‘He thought something like “Iʼll be damnʼd if I do that!”.’
b. Hij had een houding van/ ?als ... dat vertik ik!
he had an attitude VAN/like that refuse.flatly I
‘His attitude was something like “Iʼll be damnʼd if I do that!”.’
That we are dealing with sloppy quotes may find additional support in the fact,
illustrated in (181), that quotative van can easily be replaced by phrasal prepositions
like in de trant/geest van and op een manier van, which are all semantically close to
English like.
706 Syntax of Dutch: Verbs and verb phrases

(181) a. Hij zei iets in de geest/trant van ... wat maakt het uit?
he said something in the spirit/manner of what makes it prt.
‘He said something like “What difference does it make?”.’
b. Hij keek op een manier van ... wat willen die mensen van me?
he looked in a manner of what want those people from me
‘He looked like he was thinking “What do these people want from me?”.’
In fact, the quote may even be invented by the speaker himself in order to give a
subjective typification of some (aspect of) a person. This is what is the case in
examples (179a) and (180b) above: the quote is used to provide a characterization
of the presumed attitude of the person under discussion, and need not have anything
to do with what that person actually said or thought. That we are nevertheless dealing
with some sort of reported speech construction is clear from the fact that, just like in
direct reported speech constructions without van, the part following van need not be
well-formed Dutch, but can be virtually any sound; cf. Hoeksema (2006).
(182) a. De ober zei (iets) van ... Non monsieur! Pas possible!
the waiter said something VAN non monsieur pas possible
‘The waiter said something like “Non monsieur! Pas possible!”.’
b. De trein ging van ... tjoeke, tjoeke, tjoek.
the trein went VAN tjoeke, tjoeke, tjoek
‘The trein made a sound like “choo-choo-choo”.’
Note in passing that in Foolen et al. (2006) the approximate/typificational reading
of the quotative van-construction is related to the fact that the preposition van may
also have an approximate/typificational function in non-quotative constructions.
This is illustrated in (183) by means of the phrasal predicate iets (weg) hebben van
‘to look like/resemble’.
(183) a. Hij heeft iets van Mick Jagger.
he has something VAN Mick Jagger
‘He reminds me of Mick Jagger in a way.’
b. Hij heeft iets weg van een filmster.
he has something away VAN a movie star
‘He looks a bit like a movie star.’
Related to the typificational reading of the quotative van-construction is that the
quote is often some conventionalized expression providing a more or less generally
recognizable characterization of some state of affairs. Some cases were provided
earlier but in (184) we add two, slightly abbreviated, attested examples; the quote in
(184b) is a fixed expression in Dutch. We refer to Foolen et al. (2006) for a more
extensive discussion of the pragmatic and sociolinguistic aspects related to the
actual use of the quotative van-construction.
(184) a. een wereldbeeld van je bent voor of tegen ons
a world.view VAN you are for or against us
‘a world view of the type: “Youʼre either for or against us”’
b. een sfeer van doe maar gewoon, dan doe je al gek genoeg
an atmosphere VAN do PRT. normally then do you already crazy enough
‘an atmosphere of the type: “be normal, then youʼre being crazy enough as it is”’
 Argument and complementive clauses 707

That we are dealing with a subjective typification is clear from the fact that the
quotative van-constructions are not compatible with factive predicates like
betreuren in (185); cf. Van Craenenbroeck (2002). Since betreuren normally does
not take a direct quote, we also give an example with an indirect quote.
(185) a. Jan zei/*betreurde (iets) van ... Ik ben ziek.
Jan said/regretted something VAN I am ill
b. Jan zei/*betreurde (iets) van ... dat hij ziek was.
Jan said/regretted something VAN that he ill was

3. Syntactic behavior of the quotative van-phrase

The quotative preposition van and the quote can be separated by an (optional)
hesitation marker and an intonation break. Otherwise, however, they are always
adjacent: it is not possible to place, e.g., adverbial material between the preposition
and the quote or to separate them by movement. We illustrate the latter in (186) by
means of topicalization; the trace ti indicates the normal position of the quote.
(186) a. *[Ik ben ziek]i zei Jan (iets) van ti.
I am ill said Jan something VAN
b. *[dat hij ziek was]i zei Jan (iets) van ti.
that he ill was said Jan something VAN
Probably related to the adjacency requirement is that quotative van-constructions
cannot be used as parenthetical clauses, as is shown by the contrast in (187); see
also Van Craenenbroeck (2002).
(187) a. Jan is, zei Marie, al vanaf gisteren ziek.
Jan is, said Marie already since yesterday ill
‘Jan has been ill since yesterday, said Marie.’
b. *Jan is, zei Marie iets van, al vanaf gisteren ziek.
Jan is, said Marie something VAN already since yesterday ill
Example (188a&b) show that quotative van-phrases are normally placed in the
position following the verb in clause-final position; note that the indefinite pronoun
iets/zoiets (if present) must precede the clause-final verb. The (c)-examples show
that topicalization is normally not possible either, regardless of whether or not the
preposition van is stranded.
(188) a. Hij zal wel (iets) denken van ... die is gek!
he will PRT something think VAN that.one is crazy
‘Heʼll probably think something like: “That one is crazy!”.’
b. ??Hij zal wel (iets) van ... die is gek! denken.
he will PRT something VAN that.one is crazy think
c. *Van ... die is gek! zal hij wel denken.
VAN that.one is crazy will he PRT think
c. *Die is gek! zal hij wel (iets) denken van.
that.one is crazy will he PRT something think VAN
708 Syntax of Dutch: Verbs and verb phrases

Another thing to note is that quotative van-constructions differ from (in)direct

reported speech constructions in that they never contain an °anticipatory pronoun.
This contrast is illustrated in (189).
(189) a. dat Jan het eindelijk vroeg: “Wil je met me trouwen!”
that Jan it finally asked want you with me marry
‘that Jan finally asked it: “Will you marry me!”.’
a. dat Jan (*het) eindelijk vroeg van ... wil je met me trouwen.
that Jan it finally asked VAN want you with me marry
b. dat Jan het eindelijk vroeg of ze met hem wilde trouwen.
that Jan it finally asked whether she with him wanted.to marry
‘that Jan finally asked it if she would marry him.’
b. dat Jan (*het) eindelijk vroeg van of ze met hem wilde trouwen.
that Jan it finally asked VAN whether she with him wanted.to marry

B. Polar van-constructions
This subsection discusses polar van-constructions such as (190a). The name of this
construction derives from the fact that the complement of van is typically one of the
polar adverbs wel and niet, which function, respectively, as affirmative marker and
negation. We will compare a polar van-construction such as (190a) to a polar van-
construction such as (190b) which involves the polar elements ja ‘yes’ and nee
‘no’; we will see that, although the two constructions look very similar at first sight,
they exhibit a quite different behavior.
(190) a. Ik dacht van wel/niet. [polar van wel/niet-construction]
I thought VAN AFF/not
b. Ik dacht van ... ja/nee. [polar van ja/nee-construction]
I thought VAN yes/no
In order to make the comparison between the two polar van-constructions in (190)
easier, Subsection 1 begins with a brief comparison of the syntactic behavior of
polar wel/niet and polar ja/nee ‘yes/no’. This will show that the former normally
functions as a constituent of a clause while the latter does not. Subsection 2 then
continues with an investigation of a number of differences in use of the van
wel/niet- and the van ja/nee-phrases. Subsection 3 goes on to discuss a number of
syntactic properties of polar van wel/niet-phrases, and Subsection 4 concludes with
a brief discussion of a suggestion in Hoeksema (2006) that the polar van wel/niet-
and van ja/nee-constructions in (190) are special cases of, respectively, indirect and
direct quotation.

1. The syntactic function of ja/nee ‘yes/no’ and wel/niet

This subsection discusses several differences between ja/nee ‘yes/no’ and wel/niet.
A first difference is that ja and nee are used in answering yes/no-questions, whereas
wel and niet are adverbs used as an affirmation and a negation marker, respectively.
The (b)-examples in (191) show that ja/nee and wel/niet crucially differ: the former
can be used as independent utterances in response to a question whereas the latter
cannot.
 Argument and complementive clauses 709

(191) a. Komt Jan morgen? [speaker A]

comes Jan tomorrow
‘Will Jan come tomorrow?’
b. Ja/Nee. b. *Wel/niet. [speaker B]
yes/no AFF/not

Another difference between ja/nee and wel/niet is that the former are never clausal
constituents, whereas the latter must be. This is demonstrated by the (b)-and (c)-
examples in (192) which show that ja/nee must precede the element in sentence-
initial position, and must therefore be sentence-external. The polarity adverbs, on
the other hand, always occupy a sentence-internal position, preferably in the middle
field of some clause. The (b)- and (c)-examples in (192) are intended as answers to
the question in (192a).
(192) a. Komt Jan morgen?
comes Jan tomorrow
b. Ja, ik denk dat hij komt. b. Nee, ik denk dat hij niet komt.
yes, I think that he comes No I think that he not comes
‘Yes, I think he will. ‘No, I think he wonʼt.’
c. Ik denk dat hij wel komt. c. Ik denk dat hij niet komt.
I think that he AFF comes I think that he not comes
‘I think he will.’ ‘I think he wonʼt.’
Note in passing that the examples in (193) show that the elements welles and nietes,
which are used to bring about a truth transition by contradicting some immediate
preceding assertion in discourse, behave in this respect like ja/nee and not like
wel/niet. Contrary to what is claimed by Hoeksema (2006:fn.2), the forms welles
and nietes also occur in polar van-constructions, as is clear from, e.g., the
following, completely natural example: Jambers zegt van nietes, De Pauw zegt van
welles waarop [...] ‘Jambers says it is the case, De Pauw says it is not, after which
[...]’ (Nieuwsblad.be, September 3, 2004). We will, however, not discuss nietes and
welles here, but simply assume that at least for some speakers they behave like ja
and nee in the relevant respects.
(193) a. Jan is niet hier. a. Welles, ik zag hem net.
Jan is not here he.is I saw him just.now
‘Jan isnʼt here. Yes, he is, I saw him just now.’
b. Jan is er al. b. Nietes, hij belde net dat hij ziek is.
Jan is here already he.is.not he phoned just that he ill is
‘Jan is already here. No, he isnʼt, he just phoned to tell that heʼs ill.’
The fact that the ja/nee cannot be used as clausal constituents leaves us with no
other option than to analyze examples such as (190b) as quotative van-constructions
with a direct quote. That the polar adverbs wel/niet do not occur as independent
utterances (apart from cases of ellipsis) makes such an analysis unlikely for
example (190a). This conclusion is supported by examples (194b&c); whereas ja
and nee are quite normal as direct quotes in reported speech constructions without
van, the polar adverbs wel and niet are not. We added the primed examples to show
that ja/nee and wel/niet are possible in the corresponding van-constructions.
710 Syntax of Dutch: Verbs and verb phrases

(194) a. Marie vroeg: “Komt Jan morgen?” [speaker A]

Marie asked comes Jan tomorrow
b. Ik antwoordde snel: “Ja/nee”. [speaker B]
I answered quickly yes/no
b. Ik antwoordde snel van ja/nee.
c. *Ik antwoordde snel: “wel/niet”. [speaker B]
I answered quickly AFF/not
c. Ik antwoordde snel van wel/niet.
The claim that the polar van-constructions with ja/nee are quotative van-
constructions with a direct quote receives further support from example (195a),
which shows that the complement of van can be supplemented with all kinds of
other material. Example (195b), on the other hand, does not allow such
supplements, which again suggests that van-constructions with wel/niet do not
involve a direct quote.
(195) a. Ik dacht van ... ja/nee.
I thought VAN yes/no
a. Ik dacht van ... ja/nee, (maar) dat wil ik ook!
I thought VAN yes/no but that want I also
b. Ik dacht van wel/niet.
I thought VAN AFF/not
b. *Ik dacht van wel/niet, (maar) dat wil ik ook!
I thought VAN AFF/not that want I also

2. Differences in use between van ja/nee and van wel/niet phrases

The two polar van-constructions in (190), which are repeated below as (196), are
subject to different conditions on their use as well. Whereas the van-construction
with ja/nee in (196b) can be used in any situation in which it is relevant to report
the speaker’s thoughts, the van-construction with wel/niet in (196a) is used in
specific circumstances only.
(196) a. Ik dacht van wel/niet.
I thought VAN AFF/not
b. Ik dacht van ... ja/nee.
I thought VAN yes/no
In fact, the two constructions often seem to be in complementary distribution. A
first illustration of this is provided by the question-answer pair in (197): example
(197b) is infelicitous because it does not provide an answer to question (197a) but
simply presents the speaker’s thoughts; example (197c), on the other hand, is
completely natural as an answer to (197a).
(197) a. Komt Peter morgen? [question]
comes Peter tomorrow
‘Will Peter come tomorrow?’
b. $Ik denk van ja/nee. [answer]
c. Ik denk van wel/niet. [answer]
 Argument and complementive clauses 711

It is interesting to compare the question-answer pair in (197) to the one in (198)

since this shows that we find a similar contrast between direct and indirect reported
speech constructions. It seems to confirm that van ja/nee phrases function as direct
quotes, and it might suggest that polar van wel/niet phrases have a status similar to
that of an indirect quote.
(198) a. Komt Peter morgen? [question]
comes Peter tomorrow
‘Will Peter come tomorrow?’
b. $Ik denk: “ja/nee”. [answer]
I think yes/no
c. Ik denk dat hij wel/niet komt. [answer]
I think that he AFF/not comes
‘I think that he will/wonʼt.’
That the two polar van-constructions are in complementary distribution is also
suggested by the discourse chunk in (199), which involves the denial of some
presupposed truth. Again, the response in (199b) is not felicitous given that quotes
are normally not the most suitable items to perform this function; the response in
(199c), on the other hand, does have the intended effect of denying the presupposed
truth of the proposition “Jan does not come tomorrow”.
(199) a. Jan komt morgen niet. [speaker A]
Jan comes tomorrow not
‘Jan wonʼt come tomorrow.’
b. $Dat is niet waar. Hij zei daarnet nog van ja. [speaker B]
that is not true he said just/now still VAN yes
c. Dat is niet waar. Hij zei daarnet nog van wel. [speaker B]
that is not true he said just/now still VAN AFF
‘That isnʼt true. He just told me that he would.’
The examples in (200) show again that direct quotes in direct reported speech
constructions without van pattern with van ja/nee-phrases, whereas indirect quotes
pattern with polar van wel/niet-phrases.
(200) a. Jan komt morgen niet. [speaker A]
Jan comes tomorrow not
‘Jan wonʼt come tomorrow.’
b. $Dat is niet waar. Hij zei daarnet nog: “Ja”. [speaker B]
that is not true he said just/now still yes
c. Dat is niet waar. Hij zei daarnet nog dat hij wel komt. [speaker B]
that is not true he said just/now still that he AFF comes
‘That isnʼt true. He just told me that he would.’
A final illustration of the complementary distribution of van ja/nee- and van
wel/niet-phrases is given in (201), in which speaker B indicates that the information
provided by speaker A clashes with the information available to him and, implicitly,
that he will update his knowledge state.
712 Syntax of Dutch: Verbs and verb phrases

(201) a. Jan komt morgen niet. [speaker A]

Jan comes tomorrow not
‘Jan wonʼt come tomorrow.’
b. $Bedankt, ik dacht van ja. [speaker B]
thank.you I thought VAN yes
c. Bedankt, ik dacht van wel. [speaker B]
thank.you I thought VAN AFF
‘Thanks for telling me, because I thought he would.’
Again, the examples in (202) are again added to show that van ja/nee-phrases
behave like direct quotes in reported speech constructions without van, whereas van
wel/niet-phrases behave like indirect quotes.
(202) a. Jan komt morgen niet. [speaker A]
Jan comes tomorrow not
‘Jan wonʼt come tomorrow.’
b. $ Bedankt, ik dacht: “Ja”. [speaker B]
thank.you I thought yes
c. Bedankt, ik dacht dat hij wel kwam. [speaker B]
thank.you I thought that he AFF came
‘Thanks for telling me, because I thought he would.’
To summarize the findings above, we can say that the two van-constructions in
(196) differ in that van wel/niet-phrases are normally used as a response to some
question, as a denial of some presupposed truth, or to indicate a mismatch in
information, whereas van ja/nee-phrases are simply used as direct quotes. A similar
difference in use can be observed between indirect and direct reported speech
without van, which may have led Hoeksema (2006) to the claim that, whereas van-
constructions with ja/nee are instantiations of the direct quotative van-construction,
polar van-constructions are instantiations of the indirect quotative van-construction.
We return to this suggestion in Subsection 4 after having investigated some of the
syntactic properties of van wel/niet-phrases.

3. Syntactic behavior of the van wel/niet-phrases

The previous subsection has established that direct quotes cannot be used to answer
questions; see the discussion of (197) and (198). This means that we can use
question-answer pairs to exclude unwanted intervention of direct quotative van-
readings; this is what we will do in this subsection in order to investigate the
syntactic behavior of polar van wel/niet-phrases in more detail. Paardekooper
(1986: 149-50) has shown that the internal make-up of such phrases is quite rigid.
First, the affirmative and negative adverbs wel and niet are part of a severely
restricted paradigm. Although examples such as (203) with the epistemic modals
zeker ‘certainly’ and mogelijk ‘possibly’ are sometimes taken to be acceptable, we
were not able to find any clear cases on the internet; since we consider them
degraded, we mark them with a number sign.
 Argument and complementive clauses 713

(203) a. Komt Jan straks? [yes/no question]

comes Jan later
b. #Ik denk van zeker/natuurlijk/misschien/mogelijk.
I think VAN certainly/naturally/maybe/possibly
The examples in (204) show that other adverbial phrases are also straightforwardly
excluded, which implies that polar van wel/niet-constructions are not usable as
answers to wh-questions, but always pertain to the truth or falsehood of some
proposition.
(204) a. Wie komt er morgen? a. *Ik denk van Jan. [wh-question]
who comes there tomorrow I think VAN Jan
‘Who is coming tomorrow?’
b. Wanneer komt Jan? b. *Ik denk van straks. [wh-question]
when comes Jan I think VAN later
‘When will Jan come?’
The fact that polar van wel/niet-constructions must involve the truth or falsehood of
some proposition immediately accounts for the fact, noticed both by Paardekooper
(1986) and Hoeksema (2006), that polar van wel/niet-phrases require the verb to be
non-factive; (205a) shows that polar van wel/niet with a factive verb like betreuren
‘to regret’ gives rise to a severely degraded result. A potential counterexample is
weten ‘to know’ in (205b), which is normally factive but common in the van
wel/niet-construction when combined with the adverb zeker; the reason for the
contrast between the construction with and without zeker is that the collocation
zeker weten can readily be interpreted non-factively as “to be convinced of”; see
Hoeksema (2006: 142) for further discussion of more apparent exceptions.
(205) a. *Ik betreur van wel/niet.
I regret VAN AFF/not
b. Jan wist *(zeker) van wel/niet.
Jan knew for.sure VAN AFF/not
For completeness’ sake, the examples in (206) show that the factivity restriction
also holds for non-verbal predicates. Non-factive bang zijn ‘to fear’ does allow a
polar van wel/niet-phrase whereas factive gek zijn ‘to be strange’ does not.
(206) a. Ik ben bang dat ze Peter ontslaan. [non-factive]
I am afraid that they Peter fire
‘Iʼm afraid that they will fire Peter.’
a. Ik ben bang van wel/niet.
I am afraid VAN AFF/not
b. Het is gek dat ze Peter ontslaan. [factive]
it is strange that they Peter fire
‘It is strange that they will fire Peter.’
b. *Het is gek van wel/niet.
it is strange VAN AFF/not
714 Syntax of Dutch: Verbs and verb phrases

Hoeksema (2006) collected a sample of verbs that may occur with a polar van
ja/nee- or van wel/niet-phrase on the basis of 1.000 occurrences from written
sources published after 1600. Most of these verbs occur infrequently in this
construction; we have selected those verbs that occur at least five times in the
corpus, resulting in Table (207) where the numerals indicate the number of
instances found by Hoeksema. Unfortunately, Hoeksema does not distinguish the
two constructions, and we have therefore added our own intuitions on whether the
verb in question is more normal with a van wel/niet or a van ja/nee phrase: w>j
indicates that van wel/niet is the preferred form, j>w indicates that van ja/nee is the
preferred form, and w indicates that the use of a van ja/nee-phrase is infelicitous or
even excluded. These judgments were confirmed by a more or less impressionistic
investigation of the results of a Google search on the strings [V van ja/nee] and
[V van wel/niet]. Table (207) supports Hoeksema’s (2006:150ff.) conclusion from
his diachronic investigation of polar van-constructions that the constructions with
van wel/niet-phrases are much more common in present-day Dutch than
constructions with van ja/nee-phrases (contrary to what was the case in earlier
stages of the language).
(207) Frequently occurring verbs in van ja/nee and van wel/niet phrases
aannemen ‘to assume’ 16 w schijnen ‘to seem’ 7 w
antwoorden ‘to reply’ 39 j>w schudden ‘to shake’ 52 j>w
beweren ‘to claim’ 15 w vermoeden ‘to suspect’ 16 w
denken ‘to think’ 208 w>j verzekeren ‘to assure’ 5 w
dunken ‘to think’ 13 w>j volhouden ‘to maintain’ 5 w
geloven ‘to believe’ 87 w>j vinden ‘to consider/think’ 69 w
hopen ‘to hope’ 51 w vrezen ‘to fear’ 35 w
knikken ‘to nod’ 34 j>w wedden ‘to bet’ 6 w
menen ‘to suppose’ 57 w zeggen ‘to say’ 104 w>j

Many of the verbs in Table (207) can also be used as °bridge verbs licensing
extraction of a wh-phrase from their complement clause; see Table (331) in Section
5.1.5, sub I. Of course we expect this because such bridge verbs must also be non-
factive, just like verbs taking a van wel/niet-phrase. It is interesting to note, though,
that three out of the seven verbs that do not occur in the list of bridge verbs prefer a
van ja/nee-phrase; we return to these verbs in Subsection 4.
Polar van wel/niet-phrases are also quite rigid when it comes to modification.
Modal adverbs are occasionally judged as acceptable and also occur with a very
low frequency on the internet, as was shown by a Google search (2/29/2012), on the
string [denk[±past] van ADVERB wel/niet] for the adverbs zeker ‘certainly’, misschien
‘maybe’, natuurlijk ‘naturally’, mogelijk ‘possibly’ and helaas ‘unfortunately’. We
found that zeker is used to modify both wel and niet, misschien is used to modify
wel, and helaas is used to modify niet. We did not find any cases in which the
adverbs natuurlijk or mogelijk were used as modifiers. Other adverbs seem
categorically excluded.
 Argument and complementive clauses 715

(208) a. Ik denk van zeker/natuurlijk wel. [attested cases]

I think VAN certainly/naturally AFF
b. Ik denk van zeker/helaas niet. [attested cases]
I think VAN certainly not
c. *Ik denk van morgen/hier wel/niet.
I think VAN tomorrow/here AFF/not
Paardekooper also observed that polar van wel/niet-phrases must follow the verbs in
clause-final position. The contrast between the (b)-examples in (209) shows that
they cannot occur in the middle field of the clause.
(209) a. Komt er een reorganisatie? [question]
comes there a reorganization
‘Will there be a reorganization?’
b. Jan liet duidelijk blijken dat hij dacht van wel. [answer]
Jan let clearly show that he thought VAN AFF
‘Jan made it perfectly clear that he though that there would be.’
b. ??Jan liet duidelijk blijken dat hij van wel dacht.
Jan let clearly show that he VAN AFF thought
Moreover, the examples in (210) show that topicalization of polar van wel/niet-
phrases also leads to a degraded result, regardless of whether the preposition van is
stranded or not. See Hoeksema (2008) for the same observations.
(210) a. Ik denk van wel/niet.
I think VAN AFF/not
*?
b. Van wel/niet denk ik.
b. *Wel/Niet denk ik van.
Paardekooper concluded from the immobility of polar van wel/niet-phrases that we
are not dealing with regular quotes as these normally do allow topicalization.
Subsection A has shown, however, that quotes from quotative van-constructions
must also follow the verbs in clause-final position, and this still leaves open the
possibility that polar van-constructions are indirect quotative van-constructions, as
was indeed suggested in Hoeksema (2006).

4. Are polar van-constructions quotative van-constructions?

The previous subsections have shown that polar van ja/nee-constructions must be
analyzed as direct quotative van-constructions constructions, leading to the
prediction that this type of van-construction can only occur with verbs that may
take direct quotes in reported speech without van. Hoeksema (2006) is probably
correct in claiming that this expectation is borne out; the primed examples in (211)
with beweren ‘to claim’ and geloven ‘to believe’ feel uncomfortable, and, although
they do occur on the internet, they have a much lower frequency than the
corresponding examples with zeggen ‘to say’ and denken ‘to think’. We indicate
this by using the percentage mark.
716 Syntax of Dutch: Verbs and verb phrases

(211) a. Jan zei/??beweerde: “Peter is ziek”.

Jan said/claimed Peter is ill
a. Jan zei/%beweerde van ja/nee.
Jan said/claimed VAN yes/no
b. Jan dacht/*geloofde: “Els is aardig”.
Jan thought/believed Els is nice
b. Jan dacht/%geloofde van ja/nee.
Jan thought/ believed VAN yes/no
Hoeksema also suggests that polar van wel/niet-constructions are indirect quotative
van-constructions, which is in keeping with the findings of Subsection 2 that
indirect quotes behave like polar van wel/niet phrases in that they can be used as a
response to some question, as a denial of some presupposed truth, or to indicate a
mismatch in information. Furthermore, it correctly predicts that polar van wel/niet
phrases can be used as the complement of verbs taking indirect quotes.
(212) a. Jan zei/beweerde dat Peter ziek was.
Jan said/claimed that Peter ill was
a. Jan zei/beweerde van wel/niet.
Jan said/claimed VAN AFF/not
b. Jan dacht/geloofde dat Els aardig is.
Jan thought/believed that Els nice is
b. Jan dacht/geloofde van wel/niet.
Jan thought/believed VAN AFF/not
The suggested analyses of the two polar van-constructions correctly account for the
placement of the van wel/niet-phrases: examples (209)-(210) in Subsection 3 have
shown that they behave like van-phrases in indirect quotative van-constructions
since they obligatorily follow the verbs in clause-final position. A potential problem
is that van-phrases in direct quotative van-constructions have the same property,
and we would therefore predict that polar van ja/nee-phrases also need to follow the
verbs in clause-final position. However, Hoeksema (2008:74ff.) found that this
expectation is not borne out: in about 5% of the cases the van ja/nee-phrase may
occur in the middle field of the clause. Hoeksema relates this to the fact that (short)
direct quotes may occasionally also occur in the middle field of the clause; see
Subsection IIA, for a discussion of such cases.
(213) a. dat Jan <van ja> zei <van ja>.
that Jan VAN yes said
b. dat Jan <“ja”> zei <”ja”>.
that Jan yes said
It seems that this suggestion can be supported by the examples in (214). The
primeless examples show that the verbs knikken ‘to nod’ and schudden ‘to shake
(one’s head)’ can only be combined with a direct quote if the latter precedes the
verb in clause-final position, and our judgment of the primed examples show that
the corresponding van ja/nee-phrase likewise prefers to precede the verbs in clause-
final position. Our intuitions are confirmed by a Google search (3/2.2012) on the
strings [<van ja> geknikt <van ja>] and [<van nee> geschud <van nee>], which
 Argument and complementive clauses 717

showed that preverbal placement is more frequent than postverbal placement of the
van ja/nee-phrase; the absolute numbers are given in square brackets.
(214) a. Jan heeft <“ja”> geknikt <*“ja”>.
Jan has yes nodded
a. Jan heeft <van ja> geknikt < ?van ja>. [37/12]
Jan has VAN yes nodded
b. Jan heeft <”nee”> geschud <*“nee”>.
Jan has no shaken
b. Jan heeft <van nee> geschud < ?van nee>. [24/4]
Jan has VAN no shaken

The discussion above has shown that there may indeed be reasons to think that the
polar van ja/nee- and polar van wel/niet-constructions are special instantiations of,
respectively, direct and indirect quotative van-constructions. However, the evidence
so far is still a little scanty; a more detailed investigation may therefore be needed
to provide a solid foundation for this idea.

5.1.3. Subject clauses

This section discusses subject clauses. That subject clauses are possible is strongly
suggested by the fact that the primeless examples in (215), in which the verbs
zeggen ‘to say’ and vragen ‘to ask’ take a direct object clause, can be passivized;
the resulting primed examples are likely to have a subject clause.
(215) a. Jan zei [dat de bank beroofd was].
Jan said that the bank robbed was
‘Jan said that the bank had been robbed.’
a. Er werd gezegd [dat de bank beroofd was].
there was said that the bank robbed was
‘It was said that the bank had been robbed.’
b. Marie vroeg [of de buit groot was].
Marie asked whether the catch big was
‘Marie asked whether the catch was big.’
b. Er werd gevraagd [of de buit groot was].
there was asked whether the catch big was
‘It was asked whether the catch was big.’
The acceptability of the primed examples in (215) raises the question as to whether
subject clauses can also be selected by active main verbs. Subsection I shows that
although subject clauses do not occur with intransitive and transitive verbs, they do
occur with unaccusative verbs, that is, verbs with a derived °DO-subject; from this
we may safely conclude that subject clauses are always internal °arguments of the
°matrix verb. Subsection II and III discuss, respectively, the position of subject
clauses and the use of the anticipatory subject pronoun het.

I. Verb types
Generally speaking, subject clauses do not occur with intransitive and transitive
verbs. The reason is that such verbs normally take an external argument with the
718 Syntax of Dutch: Verbs and verb phrases

function of agent. Given that clauses refer to propositions/questions/etc., and not to

agentive entities, it is expected on semantic grounds that subject clauses cannot
occur with such verbs. The examples in (216) show that using subject clauses with
(in)transitive verbs indeed gives rise to complete gibberish.
(216) a. Jan lacht. a. $Het lacht [dat Peter zingt].
Jan laughs it smiles that Peter sings
b. Jan eet spinazie. b. $Het eet spinazie [dat Marie honger heeft].
Jan eats spinach it eats spinach that Marie hungry is
There are potential counterexamples to the claim that transitive verbs do not take
subject clauses. Example (217a), for instance, shows that the transitive verb
bewijzen ‘to prove’ can easily be combined with a clausal subject. Such cases are
special, however, in that they involve factive clauses, that is, clauses the truth of
which is presupposed by the speaker. Section 5.1.2.3 has shown that normally such
clauses can be paraphrased by means of a noun phrase het feit dat ... ‘the fact that
...’, as in (217b), and that they exhibit a number of nominal properties.
(217) a. Het bewijst niets [dat Peter geen alibi heeft].
it proves nothing that Peter no alibi has
‘It proves nothing that Peter has no alibi.’
b. Het feit [dat Peter geen alibi heeft] bewijst niets.
the fact that Peter no alibi has proves nothing
‘The fact that Peter has no alibi proves nothing.’
Subject clauses are possible if they are internal arguments of the verb, as is
clear from the fact that a transitive sentence such as (218a) is easy to passivize. The
(b)-examples show that the passive counterpart of this sentence may either contain
the °expletive er or the °anticipatory pronoun het: this is a reflex of the fact that the
anticipatory pronoun is optional in (218a).
(218) a. dat Jan (het) zei [dat Peter een nieuwe auto gekocht had].
that Jan it said that Peter a new car bought had
‘that Jan said (it) that Peter had bought a new car.’
b. Er werd (door Jan) gezegd [dat Peter een nieuwe auto gekocht had].
there was by Jan said that Peter a new car bought had
‘It was said (by Jan) that Peter had bought a new car.’
b. Het werd (door Jan) gezegd [dat Peter een nieuwe auto gekocht had].
it was by Jan said that Peter a new car bought had
‘It was said by Jan that Peter had bought a new car.’
Since the examples in (218) show that °DO-subjects may be clausal, it should not
come as a surprise that we also find subject clauses with unaccusative verbs. The
examples in (219) show that this is quite common with NOM-DAT verbs; cf. Section
2.1.3. We illustrate this in the (a)-examples by means of a NOM-DAT verb that takes
zijn in the perfect tense, and in the (b)-examples by means of a NOM-DAT verb that
takes hebben.
 Argument and complementive clauses 719

(219) a. Het viel Marie erg tegen [dat Jan erover klaagde].
it disappointed Marie a.lot prt. that Jan about.it complained
‘It disappointed Marie terribly that Jan was complaining about it.’
a. Het is Marie erg tegengevallen [dat Jan erover klaagde].
it is Marie a.lot prt.-disappointed that Jan about.it complained
b. Het bevreemde Marie zeer [dat Jan erover klaagde].
it surprised Marie much that Jan about.it complained
‘It surprised Marie greatly that Jan was complaining about it.’
b. Het heeft Marie zeer bevreemd [dat Jan erover klaagde].
it has Marie much surprised that Jan about.it complained
Subject clauses are also common with psychological predicates that take an object
experiencer; cf. Section 2.5.1.3. This holds both for (220a) with the psych-verb
ergeren ‘to annoy’ and for (220b) with the periphrastic expression kwaad maken ‘to
make angry’.
(220) a. Het ergerde Peter/hem [dat Els er niet was].
it annoyed Peter/him that Els there not was
‘It annoyed Peter/him that Els wasnʼt present.’
b. Het maakte Peter/hem erg kwaad [dat Els er niet was].
it made Peter/him very angry that Els there not was
‘It made Peter very angry that Els wasnʼt present.’
Note in passing that psych-verbs such as ergeren ‘to annoy’ and many NOM-DAT
verbs are object experiencer verbs; consequently, they can be combined
successfully with conditional als-clauses; see the examples in (221). Since Section
5.1.2.1, sub VI, has shown on the basis of similar examples with subject
experiencer verbs that such als-clauses are not arguments of the verb, we need not
elaborate on this here; as illustrated in the primed examples, the fact that preposed
als-phrases can be followed by the resumptive element dan ‘then’ suggests that we
are dealing with conditional adverbial clauses.
(221) a. Het valt me op als Jan erover klaagt. [NOM-DAT verb]
it is.conspicuous me prt. if Jan about.it complains
‘I notice it when Jan complains about it.’
a. Als Jan erover klaagt (dan) valt me dat op.
if Jan about.it complains then is.conspicuous me that prt.
b. Het staat me erg tegen als Jan erover klaagt. [NOM-DAT verb]
it palls me much on if Jan about.it complains
‘It disgusts me if he complains about it.’
b. Als Jan erover klaagt (dan) staat me dat erg tegen.
if Jan about.it complains then palls me that much on
c. Het ergert me als Els er niet is. [psych-verb]
it annoys me if Els there not is
‘It annoys me if Els isnʼt present.’
c. Als Els er niet is, (dan) ergert me dat.
if Els there not is then annoys me that
720 Syntax of Dutch: Verbs and verb phrases

A conclusive argument for assuming that the als-phrases in (221) are not subject
clauses is that the subject pronoun dat in the primed examples cannot be dropped
when they occupy the sentence-initial position (that is, when dan ‘then’ is not
present). The examples in (222) show that this is compulsory when run-of-the-mill
subject clauses introduced by the complementizer dat ‘that’ occupy the initial
position, for the simple reason that inclusion of the pronoun dat leads to a clause
with two subjects.
(222) a. Dat Jan erover klaagt valt me (*dat) op.
that Jan about.it complains is.conspicuous me that prt.
b. Dat Jan erover klaagt staat me (*dat) erg tegen.
that Jan about.it complains stands me that much counter
c. Dat Els er niet is, ergert me (*dat).
that Els there not is annoys me that
Subject clauses are also very common if they function as the subject of copular
constructions, as in (223a). This is expected because such subjects are not the
external arguments of the copular, for the same reason that the direct object in the
vinden-construction in (223b) is not an internal argument of vinden. In these two
cases we are dealing with SUBJECTs of the °complementive; cf. Section 2.2.2.
(223) a. Het is vreemd [dat Els er niet is].
it is strange that Els there not is
‘Itʼs odd that Els isnʼt present.’
b. Peter vindt het vreemd [dat Els er niet is].
Peter considers it strange that Els there not is
The copular constructions in (224) show that the adjective bekend may take either a
declarative or an interrogative subject clause. The former is always possible, but the
latter only occurs if the matrix clause is negative and/or interrogative. The
complementizer of is used in the (b)-examples if the relevant decision has not been
made public yet, the complementizer dat if the decision has been made public but
has (not yet) reached the intended public.
(224) a. Het is al bekend [dat/*of Els de nieuwe voorzitter wordt].
it is already known that/whether Els the new chairman becomes
‘It is already known that Els will be the new Chair.’
b. Het is nog niet bekend [dat/of Els de nieuwe voorzitter wordt].
it is yet not known that/whether Els the new chairman becomes
‘It isnʼt known yet that/whether Els will be the new Chair.’
b. Is het al/nog niet bekend [dat/of Els de nieuwe voorzitter wordt]?
is it already/not yet known that/whether Els the new chairman becomes
‘Is it already/not yet known that/whether Els will be the new Chair?’
Again, it should be noted that we occasionally encounter als-clauses that can easily
be misanalyzed as subject clauses. That we are not dealing with subject clauses here
is clear from the fact, illustrated in (225), that such als-clauses differ from run-of-
the-mill subject clauses introduced by the complementizer dat ‘that’ in that a
 Argument and complementive clauses 721

subject pronoun must be present if the als-clause occupies the sentence-initial

position; we must therefore be dealing with conditional clauses.
(225) a. Dat Els er niet is, is (*dat) vreemd.
that Els there not is is that strange
‘that Els isnʼt present is strange.’
b. Als Els er niet is, is *(dat) vreemd.
if Els there not is is that strange
‘If Els isnʼt present, that is strange.’
Finally, we want to point out subject clauses are possible with epistemic modal
verbs; we will return to this in Section 5.2.2.2 and 5.2.3.2.
(226) a. Het kan [dat Peter morgen in Utrecht is].
it may.be.the.case that Peter tomorrow in Utrecht is
‘It may be the case that Peter will be in Utrecht tomorrow.’
b. Het schijnt [dat Peter morgen in Utrecht is].
it seems that Peter tomorrow in Utrecht is
‘It seems to be the case that Peter will be in Utrecht tomorrow.’

II. The placement of subject clauses

Subject clauses normally follow the verbs in clause-final position, as shown by the
primed examples in (219), which are repeated here for convenience as (227).
(227) a. Het is Marie erg tegengevallen [dat Jan erover klaagde].
it is Marie a.lot prt.-disappointed that Jan about.it complained
‘It has disappointed Marie terribly that Jan complained about it.’
b. Het heeft Marie zeer bevreemd [dat Jan erover klaagde].
it has Marie much surprised that Jan about.it complained
‘It has surprised Marie greatly that Jan was complaining about it.’
Subject clauses may also occur in sentence-initial position, in which case they are
optionally followed by the resumptive demonstrative pronoun dat ‘that’.
(228) a. [Dat Jan erover klaagde] (dat) is Marie erg tegengevallen.
that Jan about.it complained that is Marie a.lot prt.-disappointed
‘That Jan complained about it has disappointed Marie terribly.’
b. [Dat Jan erover klaagde] (dat) heeft Marie zeer bevreemd.
that Jan about.it complained that has Marie much surprised
‘That Jan complained about it has surprised Marie greatly.’
The examples in (229) show that it is not possible to have the subject clause in the
°middle field of the clause; see De Haan (1974) and Koster (1978). The main
clauses in the primeless examples have a non-subject in sentence-initial position
and the subject clauses of (227) and (228) in the middle field; the primed examples
provide the corresponding embedded clauses. Such examples are deemed
ungrammatical.
722 Syntax of Dutch: Verbs and verb phrases

(229) a. *Waarschijnlijk is [dat Jan erover klaagde] Marie erg tegengevallen.

probably is that Jan about.it complained Marie a.lot prt.-disappointed
a. *dat [dat Jan erover klaagde] Marie erg tegengevallen is.
that that Jan about.it complained Marie a.lot prt.-disappointed is
b. *Waarschijnlijk heeft [dat Jan erover klaagde] Marie erg bevreemd.
probably has that Jan about.it complained Marie a.lot surprised
b. *dat [dat Jan erover klaagde] Marie erg bevreemd heeft.
that that Jan about.it complained Marie a.lot surprised has
We should note, however, that the examples seem at least marginally acceptable if
the clause is interpreted as factive: (het feit) dat Jan erover klaagde. If this is the
case, it would not be surprising, considering that Section 5.1.2.3 has shown that
factive clauses are more generally used in nominal argument positions. Example
(230) provides instances in which the subject clause is more clearly factive, and we
believe that these cases are indeed possible (provided that the clause does not
become too lengthy).
(230) a. Natuurlijk bewijst [(het feit) [dat Peter geen alibi heeft]] absoluut niets.
of.course proves the fact that Peter no alibi has absolutely nothing
‘Of course, the fact that Peter has no alibi proves absolutely nothing.’
b. dat [(het feit) [dat Peter geen alibi heeft]] absoluut niets bewijst.
that the fact that Peter no alibi has absolutely nothing proves
‘that the fact that Peter has no alibi proves absolutely nothing.’
Koster (1978) concludes from the fact that subject clauses cannot occur in the
regular subject position in the middle field of the clause that subject sentences do
not exist. He also proposes that the clauses in (228) are not sentence-internal, but
function as sentence-external satellites that bind a (possibly phonetically empty)
subject pronoun; actually, according to Koster, we are dealing with a kind of °left-
dislocation constructions. If we assume that pronouns are moved from the regular
subject position into sentence-initial position, examples such as (228a) are analyzed
as in (231a) if the demonstrative pronoun is present, and as in (231b) if it is not.
(231) a. [Dat Jan erover klaagde]i [sentence dati is ti Marie erg tegengevallen].
b. [Dat Jan erover klaagde]i [sentence proi is ti Marie erg tegengevallen].
Koster’s proposal was challenged in Klein (1979). An important reason is that the
prosody of the examples with and without the resumptive pronoun dat differ
markedly: while in the former case the clause is normally separated from the
sentence by an intonation break, the clause can be prosodically integrated in the
sentence in the latter case, as indicated in (232), in which the comma indicates the
obligatory intonation break.
(232) a. [Dat Jan erover klaagde], dat is Marie erg tegengevallen.
b. [Dat Jan erover klaagde] is Marie erg tegengevallen.
If Klein’s conclusion that the clause in (232b) is sentence-internal is correct, we
should account for the fact that the clause cannot occur in the regular subject
position in the examples in (229) by claiming that clauses cannot surface in nominal
 Argument and complementive clauses 723

argument positions. This is in fact the same conclusion drawn for object clauses in
Section 5.1.2.2, sub III, to which we refer the reader for further discussion. We will
investigate the pros and cons of Koster’s proposal in our discussion of
topicalization in Section 11.3.2.

III. The anticipatory pronoun het and expletive er

Like object clauses, subject clauses cannot be preposed in sentences that contain the
°anticipatory pronoun het, as shown in (233b). This would follow immediately from
Koster’s left-dislocation analysis as the object pronoun must be replaced by the
resumptive pronoun dat or its phonetically empty counterpart pro. The structures in
(231) show that the position of het in (233b) is already occupied by a °trace.
(233) a. Het is Marie erg tegengevallen [dat Jan erover klaagde].
it is Marie a.lot prt.-disappointed that Jan about.it complained
‘It has greatly disappointed Marie that Jan complained about it.’
b. [Dat Jan erover klaagde] is (*het) Marie erg tegengevallen.
that Jan about.it complained is it Marie a.lot prt.-disappointed
The analysis must be slightly different if we accept Klein’s conclusion that the
subject clause occupies the sentence-initial position if the demonstrative pronoun
dat is not present. We then have to assume that the subject clause has not been
moved into clause-initial position in one fell swoop but has moved via the regular
subject position; the anticipatory pronoun is then blocked given that the subject
position is occupied by a trace of the clause. See Section 5.1.2.2, sub III, for a more
extensive discussion of this option.
The (b)-examples in (234) show that subject clauses cannot be preposed in
clauses that contain expletive er either; er can only be interpreted as an adverbial
phrase of place in these examples. The reason for this is different, however, than in
the case of het; expletive er can only be used if the subject is part of the °focus (new
information) of the clause, whereas preposed subject clauses are normally
interpreted as being part of the presupposition of the clause.
(234) a. Er is gebleken [dat de software goed werkt].
there is turned.out that the software well works
‘It has turned out that the software is working well.’
b. [dat de software goed werkt] dat is (#er) gebleken.
that the software well works that is there turned.out
b. [dat de software goed werkt] is (#er) gebleken.
that the software well works is there turned.out
The option of having the anticipatory pronoun het or the expletive er is not only
affected by the position of the subject clause. In examples with a °complementive,
the position of the secondary predicate may also be relevant. With a sentence-initial
predicate het is preferably dropped and er becomes completely impossible.
724 Syntax of Dutch: Verbs and verb phrases

(235) a. Het/Er is duidelijk geworden [dat Jan de nieuwe voorzitter wordt].

it/there is clear become that Jan the new chairman becomes
‘It has become clear that Jan will become the new chairman].’
b. Duidelijk is ( ?het) geworden [dat Jan de nieuwe voorzitter wordt].
clear is it become that Jan the new chairman becomes
b. Duidelijk is (*er) geworden [dat Jan de nieuwe voorzitter wordt].
clear is there become that Jan the new chairman becomes
The examples in (236) show that we may find the same phenomenon in perfect-
tense constructions with °monadic unaccusative verbs taking subject clauses like
blijken ‘to turn out’: with topicalized participles, het and er cannot be properly
realized. Examples with het and er do occur on the internet but are very rare.
(236) a. Het/Er is gebleken [dat vette vis gezond is].
it/there is turned.out that oily fish healthy is
‘It has turned out that oily fish is healthy.’
b. Gebleken is (?het) [dat vette vis gezond is].
turned.out is it that oily fish healthy is
b. Gebleken is (?er) [dat vette vis gezond is].
turned.out is there that oily fish healthy is
Although we are not aware of any theoretical account for the markedness of the
primeless (b)-examples in (235) and (236), we hypothesize that examples of this
type involve some kind of locative inversion of the type we find in English. Den
Dikken and Næss (1993) have argued that in examples like Down the hill rolled a
baby carriage the predicative PP down the hill has been topicalized via the regular
subject position, and that the subject occupies its base position in the °small clause
headed by the moved predicate; [CP Down the hilli [TP ti rolled [SC the baby carriage
ti]]]. If we assume something similar for examples such as (235b), insertion of the
pronoun het may be blocked given that the regular subject position is occupied by a
trace of the moved predicate. A potential problem for this analysis is that this leaves
unexplained why insertion of het seems to be marginally possible. Furthermore, it
remains to be seen whether an analysis of this sort can be extended to examples
such as (236b). The the degraded status of the primed (b)-examples can again be
related to the information structure of the clause if °left dislocation/topicalization of
the predicate is only possible if it is part of the presupposition of the clause. We
leave it to future research to investigate whether proposals along these lines are
viable.
Example (218) in Subsection I has shown that in passive constructions the
choice between het and er is related to the question as to whether the object clause
in the corresponding passive construction can be combined with the anticipatory
pronoun het. It seems that, as in English, clause-final subject clauses in active
sentences can always be introduced by anticipatory het and that in many cases they
can also be combined with expletive er. The semantic difference between the two
options is not always clear, but we may suppose that the choice between the two
options depends on whether the subject clause is presented as part of the
presupposition or the focus of the sentence. A Google search (1/27/2012) shows
that the frequencies of het and er in examples such as (237) are both high.
 Argument and complementive clauses 725

(237) a. Het is duidelijk geworden dat ... [presupposition]

it is clear become that
‘It has become clear that ...’
b. Er is duidelijk geworden dat ... [focus]
there is clear become that
‘It has become clear that ...’
An appeal to the information structure of the sentence seems supported by examples
like those in (238). Given that interrogative clauses are less likely to be interpreted
as presuppositional than declarative clauses, we expect examples such as (238a) to
be extremely rare (despite being definitely grammatical). A Google search
(3/22/2013) on this string shows that this expectation is indeed borne out: it resulted
in no more than 4 hits. Strings such as (238b), on the other hand, are very frequent.
(238) a. Het werd gevraagd of ... [presupposition]
it was asked whether
‘It was asked whether ...’
b. Er werd gevraagd of ... [focus]
there was asked whether
‘It was asked whether ...’
Given the result of our Google searches mentioned above, one would also expect
the frequency of examples such as (239a) to be much lower than examples such as
(239b). This expectation is, however, not borne out: we found 225 cases of the two
strings in (239a) and only 13 of the two strings in (239b).
(239) a. Het is niet/nooit duidelijk geworden of ...
it is not/never clear become whether
‘It has not/never become clear whether ...’
b. Er is niet/nooit duidelijk geworden of ...
there is not/never clear become whether
‘It has not/never become clear whether ...’
The results of our Google searches on the examples in (239) show that there must
be other factors, yet to be identified, that must be involved in the choice between
het and er. One factor that springs to mind is that the choice is related to the type of
predicate, but we leave this for future research.

5.1.4. Prepositional object clauses?

This section on finite prepositional object (PO-)clauses is relatively short given that
many of their properties and of the °anticipatory pronominal PPs introducing them
have been discussed in Section 2.3. PO-clauses never have the form of main clauses
and come in two types: declarative clauses introduced by the complementizer dat
‘that’ or interrogative clauses introduced by the complementizer of ‘whether’ or
some wh-phrase. Some examples are given in (240). The question as to whether a
declarative or an interrogative clause will be used depends on the verb: a verb like
klagen (over) ‘to complain (about)’ in (240a) selects a declarative clause, whereas
the verb twijfelen (over) in the (b)-examples selects an interrogative clause.
726 Syntax of Dutch: Verbs and verb phrases

(240) a. dat Jan (erover) klaagde [dat Marie hem steeds plaagt]. [declarative]
that Jan about.it complained that Marie him always teases
‘that Jan complained about it that Marie teases him all the time.’
b. dat Jan (erover) twijfelt [of hij het boek zal lezen]. [interrogative]
that Jan about.it is.in.doubt whether he the book will read
‘that Jan is in doubt about whether heʼll read the book.’
b. dat Jan (erover) twijfelt [welk boek hij zal lezen]. [interrogative]
that Jan about.it is.in.doubt which book he will read
‘that Jan is in doubt about which book heʼll read.’
The examples in (240) show that clause-final PO-clauses can be introduced by an
anticipatory pronominal PP in the °middle field of the clause. Depending on the
verb in question, this PP can be optional or obligatory. The former holds for the
verbs in (240) and the latter for the verbs in (241). An extensive sample of PO-
verbs that can or cannot drop the anticipatory pronominal PP can be found in
Section 2.3.1, sub VI.
(241) a. dat Jan *(ervan) geniet [dat hij rijk is].
that Jan of.it enjoys that he rich is
‘that Jan enjoys it that heʼs rich.’
b. dat Jan *(erop) rekent [dat Marie zal komen].
that Jan on.it counts that Marie will come
‘that Jan counts on it that Marie will come.’
PO-clauses can be left-dislocated, in which case the anticipatory pronoun is
replaced by a resumptive pronominal PP in the form of daar + P. This is illustrated
in (242) for the examples in (240); the (discontinuous) resumptive PP is given in
italics.
(242) a. [Dat Marie hem steeds plaagt], daar klaagde Jan over.
that Marie him always teases there complained Jan about
b. [Of hij het boek zal lezen], daar twijfelt Jan over.
whether he the book will read there is.in.doubt Jan about
‘Whether heʼll read the book, that Jan is in doubt about.’
b. [Welk boek hij zal lezen], daar twijfelt Jan over.
which book he will read there is.in.doubt Jan about
Although some speakers seem to allow omission of the pronominal part of the
resumptive PP, most people reject this. This is indicated in (243), in which pro
stands for the empty/deleted resumptive pronominal part, by means of a percentage
sign.
%
(243) a. [dat Marie hem steeds plaagt] pro klaagde Jan over.
that Marie him always teases complained Jan about
b. %[of hij het boek zal lezen] pro twijfelt Jan over.
whether he the book will read is.in.doubt Jan about
‘Whether heʼll read the book, Jan doubts.’
%
b. [Welk boek hij zal lezen] pro twijfelt Jan over.
which book he will read is.in.doubt Jan about
 Argument and complementive clauses 727

Note in passing that the fact that most speakers do not accept examples such as
(243) may be a potential problem for Koster’s proposal discussed in Sections
5.1.2.2, sub III, and 5.1.3, sub II, the substance of which was that apparent
sentence-initial object and subject clauses are actually left-dislocated and that the
first position of the sentence is in fact filled by an empty pronominal element pro. If
the resumptive pronoun can be phonetically empty in the case of subject and object
clauses, why is this excluded for most speakers in the case of PO-clauses? Note also
that the examples in (243) do not improve if the prepositional part of the resumptive
pronominal PP is left out. Although some speakers may perhaps marginally accept
examples such as (244), they contrast sharply with the examples without an
anticipatory PP in (240), which are fully grammatical.
??
(244) a. [Dat Marie hem steeds plaagt] klaagde Jan.
that Marie him always teases complained Jan
b. *?[Of hij het boek zal lezen] twijfelt Jan.
whether he the book will read is.in.doubt Jan
‘that Jan is in doubt whether heʼll read the book.’
b. *?[Welk boek hij zal lezen] twijfelt Jan.
which book he will read is.in.doubt Jan
PO-clauses cannot be placed in the middle field of the clause, irrespective of
whether or not an anticipatory PP is present. PO-clauses normally do not occur as
part of the PP-complement of the verb either; examples such as (245) are quite
marked compared to examples such as (240), which is indicated here by means of a
question mark (although Section P2.4.1.1, discusses a number of exceptional
circumstances that do seem to license PPs of the sort in (245)).
?
(245) a. dat Jan klaagde [PP over [dat Marie hem steeds plaagt]].
that Jan complained about that Marie him always teases
b. ?dat Jan twijfelt [PP over [of hij het boek zal lezen]].
that Jan is.in.doubt about whether he the book will read
b. ?dat Jan twijfelt [PP over [welk boek hij zal lezen]].
that Jan is.in.doubt about which book he will read
We want to conclude by noticing that there are als-clauses that can easily be
erroneously analyzed as PO-clauses; We refer the reader to Paardekooper
(1986:1.18.9, B2) for a concrete case of such a misanalysis, but we will use
example (246a) for reasons of representation. The two examples in (246a) differ,
however, in that the als-clause but not the dat-clause can be followed by dan ‘then’
if the als-clause occurs on the first position of the utterance, suggesting that we are
dealing with a conditional adverbial clause. This suggestion is supported by the fact
that there is a sharp difference between the two variants of example (246c), in
which the clauses appear as part of the PP-complement: whereas the dat-clause
gives rise to a marked but interpretable result, the als-clause gives rise to an
unacceptable and uninterpretable result.
728 Syntax of Dutch: Verbs and verb phrases

(246) a. Jan klaagt er altijd over [dat/als het regent].

Jan complains there always about that/if it rains
‘Jan always complains about it that/if it rains.’
b. [Als/*dat het regent], dan klaagt Jan er altijd over.
if/that it rains then complains Jan there always about
‘If it rains, Jan is always complaining about it.’
c. Jan klaagt altijd [PP over [?dat/*als het regent]].
Jan complains always about that/if it rains
Literally: ‘Jan always complains about that it rains.’
A final argument for assuming that the als-clause is a conditional adverbial clause
is that it can occupy the sentence-initial position while the anticipatory pronominal
PP erover is present; if the als-clause were a PO-clause, we would end up with two
prepositional objects within a single clause. Analyzing the als-clause as a
conditional adverbial phrase, on the other hand, is unproblematical given that we
may then give (247a) a similar analysis as (247b).
(247) a. [Als het regent] klaagt Jan er altijd over.
if it rains complains Jan there always about
‘If it rains, Jan is always complaining about it.’
b. [Als het regent] klaagt Jan altijd over reuma.
if it rains complains Jan always about rheumatism
‘If it rains, Jan is always complaining about rheumatism.’
For more discussion of the incorrect analysis of conditional als-clauses as argument
clauses we refer the reader to Sections 5.1.2.1, sub VI, to 5.1.3, sub I, for similar
cases in the domain of object and subject clauses.

5.1.5. Fragment clauses

Fragment clauses cannot be immediately recognized as such because they do not
contain an overt finite verb and, consequently, look like phrases of some non-verbal
category. There are two types of fragment clauses: FRAGMENT WH-QUESTIONs and
FRAGMENT ANSWERs. Examples of the former are given in the primed examples in
(248), which show that fragment wh-questions can plausibly be analyzed as
phonetically reduced finite interrogative clauses.
(248) a. Jan heeft gisteren iemand bezocht. [speaker A]
Jan has yesterday someone visited
‘Jan visited someone yesterday.’
a. Wie heeft Jan gisteren bezocht? [speaker B]
who has Jan yesterday visited
‘Who (did he visit yesterday)?’
b. Jan heeft Marie bezocht. [speaker A]
Jan has Marie visited
‘Jan has visited Marie’
b. Wanneer heeft Jan Marie bezocht? [speaker B]
when has Jan Marie visited
‘When (did Jan visit Marie)?’
 Argument and complementive clauses 729

Ross (1967) derived fragment wh-questions by means of a deletion operation that

he referred to as SLUICING, and fragment wh-questions are therefore also known as
sluicing constructions; the suppressed information is indicated here by means of
strikethrough. At first sight, the deletion seems licensed simply by the presence of
some antecedent clause in the preceding discourse, which contains some (implicit)
correlate of the wh-phrase constituting the fragment wh-question, but our discussion
below will bear out that on closer scrutiny the situation is more complex.
The examples in (249) show that fragment answers may arise in conversation
as a response to wh-questions; the suppressed information is again indicated by
strikethrough.
(249) a. Wat heeft Jan gisteren gekocht? [speaker A]
what has Jan yesterday bought
‘What did Jan buy yesterday?’
a. Een nieuwe computer heeft Jan gisteren gekocht. [speaker B]
a new computer has Jan yesterday bought
‘A new computer (Jan bought yesterday).’
b. Wanneer heeft Jan die nieuwe computer gekocht? [speaker A]
when has Jan that new computer bought
‘When did Jan buy that new computer?’
b. Gisteren heeft Jan die nieuwe computer gekocht. [speaker B]
yesterday has Jan that new computer bought
‘Yesterday (Jan bought that new computer).’
The non-reduced clauses corresponding to the fragment clauses in the examples
above are grammatical but less felicitous, for reasons of economy, given that the
suppressed information can easily be reconstructed from the context; usually the
preceding discourse contains some antecedent clause which provides the
information suppressed in the fragment clause. Nevertheless, we cannot a priori
assume that the deletion analysis suggested above is correct, especially because it
runs into several problems. Establishing that we are dealing with some kind of
reduction will therefore be an essential part of our discussion of fragment clauses.
After having established this, we will discuss the properties of fragment clauses in
greater detail. Fragment wh-questions are discussed in Subsection I and fragment
answers in Subsection II.

I. Fragment wh-questions (sluicing)

The examples in (250) show that fragment wh-questions do not only occur as
independent utterances but also as subparts of clauses. If we are indeed dealing with
reduced clauses, this would show that sluicing can apply to °matrix and embedded
clauses alike.
(250) a. Jan heeft gisteren iemand bezocht. [speaker A]
Jan has yesterday someone visited
‘Jan visited someone yesterday.’
a. Kan je me ook zeggen wie Jan gisteren bezocht heeft? [speaker B]
can you me also tell who Jan yesterday visited has
‘Can you tell me who (Jan visited yesterday)?’
730 Syntax of Dutch: Verbs and verb phrases

b. Jan heeft gisteren iemand bezocht, maar ...

Jan has yesterday someone visited but
... ik weet niet wie Jan gisteren bezocht heeft?
I know not who Jan yesterday visited has
‘Jan visited someone yesterday, but I donʼt know who.’
The following subsections discuss fragment wh-questions in more detail.
Subsection A begins by showing that fragment wh-questions are indeed clauses, and
that we must therefore assume that some sort of sluicing operation is at work here.
This need not imply, however, that sluicing must be seen as a deletion operation.
Subsection B shows that there are at least two ways of analyzing sluicing, which in
fact both face a number of challenges. Subsection C continues by investigating to
what extent the interpretatively present but phonetically non-expressed part of the
fragment wh-question must be isomorphic to some antecedent clause. Subsection D
investigates the correlate of the wh-phrase in the antecedent clause. Subsection E
concludes with a number of specific examples that may involve sluicing.

A. Fragment wh-questions are clauses

This subsection reviews the evidence in favor of the claim that fragment wh-
questions are really clauses. We will follow the literature in mainly discussing
examples of the sort in (250b), but this is not a matter of principle; similar
arguments can be given on the basis of examples such as (250a).

1. Selection restrictions
A first argument for claiming that fragment wh-questions are clauses is based on the
selection restrictions imposed by the verb on its complements; embedded fragment
wh-questions can only occur with predicates that select interrogative clauses. The
primeless examples in (251) illustrate that verbs like weten ‘to know’ and zien ‘to
see’ may take an interrogative clause and the primed examples show that they may
likewise take an embedded fragment wh-question. Examples such as (251a) are
especially telling given that the verb weten ‘to know’ can only be combined with a
severely limited set of noun phrases, and noun phrases referring to objects are
certainly not part of this set (contrary to what is the case with its English counter-
part to know): cf. Ik weet het antwoord/*dat boek ‘I know the answer/that book’.
(251) a. Ik weet [wat Jan gekocht heeft].
I know what Jan bought has
‘I know what Jan has bought.’
a. Jan heeft iets gekocht maar ik weet niet wat.
Jan has something bought but I know not what
‘Jan bought something but I donʼt know what.’
b. Ik zag [wie er wegrende].
I saw who there away-ran
‘I saw who ran away.’
b. Er rende iemand weg en ik zag ook wie.
there ran someone away and I saw also who
‘Someone ran away, and I also saw who.’
 Argument and complementive clauses 731

The examples in (252) show that verbs like beweren ‘to claim’, which do not select
interrogative clauses, cannot be combined with fragment wh-questions either.
(252) a. *Marie beweert [wat Jan gekocht heeft].
Marie claims what Jan bought has
b. *Peter denkt dat Jan iets gekocht heeft *(en Marie beweert wat).
Peter thinks that Jan something bought has and Marie claims what

2. Coordination
A second argument for assuming that fragment wh-questions are clauses can be
based on coordination: given that coordination is normally restricted to phrases of
the same categorial type, the fact that full clauses fragment wh-questions can be
coordinated with fragment wh-questions suggests that the first are also clauses.
(253) a. Jan vroeg me [[waar ik gewoond had] en [hoe lang]].
Jan asked me where I lived had and how long
‘Jan asked me where I had lived and for how long.’
b. Ik weet niet [[wat hij gedaan heeft] of [waarom]].
I know not what he done has or why
‘I donʼt know what he has done or why.’

3. Case assignment
A third argument is based on case assignment: the wh-phrase constituting the overt
part of the fragment wh-question in (254a) is assigned the same case as the
corresponding phrase in the antecedent clause and not the case normally assigned
by the embedding predicate. One must keep in mind, however, that cases like these
may be misleading as they may involve N-ellipsis on top of sluicing. An argument
in favor of such an analysis is that the possessive pronoun wiens in (254b) does not
have a syntactic correlate in the antecedent clause, whereas the noun phrase wiens
auto does.
(254) a. Jan heeft iemands boek gelezen, maar ik weet niet wiens.
Jan has someoneʼs book read but I know not whose
‘Jan has read someoneʼs book but I donʼt know whose.’
b. Er staat een auto op de stoep, maar ik weet niet wiens.
there stands a car on the pavement but I know not whose
‘There is a car on the pavement but I donʼt know whose.’
Since Dutch has overt case marking on pronominal possessives only, we cannot
provide any better evidence than cases such as (254), but Merchant (2001/2006)
provides a number of examples from German (and other languages) that involve
nominal arguments. Although the verb wissen ‘to know’ governs °accusative case,
the wh-phrase that constitutes the fragment wh-question in (255) has °dative case
just like the complement of the verb schmeicheln ‘to flatter’ in the antecedent
clause.
(255) Er will jemandemdat schmeicheln, aber sie wissen nicht wemdat/*wenacc.
he wants someone flatter but they know not who/who
‘He wants to flatter someone, but they donʼt know who.’
732 Syntax of Dutch: Verbs and verb phrases

4. Syntactic distribution/placement of fragment wh-questions

The most important argument for claiming that fragment wh-questions are clauses
involves the syntactic distribution of embedded fragment wh-questions like Wie?
‘Who?’ or Wat? ‘What?’. If such fragment wh-questions were noun phrases, we
would expect them to have the distribution of nominal phrases and hence to appear
before the clause-final verbs. If, on the other hand, such fragment wh-questions are
clauses, we expect them to occur in the normal position of clauses, that is, after the
clause-final verbs. The examples in (256) therefore unambiguously show that
fragment wh-questions are clauses.
(256) a. Jan heeft iets gekocht en ik denk dat ik weet wat.
Jan has something bought and I think that I know what
‘Jan has bought something and I think that I know what.’
b. *Jan heeft iets gekocht en ik denk dat ik wat weet.
Jan has something bought and I think that I what know
The examples in (257) show that, like regular object clauses, fragment wh-questions
functioning as direct object can only occur to the left of the clause-final verbs if
they are topicalized or left-dislocated. The relevant sluicing construction is given in
the second conjunct of (257b).
(257) a. [Wat hij gekocht heeft] (dat) weet ik niet.
what he bought has that know I not
‘What he bought, (that) I donʼt know.’
b. Hij heeft iets gekocht, maar wat (dat) weet ik niet.
he has something bought but what that know I not
‘He bought something but what (that) I donʼt know.’

5. The anticipatory pronoun het

Yet another argument involves the distribution of the anticipatory pronoun het. We
would expect this pronoun to be possible if fragment wh-questions are clauses, but
not if they are some non-verbal category. The examples in (258) show that the
results are somewhat mixed: the (a)-examples show that fragment wh-questions
functioning as objects cannot co-occur with the anticipatory pronoun het, whereas
the (b)-examples show that fragment wh-questions functioning as subjects can.
(258) a. Ik weet (het) nog niet [wie er morgen komt].
I know it yet not who there tomorrow comes
‘I donʼt know yet who is coming tomorrow.’
a. Er komt morgen iemand, maar ik weet (*het) nog niet wie.
there comes tomorrow someone but I know it yet not who
‘Someone will be coming tomorrow, but I donʼt know yet who.’
b. Het is nog niet duidelijk [wie er morgen komt].
it is yet not clear who there tomorrow comes
‘It isnʼt clear yet who will come tomorrow.’
b. Er komt morgen iemand, maar het is nog niet duidelijk wie.
there comes tomorrow someone but it is yet not clear who
‘Someone will be coming tomorrow, but it isnʼt clear yet who.’
 Argument and complementive clauses 733

A possible account for the contrast between the two primed examples in (258) may
be that fragment wh-questions are always part of the °focus (new information) of
the clause, as is clear from the fact that they are always assigned focus accent.
Section 5.1.1, sub III, has shown that the anticipatory object pronoun het tends to
trigger a presuppositional reading of the object clause; so it may be that combining
it with a fragment wh-question results in an incoherent information structure, which
may account for the judgment given in (258a). Although Section 5.1.3, sub III, has
shown that the anticipatory subject pronoun het can sometimes likewise trigger a
presuppositional reading of the subject clause, there are also many cases in which
this effect does not arise; this means that the information structure of example
(258b) may be fully coherent, regardless of whether the anticipatory pronoun is
present or not. We leave it to future research to establish whether this account of the
contrast between the two primed examples in (258) is tenable, but conclude for the
moment that the acceptability of the anticipatory pronoun het in examples such as
(258b) provides support for the claim that fragment wh-questions are clauses.

6. Left dislocation
The argument on the basis of the anticipatory pronoun can be replicated in a slightly
more straightforward form on the basis of °left-dislocation constructions such as
(259); the primed examples show that the resumptive pronoun dat ‘that’ is possible
with fragment wh-questions, irrespective of the latter’s function.
(259) a. [Wie er morgen komt] dat weet ik nog niet.
who there tomorrow comes that know I not yet
‘Who is coming tomorrow, that I donʼt know yet.’
a. Er komt morgen iemand, maar wie dat weet ik nog niet.
there comes tomorrow someone but who that know I yet not
‘Someone will be coming tomorrow, but who, that I donʼt know yet.’
b. [Wie er morgen komt] dat is nog niet duidelijk.
who there tomorrow comes that is yet not clear
‘Who is coming tomorrow, that isnʼt clear yet.’
b. Er komt morgen iemand, maar wie dat is nog niet duidelijk.
there comes tomorrow someone but who that is yet not clear
‘Someone will be coming tomorrow, but who, that isnʼt clear yet.’
It should be noted that the possibility of left dislocation strongly disfavors the
nominal analysis of fragment wh-questions. First, example (260) shows that left
dislocation is normally excluded with wh-phrases.
(260) a. Wat (*dat) wil je kopen?
what that want you buy
‘What do you want to buy?’
b. Welke boeken (*die) wil je kopen?
which books these want you buy
‘Which books do you want to buy?’
Second, the primeless examples in (261) show that resumptive pronouns normally
exhibit number agreement with left-dislocated noun phrases, whereas the primed
734 Syntax of Dutch: Verbs and verb phrases

examples show that left dislocation of fragment wh-clauses involves the invariant
form dat ‘that’, that is, the form normally found with left-dislocated clauses.
(261) a. Het boek, dat wil ik kopen.
the book that want I buy
a. Jan wil een boek kopen, maar welksg dat weet ik niet.
Jan wants a book buy but which that know I not
b. De boeken, die/*dat wil Jan kopen.
the books those/that want Jan buy
b. Jan wil wat boeken kopen, maar welkepl dat/*die weet ik niet.
Jan wants some books buy, but which that/these know I not

7. Nominalization
Nominalization also provides evidence for the claim that fragment wh-questions are
clauses. First, the (a)-examples in (262) show that nominal objects of verbs
normally appear as van-PPs in the corresponding nominalizations; cf. N2.2.3.2.
Second, the (b)-examples show that object clauses are never preceded by a
preposition. The fact that the nominalization in (262b) does not contain the
preposition van thus shows that fragment clauses are not nominal, but clausal.
(262) a. Jan rookt sigaren.
Jan smokes cigars
a. [Het roken *(van) sigaren] is ongezond.
the smoking of cigars is unhealthy
b. Marie vroeg [waarom Jan sigaren rookt]. b. Marie vroeg waarom.
Marie asked why Jan cigars smokes Marie asked why
‘Marie asked why Jan smokes cigars.’
b. de vraag [waarom Jan sigaren rookt] b. de vraag waarom
the question why Jan cigars smokes the question why
‘the question as to why Jan smokes cigars’

8. Subject-verb agreement
The final argument again pertains to fragment wh-questions functioning as subjects.
If fragment wh-questions are really clauses, we expect finite verbs to exhibit
(default) singular agreement throughout, whereas we would expect finite verbs to
agree in number with nominal fragment wh-questions if they are not. The examples
in (263) show that the former prediction is the correct one; finite verbs are always
singular even if the fragment wh-question has the form of a plural noun phrase.
(263) a. Het is niet duidelijk [welke boeken Jan wil hebben].
it is not clear which books Jan wants.to have
‘It isnʼt clear which books Jan wants to have.’
a. Jan wil wat boeken hebben, maar het is/*zijn niet duidelijk welke.
Jan wants.to some books have but it is/are not clear which
‘Jan wants to have some books, but it isnʼt clear which.’
 Argument and complementive clauses 735

b. [Welke boeken Jan wil hebben] is niet duidelijk.

which books Jan wants.to have is not clear
‘Which books Jan wants to have isnʼt clear.’
b. Jan wil wat boeken hebben, maar welke is/*zijn niet duidelijk.
Jan wants some books have but which is/are not clear
‘Jan wants to have some books, but which ones isnʼt clear.’

B. What is Sluicing?
The previous subsection has shown that there is overwhelming evidence in favor of
the claim that fragment wh-questions are clausal in nature, and hence that
something like sluicing must exist. Let us assume the standard generative claim
discussed in Section 9.1 that embedded finite interrogative clauses have the CP/TP
structure in (264a), and that the wh-element occupies the position preceding the
(phonetically empty) complementizer indicated by C. Sluicing can then be derived
in at least two ways: the phonetic content of TP might be deleted under identity
with its antecedent clause in the preceding discourse, or the TP might be
phonetically empty right from the start and function as a pro-form that can be
assigned an interpretation on the basis of its antecedent clause. The two options
have been indicated in the (b)-examples in (264), in which strikethrough stands for
deletion of the phonetic content of the TP and e for an empty pro-form replacing
TP.
(264) a. Ik weet niet [CP wati C [TP Jan gekocht ti heeft]].
I know not what Jan bought has
‘I donʼt know what Jan has bought.’
b. Ik weet niet [CP wati C [TP Jan gekocht ti heeft]].
b. Ik weet niet [CP wat C [TP e ]].
We will not attempt to compare the two analyses here, but confine ourselves to
mentioning a series of problems that must be solved by any proposal that claims
that fragment wh-questions are CPs with a phonetically empty TP; readers who are
interested in a comparison of the two analyses are referred to Merchant
(2001/2006), who also discusses a number of other proposals, such as the idea that
fragment wh-questions are reduced wh-cleft-constructions: Wat is het dat Jan
gekocht heeft ‘What is it that Jan has bought?’. Because it is easier for reasons of
exposition, we will follow Merchant’s (2001/2006) wh-movement + TP deletion
approach in (264b) in our structural representations, without intending to imply,
however, that we consider this approach superior or inferior to the TP pro-form
approach.

1. Sluicing is possible in wh-questions only

A first problem that should be accounted for is that sluicing is generally impossible
outside the domain of fragment wh-questions. This is illustrated in the (a)-examples
in (265): the first conjunct Jan is hier may not give rise to sluicing in the declarative
object clause in the second conjunct, although it can be pronominalized by means of
the pronoun het/dat. The same thing is illustrated in the (b)-examples which involve
an embedded yes/no question. The unacceptability of the primeless examples shows
736 Syntax of Dutch: Verbs and verb phrases

that we need to formulate certain non-trivial conditions on the application of

sluicing to ensure that it gives rise to fragment wh-questions only.
(265) a. *Jan is hier maar Peter mag niet weten [dat hij hier is]
Jan is here but Peter is.supposed.to not know that he here is
a. Jan is hier maar Peter mag het/dat niet weten
Jan is here but Peter is.supposed.to it/that not know
‘Jan is here but Peter isnʼt supposed to know it/that,’
b. *Jan komt misschien maar niemand weet zeker [of hij komt].
Jan comes maybe but nobody knows for.sure whether he comes
b. Jan komt misschien maar niemand weet het/dat zeker.
Jan comes maybe but nobody knows it/that for.sure
‘Jan may be coming but nobody knows it/that for certain.’
For completeness’ sake, example (266b) shows that sluicing is not possible in the
domain of wh-exclamatives either.
(266) a. Het is ongelooflijk [wat een boeken Els geschreven heeft]!
it is incredible what a books Els written has
‘It is incredible how many books Els has written!’
b. *Els heeft veel geschreven; het is vooral ongelooflijk wat een boeken.
Els has a.lot written it is especially incredible what a books

2. The overt part does not include elements in the C-position

A second problem that should be solved is that fragment wh-questions normally
cannot contain material that is not part of the wh-phrase. Some speakers of Dutch
allow the overt realization of the complementizer of in embedded clauses, but,
contrary to what is to be expected on the basis of the analyses in (264), example
(267a) shows that the complementizer does not surface in embedded fragment wh-
questions. The (b)-examples further show that, under the standard analysis that
finite verbs occupy the C-position in interrogative main clauses, it would predict
wrongly that non-embedded fragment wh-questions like Wat? ‘What?’ should
contain a finite verb.
(267) a. Ik weet niet [CP wati of [TP hij ti zei]].
I know not what COMP he said
‘I donʼt know what he said.’
a. *Hij zei iets maar ik weet niet [CP wati of [TP hij ti zei]].
he said something but I know not what COMP he said
‘He said something but I donʼt know what.’
b. Hij zei iets.
he said something
b. *[CP Wat [C zei] [TP hij ti tv]]?
what said he
Naturally, the ungrammaticality of the primed examples in (267) may be solved by
assuming that sluicing affects the sequence C + TP, but this assumption is less
desirable given that deletion and pronominalization normally involve maximal
 Argument and complementive clauses 737

°projections. If we want to stick to this standard assumption, the analyses in (264)

require additional stipulations to be made; see cf. Merchant (2001:281ff.).

3. The overt part sometimes includes TP-internal material

The third problem is in a sense the reverse of the second one: if sluicing involves
deletion or pronominalization of the TP projection, we wrongly predict that TP-
internal material will never surface. A first case that proves that this is wrong has to
do with multiple questions. Example (268a) shows that, like in English, Dutch
multiple questions allow at most one wh-phrase in the CP projection, which predicts
that fragment wh-questions also consist of at most one wh-phrase. The (b)-examples
in (268) show, however, that the presumed TP-internal wh-phrase in multiple wh-
questions must be overtly expressed in fragment questions: leaving it out leads to
unacceptability; we refer the reader to Merchant (2001:285ff.) for a more detailed
discussion of this issue.
(268) a. Ik weet [CP wiei C [TP ti gisteren wat las]].
I know who yesterday what read
‘I know who read what yesterday.’
b. Iedereen las gisteren iets maar ik weet niet wie wat.
everyone read yesterday something but I know not who what
b. *Iedereen las gisteren iets maar ik weet niet wie.
everyone read yesterday something but I know not who
It may be relevant in this connection that although multiple questions can be
straightforward main clauses, non-embedded multiple fragment wh-questions are
very marked. This is illustrated in the examples in (269).
(269) a. Wie heeft wat gelezen?
who has what read
‘Who read what?’
b. A: Iedereen heeft iets gelezen. B: *?Wie wat?
A: everyone has something read B: who what
‘Everyone has read something. Who what?’
A second case, not mentioned in Merchant (2001/2006), involves constructions
with °floating quantifiers. The examples in (270) show that although the Dutch
floating quantifiers nog meer ‘else’ and allemaal ‘all’ must appear TP-internally in
embedded interrogative clauses, they nevertheless seem to survive sluicing. Note
that the problem does not occur in English given that the quantifiers may be
adjacent to the wh-phrase in regular wh-questions; cf. Merchant (2006:122).
(270) a. Ik ben vergeten [CP wie <*nog meer> C [TP er <nog meer> waren]].
I am forgotten who else there were
‘Iʼve forgotten who else were there.’
a. Jan was er, maar ik ben vergeten wie nog meer.
Jan was there but I am forgotten who else
‘Jan was there but Iʼve forgotten who else.’
738 Syntax of Dutch: Verbs and verb phrases

b. Ik ben vergeten [CP wie <*allemaal> C [TP er <allemaal> waren]].

I am forgotten who all there were
‘Iʼve forgotten who all were there.’
b. Er waren veel mensen, maar ik ben vergeten wie allemaal.
there were many people but I am forgotten who all
‘There were many people but Iʼve forgotten who all.’
The primed (a)- and (b)-examples in (271) show that we find the same facts in main
clauses. In fact, the primed (c)-examples seem to show that it is even possible in
such cases to construct fragment wh-questions that contain adverbial-like material.
(271) a. Wie <*nog meer> waren er <nog meer>?
who else were there
a. A: Jan was er. B: Leuk! Wie nog meer?
A: Jan was there B: nice who else
b. Wie <*allemaal> waren er <allemaal>?
who all were there
b. A: Er waren veel mensen. B: Leuk! wie allemaal?
A: there were many people B: nice who all
c. Wie <*dan> heeft hij <dan> uitgenodigd?
who then has he prt.-invited
‘Who did he invite then?’
c. A: Jan heeft een speciale gast uitgenodigd. B: O, wie dan?
A. Jan has a special guest prt.-invited B: o who then
‘Jan has invited a special guest. O, who then?’

4. Sluicing is not island-sensitive

We conclude our list of potential problems with the fact that has received most
attention in the literature, namely, that sluicing is not °island-sensitive. In short, the
problem is that there are fragment wh-questions for which it is not immediately
clear that they can be derived by means of wh-movement followed by TP deletion,
because wh-movement is blocked in the corresponding non-reduced wh-questions.
First consider the examples in (272a&b), which show that relative clauses are
°islands for wh-extraction. If non-embedded fragment wh-questions are derived by
deletion of the TP of the matrix clause, we expect that (272c) could not be used to
enquire more closely as to the nature of the thing stolen, but this is clearly wrong as
this fragment wh-question would be an entirely natural response to the assertion in
(272a).
(272) a. Jan ontmoette iemand [REL-CLAUSE die iets gestolen had].
Jan met somebody who something stolen had
‘Jan met someone who had stolen something.’
b. *Wati ontmoette Jan iemand [REL-CLAUSE die ti gestolen had]?
what met Jan somebody who stolen had
c. Wat?
what
 Argument and complementive clauses 739

The examples in (273a&b) illustrate the so-called °coordinate structure constraint,

according to which wh-extraction from a coordinate structure is impossible. If non-
embedded fragment wh-questions were derived by deletion of the TP of the matrix
clause, we would expect that (273c) could not be used to ask who the second person
involved was, but again this is clearly wrong as this fragment wh-question would be
a natural response to the statement in (273a).
(273) a. Zij heeft gisteren [[Peter] en [nog iemand anders]] ontmoet.
she has yesterday Peter and yet someone else met
‘She met Peter and one other person yesterday.’
b. *Wiei heeft zij gisteren [[Peter] and [ti]] ontmoet?
who has she yesterday Peter and met
c. Wie?
who
The examples in (274) illustrate the so-called wh-island constraint, according to
which wh-extraction from an embedded interrogative clause is impossible. We see
again that fragment wh-questions are not sensitive to this type of island as the
fragment wh-question in (274c) would again be an entirely natural response to the
sentence in (274a).
(274) a. Marie weet [wie iets gestolen heeft].
Marie knows who something stolen has
‘Marie knows who has stolen something.’
b. *Wati weet Marie [wie ti gestolen heeft]?
what knows Marie who stolen has
‘Marie knows who has stolen what?’
c. Wat?
what
The examples in (275) show that while wh-movement from adverbial °adjunct
clauses is prohibited, fragment wh-questions are not sensitive to it: the fragment
wh-question (275c) is a completely natural response to what is asserted in (275a).
(275) a. Marie is boos op Jan [ADJUNCT omdat hij iets gestolen heeft].
Marie is angry at Jan because he something stolen has
‘Mary is angry at Jan because he has stolen something.’
b. *Wati is Marie boos op Jan [ADJUNCT omdat hij ti gestolen heeft].
what is Marie angry at Jan because he stolen has
c. Wat?
what
Examples (272) to (275) make it patently clear that fragment wh-questions are not
sensitive to islands for wh-extraction. Although Merchant (2001/2006) mentions
many more cases, we will add one slightly more problematic example of potential
island-insensitivity, which involves extraction of attributive modifiers from noun
phrases. Although such extractions are normally not possible in Dutch wh-
questions, fragment wh-questions consisting of APs that correlate with an
attributive modifiers in their antecedent clause are normally judged acceptable by
740 Syntax of Dutch: Verbs and verb phrases

Dutch speakers; they generally prefer Merchant’s example in (276b) to the one in
(276b), in which the full noun phrase is pied-piped (and the noun man may be
omitted as the result of N-ellipsis).
(276) a. Zij hebben een lang-e man aangesteld, maar ik weet niet ...
they have a tall-AGR man hired but I know not
‘Theyʼve hired a tall man, but I donʼt know ...’
b. ... hoe lang/*lang-e.
how tall/tall-AGR
b. ??... een hoe lange (man).
a how long man
A potential problem for the wh-movement + TP deletion approach is, however, that
the extracted adjective, which is supposed to have an attributive function, does not
exhibit the expected attributive -e inflection. Moreover, some of our informants
indicate that even the use of the non-inflected adjective in (276b) is marked (just
like the German speakers consulted by Merchant). It is therefore not entirely clear
whether it is fully justified to use examples such as (276b) as an illustration of the
island-insensitivity of sluicing.
To conclude our discussion of the island-insensitivity of sluicing, we want to
note that Merchant found one case in which Dutch fragment wh-questions seem to
be island-sensitive: fragment wh-questions do obey the language-specific constraint
on preposition stranding. First of all, the examples in (277b) show that wh-
movement of wh-phrases from PPs is normally impossible in Dutch. Preposition
stranding is only possible (and actually preferred) if we are dealing with a
pronominal PP, that is, a PP consisting of an R-word and a preposition, like
waarnaar ‘to what’ in (277c); we refer the reader to P5 for a detailed discussion of
this.
(277) a. Jan luistert graag naar Peter/de radio.
Jan listens gladly to Peter/the radio
‘Jan likes to listen to Peter/the radio.’
b. *Wiei luister Jan graag naar ti? b. [Naar wie]i luistert Jan graag ti?
who listens Jan gladly to to whom listens Jan gladly
(?)
c. Waar luistert Jan graag naar ti? c. Waarnaari luistert Jan graag ti?
where listens Jan gladly to where-to listens Jan gladly
‘What does Jan like to listen to?’ ‘What does Jan like to listen to?’
If fragment wh-questions are not island-sensitive, we would expect that none of the
sluiced counterparts of the questions in (277) need to include the preposition. The
examples in (278) show, however, that the preposition is preferably expressed if the
question word is a pronoun, and (perhaps even more surprisingly) obligatory if the
question word is an R-word. We refer the reader to Subsection E for the discussion
of one notable exception to the generalization that the wh-remnant preferably
includes the preposition.
 Argument and complementive clauses 741

(278) a. Jan luistert naar iemand, maar ik weet niet ?(naar) wie.
Jan listens to someone but I know not to whom
‘Jan is listening to someone, but I donʼt know who.’
b. Jan luistert ergens naar, maar ik weet niet waar*(naar).
Jan listens somewhere to but I know not where-to
‘Jan is listening to something, but I donʼt know what.’
Other cases of apparent island-sensitivity are provided in (279) and involve
adverbial degree modification. First, consider the (a)-examples, which show that
degree modifiers like hoe ‘how’ must pied-pipe the adjective kwaad in regular wh-
questions. The fact that the adjective kwaad cannot be omitted in the corresponding
fragment question in the primed example is of course surprising if fragment wh-
questions are not island-sensitive. The (b)-examples provide somewhat more
complex cases in which the element hoe ‘how’ is part of the more elaborate degree
modifier hoe zwaar ‘how very’, which can itself be extracted from the adjective hoe
zwaar behaard “how very hairy’.
(279) a. Hoe <kwaad> is Jan <*kwaad>?
how angry is Jan
a. Jan is kwaad, maar ik weet niet hoe *(kwaad).
Jan is angry but I know not how angry
b. Hoe <zwaar> is Jan <*zwaar> behaard?
how very is Jan hairy
‘How hirsute is Jan?’
b. Jan is zwaar behaard, maar ik weet niet hoe *(zwaar).
Jan is very hairy but I know not how severely

C. The antecedent clause need not be isomorphic to the phonetically empty TP

On the assumption that a fragment wh-question contains a phonetically empty TP,
we may expect that the empty TP would be syntactically/semantically similar to the
TP of the antecedent clause: deletion normally applies under syntactic identity, and
pro-forms receive an interpretation on the basis of some phrase in the preceding
discourse. This expectation is not borne out, however.

1. No syntactic isomorphism
Dutch is a very suitable language for illustrating that the phonetically empty TP is
not syntactically isomorphic to the TP of its antecedent clause because of the °verb-
second phenomenon found in main clauses: whereas finite verbs are clause-final
(=TP-internal) in embedded clauses, they are in second position in interrogative
main clauses (which is standardly assumed to be the C-position). Consequently, if
the phonetically empty TP in a fragment wh-question must have the same syntactic
structure as the TP of the antecedent clause, we expect that embedded fragment wh-
questions can only take an embedded clause as their antecedent clause, whereas
independent fragment wh-questions can only take a main clause as their antecedent
clause. The examples in (280) show that this expectation does not come true: the
main clause Er is iemand in de kamer ‘there is someone in the room’ in the (a)-
examples can be the antecedent of both independent and embedded fragment wh-
742 Syntax of Dutch: Verbs and verb phrases

questions, and the same thing holds for the embedded clause dat er iemand in de
kamer is ‘that there is someone in the room’ in the (b)-examples.
(280) a. A. Er is iemand in de kamer. B. Wie?
A. there is someone in the room B. who
a. A. Er is iemand in de kamer. B. Weet je ook wie?
A. there is someone in the room B. know you also who
‘There is someone in the room. Do you know who?’
b. A. Ik zie dat er iemand in de kamer is. B. Wie?
A. I see that there someone in the room is B. who
‘I see that there is someone in the room. Who?’
b. A. Ik zie dat er iemand in de kamer is. B. Kan je ook zien wie?
A. I see that there someone in the room is B. can you also see who
‘I see that there is someone in the room. Can you see who?’

2. No semantic isomorphism
The previous subsection has shown that there is no syntactic isomorphism between
the fragment wh-question and the antecedent clause. In fact, example (281a) reveals
that is not even required that the two have an isomorphic semantic representation;
the phonetically empty TP in the fragment wh-question is not interpreted in such a
way that it contains the modal willen ‘to want’ that we find in the antecedent
clause—the interpretation rather involves a modal meaning normally expressed by
kunnen ‘can’ or moeten ‘must’. A similar example can be found in (282a).
(281) a. Ik wil de fiets wel repareren maar dan moet je me vertellen hoe.
I want the bike prt repair but then must you me tell how
‘Iʼm willing to repair the bike, but then you have to tell me how.’
b. hoe ≠ hoe ik de fiets wel wil repareren ‘how I am willing to repair the bike’
b. hoe = hoe ik de fiets kan/moet repareren ‘how I can/should repair the bike’
(282) a. Ik zou je graag helpen, maar ik weet niet hoe.
I would you gladly help but I know not how
‘Iʼd like to help you, but I donʼt know how.’
b. hoe ≠ hoe ik je graag zou helpen ‘how I would like to help you’
b. hoe = hoe ik je kan helpen ‘how I can help you’
An example of a slightly more complex nature is (283). In reply to a pupil’s remark
in (283a), a teacher may react by saying (283b), in which it is clear that the elided
part cannot be isomorphous to what the pupil said given that the anaphor mezelf
cannot be bound by the interrogative pronoun wie: cf. Wie heeft zichzelf/*mezelf
nog niet opgegeven ‘who did not yet enroll?’.
(283) a. Mijnheer, ik heb mezelf nog niet opgegeven voor deze cursus.
master I have REFL yet not enrolled for this course
‘Master, I havenʼt enrolled yet for this course.’
b. Zo, ik vraag me af wie nog meer niet.
well, I wonder REFL prt. who yet more not
‘Well, I wonder who else (did not yet enroll).’
b. wie = wie zichzelf heeft opgegeven
 Argument and complementive clauses 743

The fact that semantic isomorphism need not hold in full does not mean that
anything goes, because the propositional content of the fragment wh-question is still
dependent on the propositional content that we find in the antecedent clause. This
can be illustrated by means of example (284), which shows that minimally the
proposition expressed by the main verb of the antecedent clause and its arguments
must be preserved in the interpretation of the fragment wh-question. See Merchant
(2006:ch.1) for an attempt to formally define this notion of “closeness in meaning”.
(284) a. Marie noemt veel mensen stom, maar ik weet niet precies wie.
Marie calls many people stupid but I know not precisely who
‘Marie calls many people dumb, but I donʼt know precisely who.’
b. wie = wie ze stom noemt ‘who she calls dumb’
c. wie ≠ wie ze beledigt ‘who she is insulting’

D. The correlate of the wh-phrase in the antecedent clause

The fact established in the previous subsection that the phonetically empty TP need
not be syntactically isomorphic to the TP of the antecedent clause could also have
been demonstrated on the basis of the position of the wh-phrase of the fragment wh-
question and its correlate in the antecedent clause. Again assume the wh-movement
+ TP deletion approach to fragment wh-questions. What examples such as (285)
show then is that the antecedent clause differs from the phonetically empty in the
fragment wh-question in that it has a noun phrase where the TP has a °trace.
(285) Ik geloof [CP dat [TP Jan iets leuks gelezen heeft]], maar ...
I believe that Jan something nice read has but ...
... ik weet niet [CP wati C [TP Jan ti gelezen heeft]].
... I know not what Jan read has
Actually, example (286) shows that it is even possible to use the verb lezen ‘to read’
pseudo-intransitively, and nevertheless to have a fragment wh-question with a wh-
phrase that functions as the object of lezen; the absence of a(n overt) direct object is
indicated by the use of Ø.
(286) Ik geloof [CP dat [TP Jan Ø gelezen heeft]], maar ...
I believe that Jan read has but
... ik weet niet [CP wati C [TP Jan ti gelezen heeft]].
... I know not what Jan read has
When one analyzes pseudo-intransitive constructions as constructions without a
direct object (alternatively, one may argue in favor of some covert object), the trace
of the wh-phrase in the fragment wh-question would not have a correlate at all in
the antecedent clause. This may in fact be the normal situation in fragment wh-
questions such as (287) which consist of a wh-phrase with an adverbial function, as
it is not normally assumed that such adverbial phrases are covertly expressed in
sentences in which they are not morphologically visible.
744 Syntax of Dutch: Verbs and verb phrases

(287) a. Jan is vertrokken, maar ik weet niet wanneer.

Jan is left but I know not when
‘Jan has left, but I donʼt know when.’
b. Ik ben mijn sleutels verloren, maar ik weet niet waar.
I am my keys lost but I know not where
‘Iʼve lost my keys, but I donʼt know where.’
c. Ik wil mijn fiets repareren maar ik weet niet hoe.
I want my bike repair but I know not how
‘I want to repair my bike, but I donʼt know how.’
Example (288a) shows that if an argument wh-trace does have a correlate in the
antecedent clause, the latter must be indefinite. This is probably a semantic
restriction: the use of a definite noun phrase would make the fragment wh-question
contradictory or superfluous as in, respectively, (288a) and (288b).
(288) a. Jan heeft Max Havelaar van Multatuli gelezen, ($maar ik weet niet wat).
Jan has Max Havelaar by Multatuli read but I know not what
‘Jan has read Max Havelaar by Multatuli, but I donʼt know what.’
b. A. Jan heeft Max Havelaar van M. gelezen, B. $Weet je ook wat?
A. Jan has Max Havelaar by M. read B. know you also what
‘Jan has read Max Havelaar by Multatuli. Do you also know what?’
The (a)-examples in (289) suggest that something similar applies to adverbial wh-
phrases; these cases are only acceptable if the wh-phrase is used to indicate that the
speaker cannot determine the time/location more precisely. Similarly, the
independent question in the (b)-examples is used to solicit more precise information
about the relevant time internal/location.
(289) a. Jan is onlangs vertrokken, maar ik weet niet ?(precies) wanneer.
Jan is recently left but I know not precisely when
‘Jan left recently, but I donʼt know precisely when.’
a. A. Jan is onlangs vertrokken. B. Wanneer (precies)?
A. Jan is recently left B. when precisely
‘Jan left recently. When precisely?’
b. Ik ben mijn sleutels thuis verloren, maar ik weet niet precies waar.
I am my keys at.home lost but I know not precisely where
‘Iʼve lost my keys at home, but I donʼt know precisely where.’
b. A. Ik ben mijn sleutels thuis verloren. B. Waar (precies)?
A. I am my keys at.home lost B. where precisely
‘Iʼve lost my keys at home. Where precisely?’
Universally quantified phrases are similarly excluded as correlates of wh-phrases in
fragment wh-questions. This can again be seen as a semantic restriction: if all
relevant entities in the given domain of discourse were to be included, the fragment
wh-question would become contradictory or superfluous, as illustrated in,
respectively, (290a) and (290b).
 Argument and complementive clauses 745

(290) a. Jan heeft alle romans van Boon gelezen, ($maar ik weet niet wat/welke).
Jan has all novels by Boon read but I know not what/which
‘Jan has read all the novels by Boon, but I donʼt know what/which.’
b. A. Jan heeft alle romans van Boon gelezen. B. $Wat/Welke?
A. Jan has all novels by Boon read B. what/which
‘Jan has read all the novels by Boon. Do you know what/which?’
There is, however, one exception: example (291a) shows that universally quantified
phrases are possible as the correlate of the first wh-phrase in multiple fragment wh-
questions. By means of examples like these the speaker expresses that he is not able
to supply the reader with a paired list of persons and things <x,y> such that person x
bought thing y. It is important to observe that the correlates of the wh-phrases in
such multiple fragment wh-questions must be clause mates; this condition is
satisfied in example (291a), but not in (291b), and as a result the multiple fragment
wh-question is unacceptable in the latter case as a result.
(291) a. Iedereen had iets gekocht maar ik weet niet wie wat.
everyone has something bought but I know not who what
‘Everyone had bought something, but I donʼt know who [bought] what.’
b. Iedereen zei dat Jan iets las, (*maar ik weet niet wie wat).
everyone said that Jan something read but I know not who what
‘Everyone said that Jan was reading something (but ...).’
Example (292a) is an apparent counterexample to this clause-mate condition: the
fragment wh-question can only be used if the subject pronoun in the object clause of
the antecedent clause is interpreted as a °bound variable, that is, as referentially
dependent on the quantified expression iedereen; the fact that the second correlate
is a clause mate of the bound variable is apparently enough to satisfy the clause-
mate condition. Example (292b) provides a comparable case in which the
phonetically empty pronoun PRO of the infinitival clause functions as a variable
bound by the universally quantified pronoun iedereen in the matrix clause.
(292) a. Iedereeni zei dat hiji iets las, maar ik weet niet wie wat.
everyone said that he something read but I know not who what
‘Everyone said that he was reading something (but I donʼt know who what).’
b. Iedereeni beloofde [PROi iets te lezen], maar ik weet niet wie wat.
everyone promised something to read but I know not who what
‘Everyone promised to read something (but I donʼt know who what).’

E. Sluicing-like constituents
The sluicing constructions discussed in the previous subsections all occur as
independent sentences mostly given in conjunction with a sentence containing the
correlates of the wh-phrase. We want to conclude our discussion by noting that
sluicing-like constructions can also be used as constituents of clauses and smaller
word groups; cf. Lakoff (1974). The examples in (293), adapted from Van
Riemsdijk (2000) and especially Kluck (2011), show that the sluicing-like phrase,
given in square brackets, can be used as an argument (subject/object), a
746 Syntax of Dutch: Verbs and verb phrases

°complementive, an adverbial phrase, and even as a part of a quantifier or an

attributive modifier of a noun phrase.
(293) a. [Je weet wel wie] was hier. [subject]
you know AFF who was here
‘You know who was here.’
b. Jan heeft [je raadt nooit wat] gelezen. [direct object]
Jan has you guess never what read
‘Jan has read youʼll never guess what.’
c. Jan stuurt Marie altijd [ik weet niet waar naartoe]. [complementive]
Jan sends Marie always I know not where to
‘Jan always sends Marie I never know where.’
d. Marie heeft [je raadt wel waar] geklaagd. [adverbial phrase]
Marie has you guess AFF where complained
‘Marie has complained you can guess where.’
e. Marie heeft [ik weet niet/God weet hoeveel] boeken. [quantifier]
Marie has I know not/God knows how.many books
‘Marie has I donʼt know/God knows how many books.’
f. Jan heeft een [je wil niet weten hoe grote] televisie. [attributive mod.]
Jan has a you want not know how big television
‘Jan has gotten an immensely big television.’
The matrix clauses of such sluicing-like phrases often consist of more or less fixed
collocations; they are often headed by verbs like weten ‘to know’ or raden ‘to
guess’ supplemented by the negative/affirmative markers niet/wel or a negative
adverb like nooit ‘never’. Example (293e-f) shows that the matrix clause can even
be a completely idiomatic phrase like God weet ‘God knows’ + wh-phrase or je wil
niet weten ‘you donʼt wanna know’ + wh-phrase; see Kluck (2010).
The bracketed phrases in (293) all have main clause word order, with the finite
verb in second and the subject in first position. Although this suggests that we
cannot be dealing with regular embedding, the phrases do not have the distribution
of clauses either but occupy the same position as the non-clausal elements with the
same syntactic function in (294).
(294) a. Peter was hier. [subject]
Peter was here
b. Jan heeft je dissertatie gelezen. [direct object]
Jan has your dissertation read
c. Jan stuurt Marie altijd naar Groningen. [complementive]
Jan sends Marie always to Groningen
d. Marie heeft bij de directie geklaagd. [adverbial phrase]
Marie has with the management complained
e. Marie heeft veel boeken. [quantifier]
Marie has many books
f. Jan heeft een grote televisie. [attributive modifier]
Jan has a big television
 Argument and complementive clauses 747

This paradox is solved in Kluck (2011) by assuming that the sluicing-like phrases
are actually parenthetical clauses; Examples like (293a-b) have a similar structure
as the examples in (295), the only difference is that the correlates of the wh-phrases
in the parenthetical clauses, iemand ‘someone’ and iets ‘something’, are not overtly
expressed in (293a&b). Her proposal implies that for some of the cases in (293),
there are only phonetically empty correlates.
(295) a. Iemand — je weet wel wie — was hier. [subject]
someone you know AFF who was here
‘You know who was here.’
b. Jan heeft iets — je raadt nooit wat — gelezen. [direct object]
Jan has something you guess never what read
‘Jan has read youʼll never guess what.’
An argument in favor of analyzing the bracketed phrases as sluicing
constructions can be built on cases in which the sluice is a prepositional object.
Subsection B has shown that in such cases the wh-remnant preferably includes the
preposition. The (b)-examples show that we seem to find the same preference in the
case of the constituents under discussion (albeit that our informants seem more
lenient towards (296b)).
(296) a. Jan roddelt over iemand, maar ik weet niet ?(over) wie.
Jan gossips about someone but I know not to whom
‘Jan is gossiping about someone, but I donʼt know who.’
b. Jan heeft [je weet wel over wie] geroddeld.
Jan has you know PRT about who gossiped
‘Jan has gossiped about you know who.’
b. (?)Jan heeft over [je weet wel wie] geroddeld.
Jan has about you know PRT who gossiped
‘Jan has gossiped about you know who.’
More evidence is provided in Kluck (2011:202), who observes that the wh-remnant
preferably does not include the preposition in examples such as (297a), in which the
form (op) wat is used instead of the more common pronominal PP form waarop.
This exceptional behavior is also reflected in the (b)-examples: the bracketed phrase
preferably does not include the preposition op but is itself the complement of op.
(297) a. Jan rekent ergens op, maar ik weet niet ( ?op) WAT.
Jan counts something on but I know not on what
‘Jan is counting on something but I donʼt know what.’
b. Jan heeft op [ik weet niet wat] gerekend.
Jan has on I know not what counted
‘Jan has counted on I not know what.’
b. ??Jan heeft [ik weet niet op wat] gerekend.
Jan has I know not on what counted
‘Jan has counted on I donʼt know what.’
For completeness’ sake, the examples in (298) show that sluicing also allows the
more regular form waarop. Given that the preposition is obligatory in (298a), we
748 Syntax of Dutch: Verbs and verb phrases

correctly expect the bracketed phrase in the (b)-examples to obligatorily include the
preposition.
(298) a. Jan rekent ergens op, maar ik weet niet waar *(op).
Jan counts something on but I know not where on
‘Jan is counting on something but I donʼt know what.’
b. Jan heeft [ik weet niet waarop] gerekend.
Jan has I know not where-on counted
‘Jan has been counting on I donʼt know what.’
b. *Jan heeft op [ik weet niet waar] gerekend.
Jan has on I know not where counted
The fact that the bracketed phrases in the (b)-examples in (296) to (298) exhibit a
similar behavior as the unequivocal sluicing constructions in the (a)-examples
strongly supports a sluicing analysis of the former. For more evidence in favor of
this conclusion, we refer the reader to Kluck (2011:ch.5).
A construction that seems to belong to the same domain is given in (299); the
construction resembles the regular sluicing construction in that we may add
°floating quantifiers like allemaal to the wh-phrase: compare examples like Jan
heeft weet ik wat allemaal gelezen and Jan stuurt Marie altijd weet ik waar
allemaal naartoe with the relevant examples in Subsection IB3. The examples in
(299) seem to have a similar meaning as the corresponding examples in (293), but
are structurally completely different: although the bracketed constituent does have
the order of a main clause, the finite verb and the subject are inverted. The
construction seems more restricted than the construction type in (293) in the sense
that the verb is typically weten ‘to know’, and seems to express some form of high
degree quantification. To our knowledge, this construction has not been discussed
in the literature so far.
(299) a. Jan heeft [weet ik wat] gelezen.
Jan has know I what read
‘Jan has read all kinds of stuff.’
b. Jan stuurt Marie altijd [weet ik waar naartoe].
Jan sends Marie always know I where to
‘Jan is always sending Marie I never know where.’
c. Jan heeft [weet ik waar] gestudeerd.
Jan has know I where studied
‘Jan has studied at all kinds of places.’
d. Jan heeft [weet ik hoeveel] boeken.
Jan has know I how.many books
‘Jan owns a tremendous number of books.’

F. Conclusion
The previous subsections have looked in some detail at fragment wh-questions.
Subsection A has shown that these fragment wh-questions exhibit the behavior of
clauses and so cannot be seen as projections of a non-verbal nature. Subsection B
investigated the internal structure of fragment wh-phrases in more detail: we have
seen that fragment wh-questions do not overtly express the °head of the CP-
 Argument and complementive clauses 749

projection (they never contain a complementizer or a finite verb), do not contain

any TP-internal material (although there are some potential exceptions to this
claim), and are not island-sensitive (with, again, a number of potential exceptions).
Subsection C discussed the relation between the supposedly elided TP and its
antecedent clause and showed that, although the two share the same core
proposition, they need not be identical in syntactic structure. Subsection D
discussed the relation between the wh-phrases in fragment wh-questions and their
non-wh-correlates in the antecedent clause, and later showed that the latter cannot
normally be definite or universally quantified (with the notable exception of the
correlate of the first wh-phrase in multiple fragment wh-questions). We concluded
in subsection E with a brief remark on sluicing-like constructions that are used as
constituents with a non-clausal behavior. Much of what we presented here was
based on Merchant (2001/2006), to which we refer the reader for a more detailed
discussion as well as a critical review of a variety of theoretical approaches to
sluicing.

II. Fragment answers

This subsection discusses the second type of fragment clauses, which we will refer
to as fragment answers. The examples in (300) show that fragment answers are used
in response to wh-questions and can occur either as independent utterances or as
dependent CONSTITUENTS. The overt part of the fragment answer correlates with the
wh-phrase in the antecedent clause (that is, the wh-question). Observe that fragment
answers provide new information by definition and are therefore normally assigned
sentence accent, which is indicated by a grave accent on the book title De
zondvloed in the (b)-examples of (300).
(300) a. Wat is Jan momenteel aan het lezen?
what is Jan now AAN HET lezen
‘What is Jan reading at the moment?’
b. De zòndvloed van Jeroen Brouwers. [independent]
De zondvloed by Jeroen Brouwers
b. Ik vermoed De zòndvloed van Jeroen Brouwers. [dependent]
I suppose De zondvloed by Jeroen Brouwers
The list in (301) gives a small sample of verbs that may take such fragment answers
as their complement; these are all verbs that may take a declarative clause as their
complement.
(301) Verbs that may take a fragment answer: denken ‘to think/believe’, hopen ‘to
hope’, vermoeden ‘to suppose’, vertellen ‘to tell’, vrezen ‘to fear’, zeggen ‘to
say’
Verbs taking a fragment answer as their complement are always non-factive; see
Barbiers (2000:194). This is illustrated in example (302b): whereas the non-factive
verb vrezen ‘to fear’ gives rise to a fully acceptable result, the factive verb
betreuren does not.
750 Syntax of Dutch: Verbs and verb phrases

(302) a. Wat koopt Marie voor Peter?

what buys Marie for Peter
‘What will Marie buy for Peter?’
b. Ik vrees/*betreur een drumstel.
I fear/regret a drum.set
The (visible) constituent in the fragment answer can be a nominal argument, like in
the two earlier examples, but it can also be of a different category and have a
different function. In example (303b), for instance, we are dealing with a temporal
adverbial phrase, which can appear in the form of an AP like vroeg ‘early’ or a PP
like in de ochtend ‘in the morning’
(303) a. Wanneer vertrek je morgen?
when leave you tomorrow
‘When will you leave tomorrow?’
b. Ik geloof vroeg/in de ochtend.
I believe early/in the morning
The following two subsections will argue that fragment answers are clauses and
suggest a potential analysis for them, which, like in the case of fragment wh-
phrases, raises a number of non-trivial questions.

A. Fragment answers are clauses

That fragment answers are clausal in nature can be established on the basis of their
syntactic distribution, even though we will see that the argument is not as
straightforward as in the case of fragment wh-questions discussed in Subsection I.
The basic insight is the following: if fragment answers are indeed clauses, we
predict that they normally follow the verbs in clause-final position and that they are
excluded in the °middle field of the clause; if fragment answers are not clauses but
phrases of some other category, we would predict that they must occur in front of
the verbs in clause-final position if the phrase constituting the fragment answer is
nominal in nature. Testing these predictions is not easy given that dependent
fragment answers do not readily occur in embedded clauses or clauses including a
non-main verb. Nevertheless, most speakers feel the contrast between the two
examples in (304b&c): whereas (304b) is generally judged as marked but
acceptable, example (304c) is generally considered to be degraded.
(304) a. Wat geeft Marie Peter voor zijn verjaardag?
what gives Marie Peter for his birthday
‘What will Marie give Peter for his birthday?’
b. (?)Ik weet het niet zeker, maar ik heb steeds vermoed een boek.
I know it not for.sure but I have all.the.time supposed a book
‘Iʼm not absolutely sure but my suspicion has been all along: a book.’
c. *?Ik weet het niet zeker, maar ik heb steeds een boek vermoed.
I know it not for.sure but I have all.the.time a book supposed
The contrast between the (b)- and (c)-examples is perhaps clearer when we replace
the verb vermoeden by a verb of saying/thinking, as in (305). Example (305b) is
 Argument and complementive clauses 751

generally judged as acceptable, whereas judgments on (305c) vary from very

marked to unacceptable.
(305) a. Wat geeft Marie Peter voor zijn verjaardag?
what gives Marie Peter for his birthday
‘What will Marie give Peter get for his birthday?’
b. Ik weet het niet zeker, maar Marie heeft steeds gezegd een boek.
I know it not for.sure but Marie has all.the.time said a book
‘Iʼm not absolutely sure but Marie has always said: a book.’
c. *?Ik weet het niet zeker, maar Marie heeft steeds een boek gezegd.
I know it not for.sure but Marie has all.the.time a book said
Given the subtlety of the judgments, we have also asked our informants to evaluate
examples involving manner adverbials, which like nominal arguments generally
precede the verbs in clause-final position. The net result is the same: the contrast
between the two (b)-examples in (306) shows again that fragment answers must
follow the verbs in clause-final position.
(306) a. Hoe heeft Peter dat boek gelezen: globaal of nauwkeurig?
how has Peter that book read globally or meticulously
‘How did Peter read that book: cursorily or thoroughly?’
b. Ik weet het niet zeker, maar ik zou zeggen globaal.
I know it not for.sure but I would say globally
‘Iʼm not absolutely sure, but Iʼd say: cursorily.’
c. *?Ik weet het niet zeker, maar ik zou globaal zeggen.
I know it not for.sure but I would cursorily say
The contrasts between the (b)- and (c)-examples strongly suggest that fragment
answers are clausal in nature. Additional evidence is provided by examples such as
(307), where the wh-phrase in (307a) pertains to a contextually determined set of
options: a novel, a collection of stories, a volume of poems, etc. The fact that the
neutral demonstrative pronoun dat is used in (307b) suggests that the fragment
answer is not clausal: the neutral pronoun can refer to clauses but not to non-neuter
noun phrases.
(307) a. Wat ga je morgen lezen?
what go you tomorrow read
‘What are you going to read tomorrow?’
b. Ik denk de roman, want dat is het gemakkelijkst.
I think the novel because that is the easiest
However, before we can confidently adopt the claim that fragments answers are
clauses, we have to discuss two complications. The first is that verbs of
saying/thinking may also be used in (semi-)direct reported speech constructions; see
Section 5.1.2.4, sub II. Before we can draw any conclusion from the (b)-examples
in (305) and (306), we have to establish that we are in fact dealing with fragment
answers, and not with (semi-)direct quotes. A first argument in favor of the first
option is provided by the meaning of example (305b): it does not express that Marie
literally said “Een boek”, but that Marie has said various things from which the
752 Syntax of Dutch: Verbs and verb phrases

speaker has drawn the conclusion that she would give Peter a book. The same thing
is even clearer for (306b), in which the speaker does not quote himself but provides
an opinion. A second argument can be based on example (308b) below. The fact
that the pronoun zij ‘she’ can be used to refer to Marie and the pronoun ik ‘I’ must
refer to the speaker of this sentence shows that we cannot be dealing with a direct
quote. The fact established in Section 5.1.2.4, sub II, that the choice between direct
and semi-direct quotes is normally free (in narratives at least) therefore suggests
that (308b) cannot be interpreted as a semi-direct reported speech construction
either.
(308) a. Wie koopt er een boek voor Peter?
who buys there a book for Peter
‘Who will buy a book for Peter?’
b. Ik weet het niet zeker, maar Marie heeft steeds gezegd zij/ik.
I know it not for.sure but Marie has all.the.time said she/I
‘Iʼm not absolutely sure, but Marie has said all the time: she/I.’
c. *Ik weet het niet zeker, maar Marie heeft steeds zij/ik gezegd.
I know it not for.sure but Marie has all.the.time she/I said
A final argument for claiming that we are dealing with fragment answers, and not
with (semi-)direct quotes, is provided by the examples in (309). If we were dealing
with a reported speech construction, we would expect that we could use any quote
as the fragment answer: this wrongly predicts that (309b) would be a felicitous
answer to the question in (309a).
(309) a. Komt Marie morgen dat boek halen?
comes Marie tomorrow that book fetch
‘Will Marie come to fetch that book tomorrow?’
b. #Marie heeft gezegd ja.
Marie has said yes
The second complication that must be discussed before we adopt the claim that
fragment answers are clausal is that Barbiers (2000:197-8) considers examples such
as (310) fully acceptable, provided that the displaced constituent is marked with
contrastive focus accent. Although these judgments are actually shared by many
(but not all) Dutch speakers, it is not immediately clear whether examples of this
type are relevant for our present discussion; Given the somewhat unclear status of
these examples, we will not discuss them in detail here and refer the reader to
Temmerman (2013) for an attempt to show that the primed examples are indeed
fragment clauses, albeit of a somewhat different sort than fragment clauses that
follow the verbs in clause-final position.
%
(310) a. Ik had MORGENi gedacht [CP dat Jan ti zou komen].
I had tomorrow thought that Jan would come
‘Iʼd thought that Jan would come tomorrow.’
a. %Ik had MORGENi gedacht.
I had tomorrow thought
 Argument and complementive clauses 753
%
b. Ik had in de TUINi gehoopt [CP dat het feest ti zou zijn].
I had in the garden hoped that the party would be
‘Iʼd hoped that the party would be in the garden.’
b. %Ik had in de TUINi gehoopt.
I had in the garden hoped
From the discussion above we can safely conclude that fragment answers are
clausal in nature. More support for this position can be found in the fact that
pronouns may appear in their subject form when they constitute (the visible part of)
a fragment answer; the examples in (311) show that the form of the pronoun is not
determined by the verb denken, but by the grammatical function of its wh-correlate
in the antecedent clause; cf. Barbiers (2000).
(311) a. A. Wie komt er vandaag? B. Ik denk Jan/hij. [subject pronoun]
A. who comes there today B. I think Jan/he
‘Who is coming today? I think Jan/he.’
b. A. Wie heeft hij bezocht? B. Ik denk Marie/haar. [object pronoun]
A. who has he visited B. I think Marie/her
‘Who did he visit? I think Marie/her.’

B. The derivation of fragment clauses

Since fragment wh-questions and fragment answers are both clausal in nature, it
seems natural to assume that the two have a more or less similar derivation.
Subsection I has shown that fragment wh-questions are derived by postulating that
the TP of the fragment clause is deleted or pronominalized; see the (b)-examples in
(312), repeated from Subsection IB, in which strikethrough stands for deletion of
the phonetic content of the TP and e for an empty pro-form replacing the TP.
(312) a. Ik weet niet [CP wati C [TP Jan gekocht ti heeft]].
I know not what Jan bought has
‘I donʼt know what Jan has bought.’
b. Ik weet niet [CP wati C [TP Jan gekocht ti heeft]].
b. Ik weet niet [CP wat C [TP e ]].
It seems that in the case of fragment answers, there is good reason to prefer the
deletion over the pronominalization approach; see also Temmerman (2013). First
consider the examples in (313a&b), which show that reflexive pronouns like
zichzelf ‘himself’ differ from referential pronouns like hem ‘him’: the former must
but the latter cannot have a syntactically realized antecedent in its own clause; see
Section N5.2.1.5 on °binding theory for more detailed discussion. The indices
indicate (lack of) coreference.
(313) a. Ik denk dat Peteri zichzelfi/*hemi het meest bewondert.
I think that Peter himself/him the most admires
‘I think that Peter admires himself the most.’
b. Peteri denkt dat ikj hemi/*zichzelfi het meest bewonder.
Peter thinks that I him/himself the most admire
‘Peter thinks that I admire him the most.’
754 Syntax of Dutch: Verbs and verb phrases

The distribution of the pronouns in the fragment answers in (314) show that these
are dependent on the subject in the antecedent wh-clause. This would follow
immediately under the TP ellipsis approach: although their phonetic content is
erased under TP ellipsis, subjects of fragment answers are nevertheless syntactically
present and can therefore function as antecedents of pronouns; the fact that the
pronouns in (314) have a similar distribution as the pronouns in (313) is therefore
expected. An account of this sort is not available if the TP is replaced by a pro-
form, as this would entirely remove the subject from the fragment question.
(314) a. A. Wie bewondert Jani het meest? B. Ik denk zichzelfi/*hemi.
A. who admires Jan the most B. I think himself/him
‘Who does Jan admire the most? I think himself.’
b. A. Wie bewonder jijj het meest? B. Ik denk hemi/*zichzelfi.
A. who admire you the most B. I think him/himself
‘Who do you admire the most? I think him.’
For completeness’ sake, the examples in (315) provide similar instances with a
°bound variable reading of the possessive pronoun zijn ‘his’; given that the bound
variable reading of pronouns only arises if the quantifier °c-commands the pronoun,
the availability of this reading in the question-answer pair in (315) again supports
the TP-ellipsis approach; cf. Temmerman (2013).
(315) a. Ik denk dat iedereeni zijni moeder het meest bewondert.
I think that everyone his mother the most admires
‘I think that everyone admires his mother the most.’
b. A. Wie bewondert iedereeni het meest? B. Ik denk zijni moeder.
A. who admires everyone the most B. I think his mother
‘Who does everyone admire the most? I think his mother.’

C. Two problems
Adopting a TP-deletion analysis for fragment answers is not wholly
unproblematical: it raises the non-trivial question what structure serves as the input
of the deletion operation. If we adopt a similar analysis as suggested in Subsection
IC, for fragment wh-questions, we should assume that the non-wh-correlate of the
wh-phrase in the antecedent (= zichzelf in (316)) is topicalized before deletion. An
example such as (314a) with zichzelf would then have the syntactic representation
in (316a). The problem of this analysis is, however, that the first position in
embedded clauses is normally only accessible to wh-phrases and relative pronouns;
topicalization of any other material is categorically excluded. This means that the
unacceptable structure in (316b) would be the input for TP deletion in order to
derive the acceptable fragment question in (316a).
(316) a. Ik denk [CP zichzelfi C [TP Jan ti het meest bewondert]].
I think himself Jan the most admires
b. *Ik denk [CP zichzelfi dat/Ø [TP Jan ti het meest bewondert]].
I think himself Jan the most admires
 Argument and complementive clauses 755

For completeness’ sake, the examples in (317) show that this problem does not
occur in independent fragment answers, although these of course raise the question
why the finite verb cannot be overtly expressed; see the discussion of the same
problem for independent fragment questions in Subsection IB.
(317) a. [CP Zichzelfi C [TP Jan ti het meest bewondert]].
himself Jan the most admires
b. [CP Zichzelfi bewondert [TP Jan ti het meest tbewondert]].
himself admires Jan the most
Barbiers (2000) suggested that dependent fragment clauses can be derived from the
structures in the primeless examples in (310), repeated here as (318), by deletion of
the postverbal CPs. This proposal runs into two problems, however: it wrongly
predicts that fragment clauses must precede the verbs in clause-final position, and it
leaves unexplained that fragment answers can also occur as independent utterances.
%
(318) a. Ik had MORGENi gedacht [CP dat Jan ti zou komen].
I had tomorrow thought that Jan would come
b. %Ik had in de TUINi gehoopt [CP dat het feest ti zou zijn].
I had in the garden hoped that the party would be
No further attempts will be made here to provide an answer to the question
pertaining to the derivation of fragment answers, but we refer to Temmerman
(2013) for a number of suggestions of a more theory-internal nature.
Merchant (2004) claims that fragment answers differ from fragment questions
in that the presumed topicalization operation preceding TP-deletion is °island-
insensitive. This is not so easy to demonstrate, however, because wh-movement in
antecedent wh-questions is island-sensitive itself; consequently, fragment answers
will obey the relevant island restrictions more or less by definition. Merchant
therefore demonstrates his claim by means of yes/no-questions of the sort in
(319a&b), which have a focus accent on an embedded constituent and can be seen
as implicit wh-questions; if the answers in the primed examples in (319) can be
analyzed in the same way as run-of-the-mill fragment answers, the unacceptability
of the answers in the primed examples shows that topicalization in fragment
answers is island-sensitive in its own right.
(319) a. Does Abby speak [ISLAND the same Balkan language that BEN speaks]?
a. *No, CHARLIE.
b. Did Ben leave the party [ISLAND because ABBY wouldnʼt dance with him]?
b. *No, BETH.
The status of the answers in the comparable Dutch question-answer pairs in (320) is
somewhat unclear to us, which we have indicated by a percentage sign.
Temmerman (2013) gives these pairs as fully acceptable, but our informants seem
to be less positive about it.
756 Syntax of Dutch: Verbs and verb phrases

(320) a. Zoeken ze [ISLAND iemand die GRIEKS spreekt]?

look.for they someone that Greek speaks
‘Are they looking for someone who can speak Greek?’
a. %Nee, (ik zou denken) ALBANEES.
no I would think Albanian
b. Vertrok Jan [ISLAND omdat MARIE niet met hem wou dansen]?
left Jan because Marie not with him wanted dance
‘Did Jan leave because Marie didnʼt want to dance with him?’
b. %Nee, (ik zou denken) ELS.
no I would think Els
If the answers in the primed examples in (320) are indeed felicitous and if these
answers ought to be analyzed as fragment answers, it would show that Dutch
fragment answers differ from their English counterparts in that they are island-
insensitive (just like fragment questions). For completeness’ sake, we want to point
out that Temmerman claims that the (postverbal) fragment answers in (320) differ
markedly from the (preverbal) fragment answers in (321), which are undisputedly
infelicitous as responses to the questions in the primeless examples in (320).
#
(321) a. Nee, ik zou ALBANEES denken.
no I would Albanian think
b. #Nee, ik zou ELS denken.
no I would Els think

D. Conclusion
The previous subsections have discussed two types of fragment clauses: fragment
wh-questions and fragment answers. It has been shown that fragment clauses have
the distribution of regular finite clauses, which suggests that these fragment clauses
are CPs with a phonetically empty TP. For instance, the fact that the overt part of
fragment answers may consist of a sole reflexive pronoun may favor a TP-deletion
over a TP-pronominalization approach. However, the TP-deletion approach also
raises a number of non-trivial questions concerning the lack of isomorphism
between the presumed empty TP of fragment clauses and the TP of their antecedent
clauses. These questions are discussed at length for fragment questions in Merchant
(2001/2006) and much subsequent work, but they will no doubt remain part of the
research agenda for some time to come.

5.1.6. Wh-extraction from argument clauses

This section discusses wh-extraction from argument clauses. The examples in (322)
show that such extractions can be applied to at least three types of phrases: wh-
phrases, relative pronouns, and contrastively stressed phrases. For convenience, we
will focus on extraction of wh-phrases, and refer the reader to Chapter 11 for a more
extensive discussion of the different kinds of wh-movement.
 Argument and complementive clauses 757

(322) a. Wati denk je [dat Marie morgen ti zal vertellen]?

what think you that Marie tomorrow will tell
‘What do you think that Marie will tell tomorrow?’
b. Hij liep naar de plaats waari hij wist [dat zijn accordeon ti stond].
he walked to the place where he knew that his accordion stood
‘He walked to the place where he knew his accordion was.’
c. Dit BOEKi denk ik [dat Marie ti wel wil hebben].
this book think I that Marie PRT wants.to have
‘This book, I think that Marie would like to have.’

I. Bridge verbs
Argument clauses are special in that they allow wh-extraction under specific
conditions. The examples in (323) show that the extracted wh-phrase may be either
an argument of the embedded verb or an °adjunct. The °traces ti refer to the original
position of the moved wh-phrases in the embedded clauses.
(323) a. Ik denk [CLAUSE dat Marie dit boek morgen zal kopen].
I think that Marie this book tomorrow will buy
‘I think that Marie will buy this book tomorrow.’
b. Wiei denk je [CLAUSE dat ti dit boek morgen zal kopen]? [subject]
who think you that this book tomorrow will buy
‘Who do you think will buy this book tomorrow?’
c. Wati denk je [CLAUSE dat Marie ti morgen zal kopen]? [object]
what think you that Marie tomorrow will buy
‘What do you think that Marie will buy tomorrow?’
d. Wanneeri denk je [CLAUSE dat Marie dit boek ti zal kopen]? [adverbial]
when think you that Marie this book will buy
‘When do you think that Marie will buy this book?’
The fact that wh-phrases can be extracted from argument clauses is rather special as
this is categorically prohibited from adjunct clauses. The examples in (324), for
example, show that conditional clauses are strong (absolute) °islands for wh-
movement; they block wh-extraction of both arguments and adjuncts.
(324) a. Jan zal blij zijn [CLAUSE als Marie dit boek morgen zal kopen].
Jan will happy be if Marie this book tomorrow will buy
‘Jan will be happy if Marie will buy this book tomorrow.’
b. *Wiei zal Jan blij zijn [CLAUSE als ti dit boek morgen zal kopen]?
who will Jan happy be if this book tomorrow will buy
c. *Wati zal Jan blij zijn [CLAUSE als Marie ti morgen zal kopen]?
what will Jan happy be if Marie tomorrow will buy
d. *Wanneeri zal Jan blij zijn [CLAUSE als Marie dit boek ti zal kopen]?
when will Jan happy be if Marie this book will buy
There are good reasons for assuming that the wh-phrases in (323) are not moved in
one fell swoop into their target position but that this involves an additional
movement step via the initial position of the embedded clause; cf. Section 11.3 and
Chomsky (1973/1977). This is indicated in the structures in (325), in which the
758 Syntax of Dutch: Verbs and verb phrases

traces refer to the positions that the moved phrase has occupied during the
derivation; the CP/TP structure of clauses assumed here is discussed in Section 9.1.
(325) a. Wiei denk je [CP ti dat [TP ti dit boek zal kopen]]?
who think you that this book will buy
b. Wati denk je [CP ti dat [TP Marie ti zal kopen]]?
what think you that Marie will buy
c. Wanneeri denk je [CP ti dat [TP Marie dit boek ti zal kopen]]?
when think you that Marie this book will buy
The main reason for assuming that the wh-phrases are moved via the initial position
of the embedded clause is that this immediately accounts for the unacceptability of
examples like (326b&c); since the clause-initial position of the embedded clause is
already occupied by the subject pronoun wie ‘who’, wh-extraction of the
object/adjunct must apply in one fell swoop and this is not allowed. Note that
(326c) is acceptable when the adverb wanneer ‘when’ is construed as a modifier of
the °matrix predicate, but the reading intended here is the one in which it modifies
the embedded predicate dit boek kopen ‘buy this book’, as is indicated by the traces.
(326) a. Jan vroeg [CP wie C [TP ti dit boek morgen zal kopen]]?
Jan asked who this book tomorrow will buy
‘Jan asked who will buy this book tomorrow.’
b. *Watj vroeg Jan [CP wiei C [TP ti tj morgen zal kopen]]?
what asked Jan who tomorrow will buy
c. *Wanneerj vroeg Jan [CP wiei C [TP ti dit boek tj zal kopen]]?
when asked Jan who this book will buy
Although more can be said about the restrictions on wh-movement (see Section
11.3.1), the above suffices for the main topic in this subsection: which verbs may
function as BRIDGE VERBs, that is, allow wh-extraction from their argument clauses?
For example, whereas verbs of saying/thinking normally license wh-extraction from
their complement clause, verbs of manner of speech such as fluisteren ‘to whisper’
normally do not; this observation is attributed by Erteschik-Shir (2006), to an
unpublished paper by Janet Dean (Fodor) from 1967
(327) a. Wati zei Jan [CLAUSE dat Marie ti gelezen had]?
what said Jan that Marie read had
‘What did Jan say that Marie had read?’
b. ??Wati fluisterde Jan [CLAUSE dat Marie ti gelezen had]?
what whispered Jan that Marie read had
‘What did Jan whisper that Marie had read?’
The examples in (328) show that °irrealis verbs expressing a hope or a wish
constitute another set of verbs that readily license wh-extraction in Dutch; cf.
Haider (2010:108) for the same observation for those varieties of German that allow
wh-extraction from embedded dass-clauses.
 Argument and complementive clauses 759

(328) a. Ik hoop [CLAUSE dat Marie dit boek morgen zal kopen].
I hope that Marie this book tomorrow will buy
‘I hope that Marie will buy this book tomorrow.’
b. Wiei hoop je [CLAUSE dat ti dit boek morgen zal kopen]? [subject]
who hope you that this book tomorrow will buy
‘Who do you hope will buy this book tomorrow?’
c. Wati hoop je [CLAUSE dat Marie ti morgen zal kopen]? [object]
what hope you that Marie tomorrow will buy
‘What do you hope that Marie will buy tomorrow?’
d. Wanneeri hoop je [CLAUSE dat Marie dit boek ti zal kopen]? [adverbial]
when hope you that Marie this book will buy
‘When do you hope that Marie will buy this book?’
Factive verbs like betreuren ‘to regret’, on the other hand, systematically seem to
block wh-extraction from their complement clause given that examples like
(329b-d) are generally judged to be unacceptable; see, e.g., Hoeksema (2006:147).
(329) a. Ik betreur [CLAUSE dat Marie dit boek morgen zal verkopen].
I regret that Marie this book tomorrow will sell
‘I regret that Marie will sell this book tomorrow.’
b. *?Wiei betreur je [CLAUSE dat ti dit boek morgen zal verkopen]? [subject]
who regret you that this book tomorrow will sell
‘Who do you regret will sell this book tomorrow?’
c. *?Wati betreur je [CLAUSE dat Marie ti morgen zal verkopen]? [object]
what regret you that Marie tomorrow will sell
‘What do you regret that Marie will sell tomorrow?’
d. *Wanneeri betreur je [CLAUSE dat Marie dit boek ti zal verkopen]? [adverbial]
when regret you that Marie this book will sell
‘When do you regret that Marie will sell this book?’
Examples like (329b&c), in which an argument is extracted seem degraded but
are often considered to be better than examples such as (329d), in which an adjunct
is extracted. This so-called argument-adjunct asymmetry is often attributed to the
referential status of arguments; see Rizzi (1990). That referential status is relevant
is clear from the fact that argument extraction becomes even better when the
argument is °discourse-linked, that is, when the hearer is asked to pick some entity
or set of entities from some presupposed list. Although there is considerable
variation in acceptability judgments on examples of this type, some speakers even
seem to consider them fully acceptable; see, e.g., Zwart (2011:209) for cases of
object extraction. If acceptable, the examples in (330) show that factive clauses are
not strong, but weak (selective) islands for wh-extraction.
%
(330) a. Welke studenti betreur je [CLAUSE dat ti dit boek zal verkopen]?
which student regret you that this book will sell
‘Which student do you regret will sell this book?’
b. %Welk boeki betreur je [CLAUSE dat Marie ti zal verkopen]?
which book regret you that Marie will sell
‘Which book do you regret that Marie will sell?’
760 Syntax of Dutch: Verbs and verb phrases

All in all, the list of bridge verbs seems to be relatively small. Hoeksema
(2006) collected a sample of such verbs selected from written sources published
after 1780. The complete collection consists of 963 tokens and 110 types. Most
types have a very low frequency: nearly fifty types occur only once. Restricting
ourselves to types occurring minimally six times, we get the result in Table (331).
Bridge verbs are not only relevant for wh-questions but also for relative clause and
topicalization constructions.
(331) Frequently occurring bridge verbs
begrijpen ‘to understand’ 18 # verzekeren ‘to assure’ 8
beweren ‘to claim’ 9 vinden ‘to consider/think’ 34
denken ‘to think’ 318 voelen ‘to feel/think’ 9
geloven ‘to believe’ 29 vrezen ‘to fear’ 10
hopen ‘to hope’ 37 wensen ‘to wish’ 17
horen ‘to hear’ 7 weten ‘to know’ 73 #
menen ‘to suppose’ 62 willen ‘to want’ 119
oordelen ‘to judge’ 7 willen hebben ‘would like’ 6
rekenen (meaning unclear) 6 # zeggen ‘to say’ 59
vermoeden ‘to suspect’ 15 zich voorstellen ‘to imagine’ 8
vertrouwen ‘to trust’ 6 zien ‘to see’ 18
verwachten ‘to expect’ 13

Since Hoeksema does not give his list of attested examples, we searched the
internet with the string [Wat V[±past] je dat] ‘what do/did you V that ...?’ in order to
check whether the verbs in Table (331) occur in the relevant type of wh-question.
The three cases for which we could not find such examples are marked by a number
sign; their use may be restricted to relative clause or topicalization constructions;
see example (322b), for instance, which was taken from Hoeksema (2006). This
leaves us with no more then twenty verbs that are regularly used as bridge verbs in
wh-questions, and there is in fact only one verb, denken ‘to think’, that is really
frequent (>300 tokens). Another relatively frequent bridge verb is the irrealis verb
willen ‘to want’ (>100), but all other verbs are relatively infrequent (<100). Other
corpus-based research has revealed a similar preference for the verb denken and, to
a lesser extent, willen; cf. Verhagen (2005:119ff.) and Schippers (2012:105).

II. Two approaches to wh-extraction

Wh-extraction has given rise to two main lines of research, which Erteschik-Shir
(2006) refers to as, respectively, the structural and the semantic approach. We will
start with arguments in favor of the structural approach, according to which bridge
verbs have some special syntactic property that makes their complement clauses
transparent for wh-movement. For example, Erteschik-Shir mentions that verbs of
manner of speech such as fluisteren ‘to whisper’ differ from verbs of speaking and
thinking in that they can occur without a propositional clause, and she suggests on
the basis of this that embedded clauses co-occurring with verbs of manner of speech
may have a syntactic status different from embedded clauses that co-occur with
verbs of speaking and thinking.
 Argument and complementive clauses 761

(332) a. Jan fluisterde/schreeuwde.

Jan whispered/yelled
b. Jan *zei/??dacht.
Jan said/thought
More support for the structural approach comes from languages like English and
German. In English, the set of bridge verbs seems to coincide more or less with the
set of verbs allowing that-deletion in embedded declarative clauses, an observation
again attributed by Erteschik-Shir (2006) to Janet Dean (Fodor). Verbs of
speaking/thinking, for example, allow that-deletion while factive verbs like to
regret do not (although it is not too hard to find cases on the internet). Since Dutch
does not allow dat-deletion in embedded declarative clauses, we cannot provide
similar evidence on the basis of this language.
(333) a. John thinks Marie will be elected Chair.
b. *John regrets Marie will be elected Chair.
As for German, Haider (1985:55) claims that most bridge verbs trigger embedded
°verb-second in the German varieties that have it (the Northern varieties as well as
standard German), although irrealis verbs like mögen ‘to like’ are an exception to
this general rule; cf. Haider (2010:124, fn. 25). The examples in (334) show that
wh-extraction even requires the embedded finite verb to be in second position in
those varieties that do not allow wh-extraction from embedded declarative dass-
clauses; wh-extraction in dialects not allowing a set-up such as in (334a)
obligatorily triggers verb-second, as in (334b). Since Dutch does not have this form
of embedded verb-second, we cannot provide similar evidence on the basis of this
language.
(334) a. Auf weni glaubte man [CP ti dass [TP sie ti gewartet habe]]? [Southern G.]
for whom believed one that she waited has
‘For whom did people think that she has waited?’
b. Auf weni glaubte man [CP ti [TP habe sie ti gewartet thabe]]? [Northern G.]
for whom believed one has she waited
‘For whom did people think that she has waited?’
In short, arguments in favor of the structural approach to wh-extraction emphasize
that bridge verbs obligatorily take a complement clause and that, in some languages
at least, such verbs may affect the form of these clauses by licensing
complementizer deletion or embedded verb-second. The semantic approach, which
originates in Erteschik-Shir’s (1973) seminal work, emphasizes the effect of
information structure on wh-extraction. The generalization is that wh-extraction is
possible only from clauses which are focused, that is, which express the new
information of the clause. This immediately accounts for the fact that wh-movement
is normally not possible from factive clauses given that these present propositions
the truth of which is presupposed by the speaker; see the discussion in Section
5.1.2.3. It may, however, also account for the contrast between the primeless and
primed examples in (335b&c). That we are dealing with a so-called weak °island is
clear from the fact that the two primed examples exhibit the argument-adjunct
asymmetry discussed in Subsection I. For completeness’ sake, it should be noted
762 Syntax of Dutch: Verbs and verb phrases

that the intended reading of the two (c)-examples is the one in which the adverbial
wanneer ‘when’ modifies the embedded clause; matrix scope is not intended here.
(335) a. Jan zei (niet) [dat Marie dat boek gisteren gekocht had].
Jan said not that Marie that book yesterday bought had
‘Jan said/didnʼt say that Marie had bought that book yesterday.’
b. Wati zei Jan [dat Marie ti gisteren gekocht had]? [argument]
what said Jan that Marie yesterday bought had
‘What did Jan say that Marie had bought yesterday?’
??
b. Wati zei Jan niet [dat Marie ti gekocht had]?
what said Jan not that Marie bought had
‘What didnʼt Jan say that Marie had bought?’
c. Wanneeri zei Jan [dat Marie dat boek ti gekocht had]? [adjunct]
when said Jan that Marie that book bought had
‘When did Jan say that Mary had bought the book?’
c. *Wanneeri zei Jan niet [dat Marie dat boek ti gekocht had]?
when said Jan not that Marie that book bought had
‘*When didnʼt Jan say that Marie had bought the book?’
Erteschik-Shir (1973:95ff.) shows that adding negation to the matrix clause has the
effect of defocusing the embedded clause; whereas example (335a) without
negation can readily be used to introduce the proposition expressed by the
embedded clause in the domain of discourse, example (335b) with negation is most
naturally interpreted as the denial of the presupposed truth of the embedded
proposition. This means that (335b), but not (335b), is in accordance with the
generalization that wh-extraction requires the embedded clause to be part of the
°focus of the clause.
Since in general the addition of information to the matrix clause makes it more
likely that the embedded clause is defocused, the generalization predicts that this
may also have a degrading effect on wh-extraction. This may account for the
contrast between the examples in (327), repeated here as (336). The verb fluisteren
‘to whisper’ is more informative than the verb zeggen ‘to say’ since it includes a
manner component: Jan expressed his assertion in a low voice. In fact, adding a
manner adverb like zachtjes ‘softly’ in (336c) seems to have a similar degrading
effect on wh-extraction, a fact that seems to have escaped attention in the literature
so far.
(336) a. Wati zei Jan [CLAUSE dat Marie ti gelezen had]?
what said Jan that Marie read had
‘What did Jan say that Marie had read?’
b. ??Wati fluisterde Jan [CLAUSE dat Marie ti gelezen had]?
what whispered Jan that Marie read had
‘What did Jan whisper that Marie had read?’
c. ??Wat zei Jan zachtjes [CLAUSE dat Marie ti gelezen had]?
what said Jan softly that Marie read had
‘What did Jan say softly that Marie had read?’
 Argument and complementive clauses 763

Erteschik-Shir’s generalization is completely in line with the findings in Verhagen

(2005:124ff.): on the basis of his corpus research mentioned at the end of
Subsection I, he claims that attested cases of wh-extraction differ only minimally
from what he assumes to be the basic template, which he takes to be the one given
in (337). He further claims that “invented examples of wh-extractions are judged
worse to the degree that they deviate more from the [..] pattern [in (337)]”.
(337) XPwh - denk(en) - pronoun2p [CP dat ...]
think you that
The nature of the evidence reviewed above suggests to us that each of the two
approaches has something different to contribute to the description of wh-
extraction. The structural approaches may be correct in claiming that wh-extraction
is subject to certain structural conditions, for example, that the embedded clause is a
complement of the verb in the matrix clause. The semantic approaches, on the other
hand, may be correct in postulating additional conditions on the use of wh-
extraction constructions, for example, that the embedded clause is the focus of
attention and therefore constitutes the new information of the clause, and that as a
consequence the semantic contribution of the matrix clause must be limited.

5.1.7. Independently used argument clauses

Argument clauses are selected by some higher predicate by definition and we
therefore expect them to occur as dependent clauses only. Nevertheless, it seems
that they sometimes can occur independently. The discourse chunks in (338) show
that this use is discourse-related in the sense that argument clauses can easily occur
independently as an answer to a wh-question.
(338) a. [A] Peter is hier. [B] Wat zei je? [A] Dat Peter hier is.
Peter is here What said you that Peter here is
‘Peter is here. What did you say? That Peter is here.’
b. [A] Kom je nog? [B] Wat vroeg je? [A] Of je nog komt.
come you still what asked you whether you still come
‘Are you coming or not? What did you ask? Whether youʼre coming or not.’
c. [A] Wat doe je? [B] Wat vroeg je? [A] Wat of je doet.
what do you what asked you what whether you do
‘What are you doing? What did you ask? What youʼre doing.’
Such examples can of course be analyzed as cases in which the context allows
omission of the underlined parts of the strings in Ik zei dat Peter hier is ‘I said that
Peter is here’, Ik vroeg of je nog komt ‘I asked whether you are coming or not’, and
Ik vroeg wat of je doet ‘I asked what you are doing ..’. A similar analysis seems
possible for echo-questions of the type in (339), where we may assume that the
underlined parts in Je vraagt me of ik nog kom? ‘Are you asking me whether I am
coming or not?’ and Je vraagt me wat ik doe? ‘Are you asking me what I am
doing?’ are omitted. We refer to De Vries (2001:514) and Den Dikken (2003:7) for
more examples.
764 Syntax of Dutch: Verbs and verb phrases

(339) a. [A] Kom je nog? [B] Of ik nog komt? I denk van niet.
come you still whether I still come I think of not
‘Are you coming or not? Whether Iʼm coming? I donʼt think so.’
b. [A] Wat doe je? [B] Wat of ik doe? Niets.
what do you what whether I do noting
‘What are you doing? What Iʼm doing? Nothing.’
Independently used interrogative non-main clauses are also very common to
express that the speaker is wondering about something. The main and non-main wh-
clauses in (340) seem more or less interchangeable, although the latter has a
stronger emotional load. This emotional load is also reflected by the fact that such
independently used interrogative clauses typically contain some modal element like
nu weer: example (340a) is completely acceptable as a neutral wh-question;
example (340), on the other hand, feels somewhat incomplete and is certainly not
construed as a neutral wh-question, as is marked by means of the “$” diacritic.
(340) a. Wie heeft dat nu weer gedaan? a. Wie heeft dat gedaan?
who has that PRT PRT done who has that done
‘Who has done that?’ ‘Who has done that?’
b. Wie dat nu weer gedaan heeft!? b. $Wie dat gedaan heeft!?
who that PRT PRT done has who that done has
‘Who (for heavenʼs sake) has done that?’
A similar emotional load can be detected in the independently used declarative non-
main clauses in the primed examples in (341); the speaker’s involvement is again
clear from the fact that while the primeless examples can be used as more or less
neutral assertions, the primed examples emphasize that the speaker makes a certain
wish, is uncertain, feels a certain indignation, etc. De Vries (2001:518) argues that
this may be a good reason for considering independently used non-main clauses as
constructions in their own right. Another reason he gives is that such examples have
intonational patterns that differ markedly from those of their embedded
counterparts: for instance, (341a) has a typical exclamation contour, (341b) a
question contour, and (341c) allows various marked intonation patterns.
(341) a. Ik hoop [dat je er lang van genieten mag].
I hope that you there long of have.pleasure may
‘I hope you may enjoy it for a long time.’
a. Dat je er lang van genieten mag! [wish]
b. Ik vraag me af [of dat nou een goed idee is].
I wonder REFL prt. whether that PRT a good idea is
‘I wonder whether that is such a good idea.’
b. of dat nou een goed idee is? [uncertainty]
c. Ik begrijp niet [waar dat nou weer goed voor is].
I understand not where that prt again good for is
‘I donʼt understand whatʼs the use of that.’
c. Waar dat nou goed voor is … [indignation]
 Argument and complementive clauses 765

Independently used non-main clauses may also have highly specialized meanings or
functions that their embedded counterparts lack. For example, when used as an
answer to the question in (342), the independently used of-clause in (342b)
expresses emphatic affirmation: the speaker is replying that he is eager to have the
book in question. This use is so common that it would in fact suffice to answer
(342) with en OF! ‘I sure do!’. Embedded of-clauses cannot perform this function,
but simply express dependent questions.
(342) a. Wil je dit boek hebben?
want you this book have
‘Do you want to have this book?’
b. En OF ik dit boek wil hebben!
and whether I this book want have
‘I sure do want to have that book!’
Because discussing the interpretational implications of the independent uses of
argument clauses would lead us into the domain of the conditions on actual
language use (performance), we will not digress on this. This topic has received
some attention in Cognitive Linguistics since Evans (2007): we refer the reader to
Verstraete et al. (2012), Tejedor (2013), Van Linden & Van de Velde (2014), and
the references cited therein.

5.1.8. Bibliographical notes

Although sentential complementation has been a central concern in syntactic
research over the last fifty years, it is often difficult to pinpoint specific studies;
often the data is already found in traditional grammars and discussed by many
authors. Of course, it is possible to identify several (especially early) seminal
studies like Kiparsky & Kiparsky (1970), Bresnan (1972), and Grimshaw (1979),
but much of what is found in this (and the following) chapter has been developed
over the years by various authors, and it is therefore easier to refer to specific
studies during our discussions. Nevertheless, we want to highlight a number of
studies we used in our discussion of a number of more special issues. The
discussion of factivity in Section 5.1.2 is based on Kiparsky & Kiparsky (1970),
supplemented by insights from Barbiers (2000) and Nye (2013). The discussion of
reported speech/parenthetical clauses was heavily influenced by Corver (1994),
Corver & Thiersch (2003), and De Vries (2006). The discussion of quotative and
polar van-constructions has profited a great deal from Van Craenenbroeck (2002),
Foolen et al. (2006), and Hoeksema (2006). The discussion of fragment clauses is
heavily indebted to the seminal work of Merchant (2001/2006). For a discussion of
the independent uses of argument clauses, we refer the reader to De Vries (2001),
Verstraete et al. (2012), Tejedor (2013), and Van Linden & Van de Velde (2014).

5.2. Infinitival argument clauses

This section discusses infinitival complementation of main verbs. As to their form,
infinitival °argument clauses can be divided into three main types: om + te-, te- and
bare infinitivals. The first type, which will be discussed in Section 5.2.1, is formally
characterized by the fact that the infinitive is preceded by the element te and that
766 Syntax of Dutch: Verbs and verb phrases

the full infinitival clause is (optionally) introduced by the element om, which is
generally considered a complementizer. Some typical examples are given in (343),
in which the infinitival clauses are in italics.
(343) a. Marie weigerde (om) haar fiets te verwijderen.
Marie refused COMP her bike to remove
‘Marie refused to remove her bike.’
b. Jan beloofde (om) dat boek te lezen.
Jan promised COMP that book to read
‘Jan promised to read that book.’
The second type, which will be discussed in Section 5.2.2 and is illustrated in
(344), formally differs from the first one in that the infinitival complementizer om
cannot be used; the infinitive, on the other hand, is preceded by the element te.
(344) a. Jan beweert (*om) dat boek gelezen te hebben.
Jan claims COMP that book read to have
‘Jan claims to have read that book.’
b. Jan verzekerde ons (*om) te mogen komen.
Jan assured us COMP to be.allowed come
‘Jan assured us that we were allowed to come.’
Given the optionality of the complementizer om in examples such as (343), it is
sometimes not a priori clear whether we are dealing with an om + te- or a te-
infinitival and Section 5.2.2.3 will therefore attempt to develop a number of
diagnostics that may help us to make the desired distinction. This section will
further show that there are a number of subtypes of te-infinitivals, which can be
distinguished on the basis of a set of conspicuously distinctive formal properties.
The third type of infinitival clause, the bare infinitivals, is discussed in Section
5.2.3 and can readily be recognized by the fact that elements om and te are both
obligatorily absent; we will see that, again, we can distinguish various subtypes.
(345) a. Jan moet dat boek lezen.
Jan must that book read
‘Jan must read that book.’
b. Ik zag Jan dat boek lezen.
I saw Jan that book read
‘I saw Jan read that book.’
The following sections will extensively discuss these three main types of infinitival
clauses and show that they can be further divided into smaller subcategories on the
basis of their semantic and syntactic properties.

5.2.1. Om + te-infinitivals
This section discusses the use of om + te-infinitivals as arguments of main verbs.
Such clauses are formally characterized by the fact that they are headed by a te-
infinitive and can be preceded by the “linker” element om. Furthermore, they
always contain an implied subject, °PRO, which is normally coreferential with
(°controlled by) the subject or the object of the °matrix clause, although it is
 Argument and complementive clauses 767

sometimes also possible for it to receive a generic interpretation; examples are

given in (346), in which coreference is indicated by means of coindexation and the
index arb(itrary) is used to indicate that the generic reading is intended.
(346) a. Jani beloofde Mariej [(om) PROi/*j dat boek te lezen]. [subject control]
Jan promised Marie COMP that book to read
‘Jan promised Marie to read that book.’
b. Jani verzocht Mariej [(om) PROj/*i dat boek te lezen]. [object control]
Jan requested Marie COMP that book to read
‘Jan requested Marie to read that book.’
c. Jan keurt het af [(om) PROarb te vloeken]. [generic interpretation]
Jan disapproves it prt. COMP to curse
‘Jan disapproves of cursing.’
The discussion is organized as follows. Section 5.2.1.1 starts by showing that
argument clauses in the form of an om + te infinitival have more or less the same
distribution as finite argument clauses: they may have the same syntactic functions
and normally follow the verb(s) in clause-final position. Section 5.2.1.2 continues
with a discussion of the categorial status of the linker om, which has been analyzed
as a regular preposition but also as a complementizer-like element; although this
issue is still not settled, we will for convenience gloss om by means of
COMP(LEMENTIZER) in the examples. Section 5.2.1.3 will conclude the discussion of
om + te infinitivals by investigating the implied subject PRO and the restrictions on
its interpretation in more detail.

5.2.1.1. The distribution of om + te-infinitivals

Infinitival argument clauses preceded by om have more or less the same distribution
as finite argument clauses (cf. Sections 5.1); they can occur as direct objects,
subjects and prepositional objects, but indirect object clauses are rare given that
these usually refer to animate objects and/or institutions. Furthermore, om + te-
infinitivals normally follow the verb(s) in clause-final position, although there are
limited possibilities for them to be topicalized or left-dislocated.

I. Direct object clauses

Object clauses normally follow the verb(s) in clause-final position and can be
optionally preceded by the anticipatory object pronoun het ‘it’; placing the object
clause in the °middle field of the clause generally leads to a highly degraded result.
This is illustrated in the (a)-examples in (347) for finite and by the (b)-examples for
infinitival clauses.
(347) a. dat Jan (het) besloot [dat hij het boek zou kopen].
that Jan it decided that he the book would buy
‘that Jan decided (it) that he would buy the book.’
a. *?dat Jan [dat hij het boek zou kopen] besloot.
b. dat Jan (het) besloot [(om) PRO het boek te kopen].
that Jan it decided COMP the book to buy
‘that Jan decided to buy the book.’
b. *dat Jan [(om) PRO het boek te kopen] besloot.
768 Syntax of Dutch: Verbs and verb phrases

For completeness’ sake, the examples in (348) show that the object clauses of the
verb besluiten ‘to decide’ can also appear as the complement of the corresponding
nominalization besluit ‘decision’.
(348) a. het besluit [dat hij het boek zou kopen]
the decision that he the book would buy
‘the decision that he would buy the book’
b. het besluit [(om) PRO het boek te kopen]
the decision COMP the book to buy
‘the decision to buy the book’
In (347), the object clause is an internal °argument of the verb besluiten ‘to decide’.
Direct object clauses can, however, also function as °logical SUBJECTs (external
arguments) of adjectival °complementives. This is illustrated by means of the
vinden-construction in (349); in these examples the °anticipatory pronoun het is
obligatory and the object clause normally follows the verb(s) in clause-final
position—placement of the object clause in the middle field leads to a severely
degraded result.
(349) a. dat Jan *(het) vervelend vindt [dat hij niet kan komen].
that Jan it annoying considers that he not is.able come
‘that Jan considers it annoying that he isnʼt able to come.’
a. *?dat Jan [dat hij niet kan to komen] vervelend vindt.
b. dat Jan *(het) vervelend vindt [(om) PRO niet te kunnen komen].
that Jan it annoying considers COMP not to be.able come
‘that Jan considers it annoying not to be able to come.’
b. *dat Jan [(om) PRO niet te kunnen komen] vervelend vindt.

II. Subject clauses

Subject clauses are possible if they originate as internal °arguments of the °matrix
verb, as is clear from the fact illustrated in (350) that the primeless examples of the
transitive constructions in (347) can be passivized. The passive construction can be
impersonal, that is, introduced by the °expletive er ‘there’, or personal, that is,
contain the anticipatory subject pronoun het ‘it’. The primed examples in (350)
show that subject clauses must follow the verb(s) in clause-final position; they
cannot be placed in the regular subject position right-adjacent to the
complementizer.
(350) a. dat er/het besloten werd [dat hij het boek zou kopen].
that there/it decided was that he the book would buy
‘that it was decided that he would buy the book.’
a. *?dat [dat hij het boek zou kopen] besloten werd.
b. dat er/het besloten werd [(om) PRO het boek te kopen].
that there/it decided was COMP the book to buy
‘that it was decided to buy the book.’
b. *dat [(om) PRO het boek te kopen] besloten werd.
 Argument and complementive clauses 769

Subject clauses also occur in °dyadic °unaccusative constructions, in which they

likewise originate as internal arguments of the verb. This is illustrated by means of
the NOM-DAT object experiencer psych-verb bevallen ‘to please’ in (351); the
primed examples show again that subject clauses cannot occur in the regular subject
position. In these examples the anticipatory pronoun is obligatory and the expletive
er cannot be used.
(351) a. dat het me niet bevalt [dat hij steeds dezelfde vraag stelt].
that it me not pleases that he constantly the.same question poses
‘that it displeases me that heʼs asking the same question all the time.’
a. *?dat [dat hij steeds dezelfde vraag stelt] me niet bevalt.
b. dat het me niet bevalt [(om) PRO steeds dezelfde vraag te stellen].
that it me not pleases COMP constantly the.same question to pose
‘that it displeases me to ask the same question all the time.’
b. *?dat [om PRO steeds dezelfde vraag te stellen] me niet bevalt.
The examples in (352) show that the same thing holds for the NOM-ACC object
experiencer psych-verb vervelen ‘to annoy’, provided that the subject functions as a
cause (and not as a causer) argument; see Section 2.5.1.3 for an extensive
discussion of these psych-verbs.
(352) a. dat het me verveelt [dat hij steeds dezelfde vraag stelt]cause.
that it me annoys that he constantly the.same question poses
‘that it annoys me that heʼs asking the same question all the time.’
a. *?dat [dat hij steeds dezelfde vraag stelt] me verveelt.
b. dat het me verveelt [(om) PRO steeds dezelfde vraag te stellen].
that it me annoys COMP constantly the.same question to pose
‘that it annoys me to ask the same question all the time.’
b. *dat [om PRO steeds dezelfde vraag te stellen] me verveelt.
In the examples above, the subject clause is an argument of the matrix verb.
Subject clauses can, however, also function as logical SUBJECTs (external
arguments) of adjectival complementives, which is clear from the fact illustrated in
(353) that the vinden-constructions in (349) can be passivized. The anticipatory
pronoun het is normally obligatory and surfaces as the subject of the construction;
placement of the subject clause in the regular subject position is impossible. Again,
the expletive er cannot be used.
(353) a. dat het vervelend gevonden wordt [dat hij niet kan komen].
that it annoying considered is that he not is.able.to come
‘that it is considered annoying that he isnʼt able to come.’
a. *?dat [dat hij niet kan komen] vervelend gevonden wordt.
b. dat het vervelend gevonden wordt [(om) PRO niet te kunnen komen].
that it annoying considered is COMP not to be.able.to come
‘that it is considered annoying not to be able to come.’
b. *dat [(om) PRO niet te kunnen komen] vervelend gevonden wordt.
770 Syntax of Dutch: Verbs and verb phrases

The same thing is illustrated by the copular constructions in (349); again, the
anticipatory pronoun het is normally obligatory and it is impossible to place the
subject clause in the regular subject position.
(354) a. dat het vervelend is [dat hij niet kan komen].
that it annoying is that he not is.able.to come
‘that it is annoying that he isnʼt able to come.’
a. *?dat [dat hij niet kan komen] vervelend is.
b. dat het vervelend is [(om) PRO niet te kunnen komen].
that it annoying is COMP not to be.able.to come
‘that it is annoying not to be able to come.’
b. *dat [(om) PRO niet te kunnen komen] vervelend is.

III. Prepositional object clauses

The examples in (355) show that finite and infinitival clauses can also be used as
PO-clauses, in which case they can be introduced by an °anticipatory pronominal
PP er + P. This pronominal PP can be omitted with certain verbs but not with all;
see Section 2.3.1, sub VI, for detailed discussion. The primed examples show that
complement clauses cannot appear in the middle field of the clause, regardless of
whether or not ernaar is present.
(355) a. dat Jan (ernaar) verlangt [dat hij weer thuis is].
that Jan for.it craves that he again home is
‘that Jan wishes that heʼs home again.’
a. *?dat Jan (naar) [dat hij weer thuis is] verlangt.
b. dat Jan (ernaar) verlangt [(om) PRO weer thuis te zijn].
that Jan for.it craves COMP again home to be
‘that Jan longs to be home again.’
b. *dat Jan (naar) [om PRO weer thuis te zijn] verlangt.
For completeness’ sake, the examples in (356) show that finite and infinitival
clauses can also be used as prepositional complements of adjectives. The
pronominal PP can be omitted with certain adjectives but not with all, and the
complement clause cannot appear in the middle field of the clause.
(356) a. dat Jan (er) bang (voor) is [dat hij te laat komt].
that Jan there afraid of is that he too late comes
‘that Jan is afraid (of it) that heʼll be late.’
a. *dat Jan bang (voor) [dat hij te laat komt] is.
b. Jan is (er) bang (voor) [(om) PRO te laat te komen].
Jan is there afraid of COMP too late to come
‘Jan is afraid (of it) to be late.’
b. *dat Jan bang (voor) [(om) PRO te laat te komen] is.
Interestingly, anticipatory pronominal PPs do not occur in noun phrases. The
nominalizations of the primeless examples in (357) can only be combined with the
pronominal PP ernaar if the clause is not realized. For completeness’ sake, note
 Argument and complementive clauses 771

that for some reason, nominalization gives rise to a somewhat marked result when
the complement is a finite clause.
?
(357) a. het verlangen (*ernaar) [dat hij weer thuis is]
the craving for.it that he again at.home is
b. het verlangen (*ernaar) [(om) PRO weer thuis te zijn]
the craving for.it COMP again at.home to be
c. het verlangen (ernaar)
the craving for.it

IV. Topicalization and left dislocation

The previous subsections have shown that infinitival argument clauses preceded by
om behave like their finite counterparts in that they normally follow the verb(s) in
clause-final position, that is, that they cannot occur in the middle field of the clause.
It is, however, possible to topicalize or left-dislocate the infinitival clause, although
the options seem a bit more limited than in the case of finite argument clauses.

A. Object Clauses
Example (358a) shows that topicalizing a finite object clause is quite normal
(provided that the anticipatory pronoun het is omitted), but that this leads to a
marked result in the case of an infinitival clause; for some speakers examples such
as (358b) improve when emphatic accent is assigned to some element in the
infinitival clause, in this case boek ‘book’.
(358) a. [Dat hij het boek zou kopen] besloot hij snel.
that he the book would buy decided he quickly
‘That he would buy the book he decided quickly.’
b. *?[(Om) PRO het boek te kopen] besloot hij snel.
COMP the book to buy decided he quickly
The contrast disappears, however, in °left-dislocation constructions, especially if
there is some contrastively focused element in the left-dislocated clause. We
illustrate this in (359) by means of contrastive accent on the direct object het boek
‘the book’, but it might equally well have been on the main verb kopen ‘to buy’.
(359) a. [Dat hij het BOEK zou kopen], dat besloot hij snel.
that he the book would buy that decided he quickly
‘That he would buy the book, that he decided quickly.’
b. [(om) PRO het BOEK te kopen], dat besloot hij snel.
COMP the book to buy that decided he quickly
A problem for passing judgments on the examples in (358) is that the resumptive
pronoun dat in (359) is optional, as a result of which the distinction between
topicalization and left dislocation rests entirely on intonation and meaning. First,
topicalized phrases are typically part of a larger prosodic unit, including the finite
verb in second position, while left-dislocated phrases typically constitute a prosodic
unit on their own. Second, topicalized phrases typically refer to known information,
772 Syntax of Dutch: Verbs and verb phrases

whereas left-dislocated phrases typically refer to new or contrastively focused

information.

B. Subject clauses
Subsection II has shown that subject clauses cannot occur in the regular subject
position. This was illustrated by showing that such clauses cannot follow the
complementizer in embedded clauses, and in (360) we show that they cannot follow
the finite verb in second position either.
(360) a. Vaak verveelt het me [dat hij steeds dezelfde vraag stelt].
often annoys it me that he constantly the.same question poses
‘It often annoys me that he always asks the same question.’
a. *?Vaak verveelt [dat hij steeds dezelfde vraag stelt] me.
b. Vaak verveelt het me [(om) PRO steeds dezelfde vraag te stellen].
often annoys it me COMP constantly the.same question to pose
‘It often annoys me to always ask the same question.’
b. *Vaak verveelt [(om) PRO steeds dezelfde vraag te stellen] me.
In the literature we find different evaluations of examples in which infinitival
subject clauses preceded by om occur in sentence-initial, that is, in topicalized
position. Paardekooper (1986: 358) provides examples of the type in (361b) without
any comment and it seems, indeed, that these are just as acceptable as examples
such as (361a) with a finite subject clause.
(361) a. [Dat hij steeds dezelfde vraag stelt]cause verveelt me.
that he constantly the.same question poses annoys me
‘That he always asks the same question annoys me.’
b. [(Om) PRO steeds dezelfde vraag te moeten stellen] verveelt me.
COMP constantly the.same question to have.to pose annoys me
‘Always having to ask the same question annoys me.’
Dik (1985:35), on the other hand, claims that om + te-infinitivals of the type in
(362b) are quite marked, especially if the linker element om is present. It is,
however, not so clear whether this observation is valid for all speakers since some
of our informants do accept examples like these.
(362) a. [Dat hij hier zwemt] is gevaarlijk.
that he here swims is dangerous
‘That he swims here is dangerous.’
b. %[(Om) hier te zwemmen] is gevaarlijk.
COMP here to swim is dangerous
In order to account for the contrast between the (b)-examples in (361) and (362), we
might of course hypothesize that the prohibition against topicalization of infinitival
subject clauses is restricted to cases in which the nominative subject is not an
argument of the verb but the logical SUBJECT of a complementive adjective.
However, this seems to go against Paardekooper’s (1985:117) judgment of example
(363b), which does seem to have a similar status as example (363a) with a finite
subject clause.
 Argument and complementive clauses 773

(363) a. [Dat ik even moest wachten] was niet zo vervelend.

that I a.while had.to wait was not so annoying
‘That I had to wait a while wasnʼt so annoying.’
b. [Om even te moeten wachten] was niet zo vervelend.
COMP a.while to have.to wait was not so annoying
‘To have to wait a while wasnʼt so annoying.’
For the moment, we therefore have to conclude that Dik’s categorical statement that
infinitival clauses preceded by om cannot occupy the sentence-initial position is not
supported by judgments of other speakers, and that infinitival subject clauses can in
fact occupy this position (although they are cases which are judged as somewhat
marked for unknown reasons). For completeness’ sake, the examples in (364) show
that left dislocation of infinitival subject clauses always gives rise to an impeccable
result.
(364) a. [(Om) PRO steeds dezelfde vragen te stellen], dat verveelt me.
COMP constantly the.same questions to pose that annoys me
‘To ask the same questions all the time, that annoys me.’
b. [(Om) hier te zwemmen], dat is gevaarlijk.
COMP here to swim that is dangerous
‘To swim here, that is dangerous.’
c. [Om even te moeten wachten], dat was niet zo vervelend.
COMP a.while to must wait that was not so annoying
‘To have to wait a while, that wasnʼt so annoying.’

C. Prepositional object clauses

The primeless examples show that topicalization of PO-clauses is impossible,
regardless of whether they are finite or infinitival. Left dislocation, on the other
hand, gives rise to an impeccable result, as is shown by the primed examples.
(365) a. *[Dat hij weer thuis is] verlangt Jan (ernaar).
that he again at.home is craves Jan for.it
a. [Dat hij weer thuis is], daar verlangt Jan naar.
that he again at.home is there craves Jan for
‘That heʼs home again, Jan longs for it.’
b. *[(Om) PRO weer thuis te zijn] verlangt Jan (ernaar).
COMP again at.home to be craves Jan for.it
b. [(Om) PRO weer thuis te zijn], daar verlangt Jan naar.
COMP again at.home to be there craves Jan for
‘To be home again, Jan longs for it.’

V. Conclusion
The previous subsections have shown that infinitival argument clauses exhibit
syntactic behavior similar to finite argument clauses. First, they may function as
subject, direct object and prepositional object. Second, they normally appear after
the verb(s) in clause-final position and can be introduced by an anticipatory
pronominal element in the middle field of the clause. The only difference seems to
be related to topicalization; whereas topicalization of finite object clauses is easily
774 Syntax of Dutch: Verbs and verb phrases

possible, topicalization of object om + te infinitivals gives rise to degraded results.

The same contrast has been claimed to hold for subject clauses but we have seen
that there are many cases in which subject clauses can quite felicitously occur in
sentence-initial position, and we therefore provisionally assumed that the reported
contrast is not real. Finite and infinitival prepositional object clauses also behave in
the same way in that they both resist topicalization. We also discussed left
dislocation and showed that in this area there are no differences at all between finite
clause and infinitival clauses preceded by om; left dislocation is always possible.

5.2.1.2. The categorial status of the element om

This section briefly discusses the linker element om that introduces om + te-
infinitivals. The fact that om is optional in argument clauses has given rise to the
claim that om is superfluous and should in fact be avoided, as was stated in the
beginning of the 20th century by Den Hertog (1973:74-5). His advice was in fact not
merely motivated by the optionality of om, but also because that he analyzed om as
a regular preposition. Since subjects and direct objects are normally not introduced
by the preposition om, the use of this preposition is claimed to be improper with
infinitival subject and object clauses. Similarly, since prepositional objects are
already introduced by fixed prepositions, the use of the additional preposition om
with PO-clauses is claimed to be pleonastic in nature.
The claim that the linker om is prepositional in nature has also been defended in
more recent years. Bennis & Hoekstra (1985), for example, argue for the
prepositional status of om on the basis of the distribution of om + te infinitivals.
Their point of departure is the observation extensively discussed in Section 5.2.1.1
that such infinitivals have the same distribution as finite clauses in their use as
arguments of verbs, nouns or adjectives. There is, however, one conspicuous
difference in the distribution of finite clauses and infinitival clauses preceded by
om; whereas the former can be used as the complement of a preposition, the latter
cannot. This is clear from the fact that examples such as (366b) are only acceptable
if om is omitted.
(366) a. Na [dat Jan de wedstrijd gewonnen had] rustte hij uit.
after that Jan the match won had rested he prt.
‘After Jan had won the match, he had a rest.’
b. Na [(*om) PRO de wedstrijd gewonnen te hebben] rustte Jan uit.
after COMP the match won to have rested Jan prt.
‘After having won the match, Jan had a rest.’
Bennis & Hoekstra relate the ungrammaticality of example (366b) with om to the
more general restriction that PPs normally cannot function as complements of
prepositions; see Section P2.2, where we also discuss a small number of exceptions.
If the linker element om is indeed prepositional, the contrast indicated in (366b) is
covered by the same generalization.
A potential p