• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
5.2.3.2.Modal verbs
quickinfo

This section discusses modal verbs like willen'want', moeten'must' and kunnen'may', which take a bare infinitival complement. It is a matter of debate whether modal verbs taking a bare infinitival complement should be classified as main or non-main verbs. Section 4.5, sub II, has discussed in greater detail why we diverge from most descriptive grammars in analyzing these modal verbs as main verbs, and in Subsection I we will briefly repeat some of these reasons.
      Since bare infinitives can be used as heads of both bare infinitival clauses and bare-inf nominalizations, it is impossible to tell without further investigation whether constructions such as (614a) involve nominal or clausal complementation. At least, this holds for Dutch since (614b) shows that, contrary to their English counterparts, modal verbs like willen, moeten and kunnen can also take non-clausal complements.

614
a. Jan wil een ijsje kopen.
  Jan wants.to  an ice.cream  buy
  'Jan wants to buy an ice cream.'
b. Jan wil een ijsje.
  Jan wants  an ice.cream
  'Jan wants to have an ice cream.'

Subsection II therefore reviews the reasons for assuming that these modal verbs take bare infinitival complement clauses, and will also discuss whether these modal verbs can be complemented by bare-inf nominalizations, subsection III continues by providing a discussion of a number of semantic and syntactic properties of the modal verbs under discussion, which adopts as its point of departure the semantic classification of modality provided by Palmer (2001), with one non-trivial addition based on observations found in Klooster (1986) and Barbiers (1995).

readmore
[+]  I.  Modal verbs are main verbs

The main reason for treating modal verbs like willen'want', moeten'must' and kunnen'may' as main verbs here is that they allow pronominalization of their complement; this is shown in the primed examples in (615).

615
a. Jan moet dat boek lezen.
  Jan has.to  that book  read
  'Jan has to read that book.'
a'. Jan moet dat.
  Jan must that
  'Jan has to do that.'
b. Jan wil een ijsje kopen.
  Jan wants.to  an ice.cream  buy
  'Jan wants to buy an ice cream.'
b'. Jan wil dat.
  Jan wants  that
  'Jan wants to do that.'

That modal verbs can function as main verbs is also clear from the fact illustrated in (616) and (617) that it is possible for these verbs to select non-clausal complements; in (616) the complement is nominal in nature and in (617) it has the form of an adjectival/adpositional complementive. We refer the reader to Section 4.5, sub II, for arguments showing that examples like these do not involve a bare infinitival complement with some phonetically empty verb corresponding to the verbs have, get, do, etc. in the English translations.

616
a. Jan wil een ijsje kopen.
  Jan wants.to  an ice.cream  buy
  'Jan wants to buy an ice cream.'
a'. Jan wil een ijsje.
  Jan wants  an ice.cream
  'Jan wants to have an ice cream.'
b. Jan moet zijn medicijnen innemen.
  Jan must  his medicines  in-take
  'Jan must take his medicines.'
b'. Jan moet zijn medicijnen nog.
  Jan must  his medicines  still
  'Jan should take his medicines.'
c. Jan kan alles doen.
  Jan can  everything  do
  'Jan can do anything.'
c'. Jan kan alles.
  Jan can everything
  'Jan can do anything.'
617
a. Deze fles moet leeg.
  this bottle  must  empty
  'This bottle must be emptied.'
b. Die lampen moeten uit.
  those lamps  must  off
  'Those lights must be switched off.'
c. Die boeken kunnen in de vuilnisbak.
  those books  may  into the dustbin
  'Those books may be thrown into the dustbin.'

The standard assumption that (pronominal) noun phrases must be assigned a thematic role (that is, be semantically licensed) by the verb, in tandem with our claim that non-main verbs are not able to do so, leads to the conclusion that modal verbs like moeten and willen are main verbs; see Section 4.5 for more detailed discussion.

[+]  II.  Modal verbs take bare infinitival complement clauses

The previous subsection has shown that modal verbs like willen'want', moeten'must' and kunnen'may' may take nominal complements. Since bare infinitives can be used as heads of both bare infinitival clauses and bare-inf nominalizations, it is therefore not a priori clear whether the primeless examples in (616) involve clausal or nominal complementation. This subsection therefore applies the tests developed in Section 5.2.3.1, repeated here as (618), in order to establish that modal verbs may indeed take bare infinitival complement clauses.

618
The verbal and nominal use of bare infinitives
infinitival clause nominalization
I is part of the verbal complex +
II precedes/follows the governing verb normally follows precedes
III triggers IPP-effect +
IV allows focus movement +
V may follow negative adverb niet'not' +
VI can be preceded by the article geen'no' +

In the examples in (619) the first two tests are applied to examples with willen'want'. First, these examples show that the bare infinitives can be construed as part of the verbal complex, as is clear from the fact that, in clause-final position, willen is able to separate them from their dependents, respectively, the direct object een ijsje'an ice cream' and the adverbial modifier hard'loudly'. Second, they show that the bare infinitives may follow the modal willen in clause-final position. From this we may conclude that the modal verbs are indeed able to take bare infinitival complement clauses.

619
a. dat Jan een ijsje <kopen?> wil <kopenV>.
  that  Jan an ice.cream    buy  wants
  'that Jan wants to buy an ice cream.'
b. dat Jan hard <gillen?> wilde <gillenV>.
  that  Jan loudly    scream  wanted
  'that Jan wanted to scream loudly.'

We marked the bare infinitives preceding the modal verbs in (619) with a question mark, since it remains to be seen whether they are indeed nominal in nature. If so, they should be able to also precede clause-final verbal sequences consisting of two or more verbs. The examples in (620) show, however, that this gives rise to a severely degraded result.

620
a. dat Jan een ijsje <*?kopenN> zou willen <kopenV>.
  that  Jan an ice.cream       buy  would  want
  'that Jan would like to buy an ice cream.'
b. dat Jan hard <*?gillenN> zou willen <gillenV>.
  that  Jan loudly       scream  would  want
  'that Jan would like to scream loudly.'

The examples in (620) suggest that modal verbs do not comfortably take bare-inf nominalizations as their complement. This conclusion is also supported by the fact that the IPP-effect is obligatory (test III); the modal verb must surface as an infinitive in the perfect-tense constructions in the primeless examples in (621). The status of the primed examples is comparable to the status of the examples in (620) with the infinitive preceding the verbs in clause-final position.

621
a. dat Jan een ijsje had willen/*gewild kopenV.
  that  Jan an ice.cream  had want/wanted  buy
  'that Jan had wanted to buy an ice cream.'
a'. *? dat Jan een ijsje kopenN had gewild.
  that  Jan an ice.cream  buy  had wanted
b. dat Jan hard had willen/*gewild gillenV.
  that  Jan loudly  had want/wanted  scream
  'that Jan had wanted to scream loudly.'
b'. *? dat Jan hard gillenN had gewild.
  that  Jan loudly  scream  had wanted

If modal verbs indeed resist bare-inf nominalizations as complements, we expect focus movement to be excluded (test IV). The examples in (622) show that it is not clear whether this is borne out; the examples are marked but it seems too strong a claim to say that they are unacceptable. Note that the perfect-tense constructions in the primed examples would become completely ungrammatical if we replace the past participle gewild by the infinitive willen.

622
a. ? dat Jan een ijsje kopenN waarschijnlijk wel zou willen.
  that  Jan an ice.cream  buy  probably  prt  would  want
  'that Jan would probably like to buy an ice cream.'
a'. ? dat Jan een ijsje kopenN waarschijnlijk wel had gewild.
  that  Jan an ice.cream  buy  probably  prt had wanted
b. ? dat Jan hard gillenN waarschijnlijk wel zou willen.
  that  Jan loudly  scream  probably  prt  would  want
  'that Jan would probably like to scream loudly.'
b'. ? dat Jan hard schreeuwenN waarschijnlijk wel had gewild.
  that  Jan loudly  scream  probably  prt  had wanted

The two negation tests again suggest that modal verbs do not easily take bare-inf nominalizations as their complement; the fact that the bare infinitive zingen cannot be preceded by the negative article geen'no' in (623b) suggests that it must be interpreted as verbal.

623
a. dat Jan niet wil zingen.
  that  Jan not  wants  sing
  'that Jan doesnʼt want to sing.'
b. dat Jan niet/*geen zingen wil.
  that  Jan not/no sing  wants
  'that Jan doesnʼt want to sing.'

The examples above have shown that modal infinitives normally do not take bare-inf nominalizations as their complement. Possible exceptions are cases such as (622), in which the bare infinitive is not adjacent to the verb sequence in clause-final position as the result of focus movement. The same thing may in fact hold for cases in which the infinitive is topicalized, as is can be inferred from the fact that the IPP-effect does not apply in the perfect-tense constructions in the primed examples of (624). We will return to this issue in Section 11.3.3, sub VIC.

624
a. Een ijsje kopen zou Jan wel willen.
  an ice.cream buy  would  Jan prt want
a'. Een ijsje kopen had Jan wel gewild/*willen.
  an ice.cream buy  had  Jan prt wanted/want
b. Hard gillen zou Jan wel willen.
  loudly scream  would  Jan  prt  want
b'. Hard gillen has Jan wel gewild/*willen.
  loudly scream  had  Jan  prt  wanted/want
[+]  III.  Types of modality

Palmer (2001) provides a semantic classification of modality based on cross-linguistic research. Following his classification, we can divide the Dutch modal verbs taking a bare infinitival complement as in (625). As Palmer also noted for modality markers in other languages, Dutch modal verbs are often ambiguous: the verbs moeten'must/be obliged', kunnen'may/be able' and zullen'will/shall', for example, can be used to express propositional or event modality. Observe that the modal verbs given in (625) are just the ones that are prototypically associated with the type of modality in question; they may, however, also have less prototypical uses, which we will discuss as we go along.

625
Classification of modal verbs taking a bare infinitival (after Palmer 2001)
a. Epistemic propositional modality:
a. Deductive: moeten'must'
b. Speculative: kunnen'may'
c. Assumptive: zullen'will'
a'. Evidential propositional modality:
a. reported: —
b. Sensory: —
b. Deontic event modality:
a. Permissive: mogen'may/be allowed'
b. Obligative: moeten'must/be obliged'
c. Commisive: zullen'shall'
b'. Dynamic event modality:
a. Ability: kunnen'can/be able'
b. Volitive: willen'will/want'

The schema in (625) shows that modal verbs taking a bare infinitival clause cannot be used to express evidential modality. The discussion of the different types of modality below will show that this does not necessarily mean that there are no specialized verbs that can have such a function, but only that they do not belong to (or are normally not considered part of) the set of verbs under discussion here; we return to this in Subsection A2.
      The examples in (626) and (627) below illustrate the basic distinction between propositional and event modality. In (626a), the modal verbs express propositional modality in the sense that they provide the speaker's evaluation of the factual status of the proposition be at home(Marie). This is clear from the fact that examples like these are frequently paraphrased in the linguistic literature as in (626b), where the modal predicate Vmod wel zo zijn'V be the case' in the main clause is clearly predicated of the embedded finite clause that functions as the logical subject of the main clause (via the anticipatory subject pronoun het'it', which is indicated by subscripts).

626
a. Marie moet/kan/zal nu wel thuis zijn.
propositional modality
  Marie must/may/will  now  prt  at.home  be
  'Marie must/may/will be at home now.'
b. Heti moet/kan/zal wel zo zijn [dat Marie nu thuis is]i.
  it  must/may/will  prt  the.case  be   that  Marie now  at.home  is
  'It must/may/will be the case that Marie is at home now.'

In (627a), the modal verbs express event modality. The speaker is not so much interested in the factual status of the proposition read (Marie, the book), which is typically not (yet) actualized at speech time, but in the moving force that is involved in the potential realization of the eventuality. This is clear from the fact that examples like these are generally paraphrased as in (627b), in which the predicate in the main clause is not predicated of the embedded finite clause but of the agent of the proposition expressed by the embedded clause (which is indicated by coindexing of the subject of the main clause and the implied PRO-subject of the embedded clause).

627
a. Marie moet/mag/zal het boek binnenkort lezen.
event modality
  Marie must/may/will  the book  soon  read
  'Marie must/may/shall read the book soon.'
b. Mariei is verplicht/in staat [om PROi het boek binnenkort te lezen].
  Marie  is obliged/in able  comp  the book  soon  to read
  'Marie is obliged/able to read the book soon.'

Further subdivisions of these two main types of modality will be discussed in the following subsections, Subsection A on propositional modality is relatively brief because the semantics of epistemic modality is also extensively discussed in Section 1.5.2 and evidential modality is normally (tacitly and perhaps wrongly) assumed not to be expressed by modal verbs in Dutch, Subsection B on event modality shows that Palmer's distinction between dynamic and deontic modality is not adequate enough, and that deontic modality in fact refers to two different types of modality with different semantic and syntactic properties. This will lead to a revision of the classification in (625) as in (628), Subsection C concludes by providing a binary feature analysis of these four types of modality.

628
Revised classification of modal verbs taking a bare infinitival
a. Epistemic (propositional modality type I)
b. Directed deontic (event modality type Ia)
c. Non-directed deontic (event modality type Ib)
d. Dynamic (event modality type II)
[+]  A.  Propositional modality

Propositional modality is related to the speaker's evaluation of the factual status of the proposition expressed by the projection of the main verb embedded under the modal verb. According to Palmer (2001), judgments may be of two different kinds: there are epistemic and evidential judgments, which are concerned with, respectively, the truth value of the proposition and the evidence that is available for the truth of the proposition.

[+]  1.  Epistemic modality

If modal verbs are used to express epistemic judgments, they indicate the likelihood of the actual occurrence of a specific eventuality. Although we will not address this issue here, the notion of actual occurrence should be understood as "actual occurrence within the present/past-tense interval"; see Section 1.5.2 for detailed discussion. This subsection focuses on the fact that Palmer distinguishes three types of epistemic judgments, which he refers to as speculative, deductive and assumptive, and which are prototypically expressed in Dutch by, respectively, kunnen'may', moeten'must' and zullen'will'.

629
a. Marie kan nu thuis zijn.
speculative
  Marie may  now  at.home  be
b. Marie moet nu thuis zijn.
deductive
  Marie must  now  at.home  be
c. Marie zal nu thuis zijn.
assumptive
  Marie will  now  at.home  be

By uttering sentences such as (629a-c), the speaker provides three different epistemic judgments about (his commitment to the truth of) the proposition be at home (Marie), as expressed by the lexical projection of the embedded main verb embedded under the modal verb. The use of kunnen'may' in (629a) presents the proposition as a possible conclusion: the speaker is uncertain whether the proposition is true, but on the basis of the information available to him he is not able to exclude it. The use of moeten'must' in (629b) presents the proposition as the only possible conclusion: on the basis of the information available the speaker concludes that the proposition is true. The use of zullen'will' in (629c) presents the proposition as a reasonable conclusion on the basis of the available evidence. The type of evidence on which the speakers evaluation is based is not important; it may consist of any information available to the speaker, including experience and generally accepted knowledge as in Het is vier uur; Marie kan/moet/zal nu thuis zijn'It is four o'clock; Marie may/must/will be at home now'.
      It is not immediately clear whether the three subtypes of epistemic modality in (629) are exhaustive. The slightly different constructions with mogen and willen in (630), for example, may be instantiations of epistemic modality but also have some additional meaning aspects (which may partly be attributed to the particles dan and nog wel eens). For example, the clause with the verb mogen in the first conjunct of (630a) is assumptive in that it indicates that the speaker accepts that the proposition Jan is smart is true, but the central meaning aspect of the sentence as a whole is concessive and somewhat depreciatory in nature; the second conjunct asserts the main message that Jan is not very clever with his hand; see also Haeseryn et al.(1997:1618). Similarly, the construction with the verb willen in (630b) seems speculative in nature but the main message of the construction as a whole is that the lamp has the tendency to topple over.

630
a. Jan mag dan erg slim zijn, maar hij is niet handig.
  Jan may  prt  very smart  be  but  he  is  not  deft
  'Jan may well be very smart, but he isnʼt clever with his hands.'
b. Die lamp wil nog wel eens omvallen.
  that lamp  wants  prt  prt  occasionally  prt.-fall
  'That lamp has the tendency to topple over.'

      Let us return to the judgments concerning the truth of the proposition be at home (Marie). It is clear from the paraphrases of (629) in (631) that the truth value of the embedded proposition is being evaluated epistemically: in the paraphrases the epistemic judgment and the proposition are expressed by different clauses; the latter is now expressed as a finite embedded clause that functions as the logical subject of the epistemic predicate in the main clause.

631
a. Heti kan zo zijn [dat Marie nu thuis is]i.
  it  may  the.case  be   that  Marie now  at.home  is
  'It may be the case that Marie is at home now.'
b. Heti moet zo zijn [dat Marie nu thuis is]i.
  it  must  the.case  be   that  Marie now  at.home  is
  'It must be the case that Marie is at home now.'
c. Heti zal zo zijn [dat Marie nu thuis is]i.
  it  will  the.case  be   that  Marie now  at.home  is
  'It will be the case that Marie is at home now.'

That we are dealing with special cases of epistemic modality in (630) might be supported by the fact that these examples can be given similar paraphrases as the examples in (629), as is shown by the examples in (632).

632
a. Heti mag dan zo zijn [dat Jan erg slim is]i, maar hij is niet handig.
  it  may  prt  the.case  be   that  Jan very smart  is but  he  is  not  deft
  'It may well be that Jan is very smart, but he isnʼt clever with his hands.'
b. Heti wil nog wel eens zo zijn [dat die lamp omvalt]i.
  it  wants prt prtoccasionally  the.case  be   that  that lamp  prt.-fall
  'That lamp has the tendency to topple over.'

Note in passing that we have used the predicate Vmod wel zo zijn'V well be the case' in (631) and (632), but that the modal verb kunnen'may' can also function autonomously as the epistemic predicate: cf. Het kan dat Marie nu thuis is'It may be that Marie is at home now'. This autonomous use seems less common with moeten, mogen and willen, and virtually impossible with zullen.
      That epistemic modal verbs are predicated of a propositional complement is also clear from the (b)-examples in (633); the modal verbs are predicated of the demonstrative pronoun dat'that', which is interpreted as referring to the proposition expressed by Marie is nu thuis'Marie is at home now'.

633
a. Wat denk je: is Marie nu thuis?
  what  think  you  is Marie now  at.home
  'What do you think: Is Marie at home at this moment?'
b. Ja, dat kan/moet wel.
  yes,  that  may/must  prt
  'Yes, that may/must be so.'
b'. Ja, dat zal wel.
  yes,  that  will  prt
  'Yes, that will be so.'

The (b)-examples in (633) also unambiguously show that epistemic modal verbs are monadic; they take just a single propositional argument. This implies that the nominative subject Marie in (629) cannot be selected by the modal verb. This, in turn, implies that this noun phrase is licensed by the main verb embedded under the modal verb and that it is subsequently promoted to the subject position of the entire clause. This so-called subject raising analysis is schematically given in (634) for the verb moeten; we will see in Subsection B that epistemic modal verbs crucially differ in this respect from modal verbs expressing dynamic and directed (but not non-directed) deontic modality.

634
Epistemic modality (Subject Raising)
a. —— moet [VP Marie nu wel thuis zijn]
b. Mariei moet [VPti nu wel thuis zijn].

That the nominative subject of the clause is selected by the embedded main verb is also supported by the fact that the subject of the clause can be part of an idiomatic construction such as (635a). If the subject was selected by the modal verb, the availability of this idiomatic reading would be quite surprising because an idiom is stored as a unit in the lexicon.

635
a. De beer is los.
  the boar  is loose
  'The fatʼs in the fire.'
b. De beer moet/kan/zal nu wel los zijn.
  the boar  must/may/will  now  prt  loose  be
  'The fat must/may/will be in the fire by now.'
[+]  2.  Evidential (reported and sensory) modality

Evidentiality does not pertain to the truth of the proposition, but to the evidence that supports it. Palmer (2001; Section 2.2) distinguishes two types of evidence. The first type involves reported evidence, and includes evidence based on second and third-hand reports, hearsay, etc. Dutch does not seem to have special modal verbs to express this type of evidential modality with, perhaps, one exception: the past-tense form of zullen'will' can be used to express that the speaker does not commit himself to the proposition but bases himself on some source of the information, which generally remains unidentified but which can, in principle, be made explicit by means of an adverbial volgens-PP.

636
a. Hij zou steenrijk zijn.
  he  would  immensely.rich  be
  'Heʼs said to be immensely rich.'
b. Hij zou volgens Peter/welingelichte kringen steenrijk zijn.
  he  would  according.to Peter/informed circles  immensely.rich  be
  'According to Peter/informed circles, heʼs immensely rich.'

Note in passing that the options in Dutch are more limited than in German, which can use the present as well as the past tense of the verb sollen and also of the verb wollen'will' to express evidential modality of this type; see Palmer (2001; Section 2.2.2) and Erb (2001:82) for discussion and examples. It should also be stated that the fact that Dutch does not have specialized modal verbs to express evidentiality of this kind does not mean that it has no means to express it: verbs of communication like zeggen'to say' are, of course, capable of performing this function.
      Palmer refers to the second type of evidential modality as sensory, and this pertains to evidence obtained by means of the senses. It may be claimed that this type of modality is expressed in Dutch by means of the perception verbs when they take a bare infinitival clause. Example (637a), for instance, expresses that the speaker has direct, auditory, evidence that the proposition Jan vertrok'Jan has left' is true. In this respect (637a) crucially differs from (637b), which indicates that the speaker does not have any direct evidence to support the truth of the proposition Jan vertrok'Jan has left'; he may have heard something from which he concludes that the proposition is true, or he may have been told so by some other person.

637
a. Ik hoorde [Jan vertrekken].
  heard   Jan leave
b. Ik hoorde [dat Jan vertrok].
  heard   that  Jan left

There are several facts supporting the idea that perception verbs may function as markers of evidential/sensory modality. First, perception verbs are like the unequivocal modals moeten, kunnen and zullen in that they take bare infinitivals as their complement, albeit that these infinitival complements may contain an (optional) overt subject. Secondly, it seems that the verbs zien'to see' and horen'to hear' are the ones that most frequently occur with a bare infinitival, which is in line with the fact that, cross-linguistically, sensory evidential modality is also most frequently expressed by markers pertaining to visual and auditory stimuli. Thirdly, it may account for the acceptability of examples like (638b&c) with the verb vinden'to consider': like constructions with perception verbs, the vinden-construction takes a bare infinitival complement typically referring to an eventuality that eventuality that can be perceived by means of the senses, while expressing further that the truth assignment to the proposition denoted by the bare infinitival clause is based on the (subjective) sensory perception of the subject of the clause.

638
a. Ik vind [Els goed dansen en zingen], (maar hij niet).
vision/hearing
  consider   Els well  dance and sing   but  he not
  'I think that Els is dancing and singing well (but he doesnʼt).'
b. Els vindt [die soep lekker ruiken/smaken] (maar ik niet).
smell/taste
  Els considers   that soup  nicely  smell/taste   but I not
  'Els thinks that the soup smells/tastes nice (but I donʼt).'
c. Ik vind [die trui naar prikken] (maar hij niet).
touch
  consider   that sweater  unpleasantly prickle   but he not
  'I think that this sweater is unpleasantly prickly (but he doesnʼt).'

By assuming that Dutch has a set of modal verbs expressing sensory evidentiality, we avoid the need of postulating a separate class of verbs consisting of the verb vinden'to consider' only which has properties virtually identical to those of the class of perception verbs (one noticeable difference being that the subject of the bare infinitival complement clause of vinden cannot be omitted). However, given that the perception verbs are normally not treated as a subtype of modal verbs, we will not pursue this option here, but discuss them in their own right in Section 5.2.3.3.
      Finally, it is to be noted that Dutch verbs like blijken'to turn out', lijken'to appear', and schijnen'to seem' in (639) are evidential in the sense that they can be used to indicate whether there is direct evidence in favor of the truth of the proposition, whether there are identifiable individuals that can be held responsible for the truth of the proposition, or whether we are dealing with hearsay/rumors; see Vliegen (2011). Since blijken, lijken and schijnen do not select bare infinitival complements they are not discussed here but in Section 5.2.2.2.

639
a. Uit deze feiten blijkt [dat Jan de dader is].
direct evidence
  from these facts  turns.out   that  Jan the perpetrator  is
  'These facts clearly show that Jan is the perpetrator.'
b. Het lijkt mij/haar [dat Jan de dader is].
identifiable source
  it  appears  me/her   that  Jan the perpetrator  is
  'It appears to the me/her that Jan is the perpetrator.'
c. Het schijnt [dat Jan de dader is].
hearsay/rumors
  it  seems   that  Jan the perpetrator  is
  'It seems that Jan is the perpetrator.'
[+]  B.  Event modality

Event modality is concerned with the moving force involved in the (potential) realization of the eventuality denoted by the lexical projection of the main verb embedded under the modal verb. The moving force may be internal to the person referred to by the subject of the full construction (ability or volition): Palmer refers to this type as dynamic modality, but a more telling name might be dispositional modality. The moving force may also be external to the person referred to by the subject of the full construction (obligation or permission), in which case we are dealing with deontic modality. In both these cases the moving force is directed towards the subject of the sentence. Klooster (1986) and Barbiers (1995) have shown, however, that there is a second type of deontic modality in which the moving force is not directed towards the subject at all; in order to distinguish these two types of deontic modality we will make a distinction between directed and non-directed deontic modality, where (non-)directed should be interpreted as "(not) directed towards the subject of the sentence". We thus distinguish the three types of event modality illustrated in (640), which will be discussed in some more detail in the following subsections.

640
a. Jan wil Marie bezoeken.
dynamic/dispositional modality
  Jan wants  Marie visit
  'Jan wants to visit Marie.'
b. Jan moet van zijn vader het hek verven.
directed deontic
  Jan has.to  of his father  the gate  pain
  'Jan has to paint the gate; his father asked him to do so.'
c. Jan moet meer hulp krijgen.
non-directed deontic
  Jan has.to  more help  get
  'Jan has to receive more help.'
[+]  1.  Dynamic/dispositional modality

Dynamic/dispositional modality describes some moving force internal to the nominative subject of the construction as a whole that favors the realization of the potential event denoted by the main verb embedded under the modal verb. Two verbs that are prototypically used in this modal function are kunnen'to be able' and willen'to want', which express ability and volition, respectively.

641
a. Jan kan dat boek lezen.
ability
  Jan is.able  that book  read
  'Jan can read that book'
b. Jan wil dat boek lezen.
volition
  Jan wants  that book  read
  'Jan wants to read that book.'

That the modal verbs in (641) function as main verbs is quite clear, as we have seen earlier, from the fact illustrated in (642) that the bare infinitival clause can be pronominalized. These examples also show that the subject of the sentence is not part of the infinitival clause, which shows that dynamic/dispositional verbs differ from epistemic modal verbs in that they are not monadic but dyadic predicates.

642
a. Jan kan dat.
  Jan is.able  that
  'Jan can do that.'
b. Jan wil dat.
  Jan wants  that
  'Jan wants to do that.'

In order to account for the fact that the nominative subject of the construction as a whole is also construed as the subject of the infinitival clause, Klooster (1986) proposed a control analysis of constructions of this type: the external argument of the modal verb functions as the controller of the implied subject PRO of the embedded infinitival clause. This is schematically represented in (643), in which coindexing indicates coreference.

643
Dynamic/dispositional modality (Control)
a. Jani kan [PROi dat boek lezen].
  Jan  is.able  that book  read
b. Jani wil [PROi dat boek lezen].
  Jan  wants  that book  read

      The modal verbs moeten'must' and zullen'will' can also be used to express dynamic/dispositional modality, in which case they express, respectively, a strong will/desire and determination. The primeless examples in (644) show that this use of moeten and zullen is rather special in that it normally requires the modal verb to have emphatic accent.

644
a. Jan moet dat boek lezen.
strong will/desire
  Jan must  that book  read
  'Jan definitely must read that book.'
b. Jan zal dat boek lezen.
determination
  Jan will  that book  read
  'Jan will read that book (nothing will stop him).'

The examples in (645) show that this use of moeten and zullen is also special in that pronominalization of the bare infinitival clause gives rise to a less felicitous result. The degraded status of (645a) under the intended reading can perhaps be accounted for by appealing to the fact that the directed deontic (obligation) reading of this example is simply the more prominent one, but a similar account is not available for the degraded status of (645b).

645
a. # Jan moet dat.
  Jan  must  that
b. *? Jan zal dat.
  Jan  will  that

The modal verb zullen often occurs in coordinated structures with the other dynamic/dispositional modal verbs in order to express determination in addition to ability, volition, desire, etc; especially the combination moet en zal is very frequent, and has the fixed meaning "nothing will stop me from ...". All examples in (646) are taken from the internet and require the modal verbs to have emphatic accent. Pronominalization of the bare infinitival clause is not illustrated here but again gives degraded results in these cases.

646
a. Ik kan en zal doen wat ik wil.
  am.able  and  will  do  whatever  like
  'Iʼm able to do whatever I want, and Iʼll definitely do so.'
b. Amerika wil en zal Julian Assange veroordelen.
  US  wants  and  will  Julian Assange  convict
  'The US wants to convict Julian Assange, and will definitely do so.'
c. Ik moet en zal goed leren zingen.
  must  and  will  well  learn  sing
  'Nothing will stop me from learning to sing well.'

      The modal verb mogen, which is normally used as a deontic verb, can occasionally be found with a dynamic/dispositional function as well if it is accompanied by the adverbial phrase graag'gladly'.

647
Ik mag graag wandelen.
  I   like.to  gladly  walk
'I like to walk.'
[+]  2.  Directed deontic modality

Directed deontic modality differs from dynamic/dispositional modality in that the moving force is not internal, but external to the subject of the sentence. Two verbs that are prototypically used with this modal function are moeten'to be obliged' and mogen'to be allowed', which express obligation and permission, respectively. Note that when the external force is some person in authority, it can be made explicit by means of an adverbial van-PP.

648
a. Jan moet dat boek lezen van zijn vader.
obligation
  Jan must  that book  read  of his father
  'Jan has to read that book (his father asked him to do so).'
b. Jan mag dat boek lezen van zijn vader.
permission
  Jan may  that book  read  of his father
  'Jan may read that book (he has his fatherʼs permission).'

The external force may also be impersonal (laws and other regulations), in which case the force can be expressed by means of a volgens-PP.

649
a. Volgens de regels moet de voorzitter de vergadering openen.
  according.to the rules  must the chairman  the meeting  open
  'According to the rules, the chairman must open the meeting.'
b. Volgens gewoonte mag de vader de bruid weggeven.
  according.to custom  may  the father  the bride  away-give
  'According to custom, the father may give away the bride.'

That the modal verbs in (648) function as main verbs is clear from the fact that the bare infinitival clause lends itself to pronominalization quite easily. The examples in (650) show that the subject of the sentence is not part of the infinitival clause, from which we may conclude that modal verbs expressing directed deontic modality are similar to modal verbs expressing dynamic/dispositional modality in that they are not monadic but dyadic predicates.

650
a. Jan moet dat van zijn vader.
  Jan must  that  of his father
  'Jan has to do that; his father asked him to do so.'
b. Jan mag dat van zijn vader.
  Jan may  that  of his father
  'Jan may do that; he has his fatherʼs permission.'

Pronominalization is perhaps somewhat marked in the case of examples such as (649); it seems preferred to substitute the verb phrase doen + dat'to do that' for the infinitival clause. It should be noted, however, that negative clauses with the deontic modal mogen'to be allowed' are very normal without the verb doen: Volgens de regels mag hij dat niet (doen)'According to the rules, he is not allowed to do that'.

651
a. Volgens de regels moet de voorzitter dat ?(doen).
  according.to the rules  must the chairman  that    do
  'According to the rules the chairman must do that.'
b. Volgens gewoonte mag de vader dat ?(doen).
  according.to custom  may  the father  that    do
  'According to custom the father may do that.'

The fact that the nominative subject of the construction as a whole is not affected by pronominalization indicates that directed deontic constructions are like dynamic/dispositional modal constructions in that they are amenable to a control analysis. This is shown in (652).

652
Directed deontic modality (Control)
a. Jani moet [PROi dat boek lezen].
  Jan  must  that book  read
b. Jani mag [PROi dat boek lezen].
  Jan may  that book  read

      The verb zullen can also be used to express directed deontic modality if the speaker wants to express that he is committed to the actualization of the proposition denoted by the lexical projection of the main verb; by using an example such as (653) the speaker indicates that he has the authority to instruct the technical department and is hence able to promise that everything will be fixed. For the same reason, examples such as (653b) are felt as rude (or even as a threat) given that the speaker indicates that he has the authority to boss the addressee about (and to take certain measures if he does not obey).

653
a. Onze technische dienst zal alles in orde brengen.
  our technical department  will  everything  in order  bring
  'Our technical department will fix everything.'
b. Je zal vanmiddag alles in orde brengen.
  you  will  this.afternoon  everything  in order  bring
  'You shall fix everything this afternoon.'

      Although the verb willen'want' cannot be used to express directed deontic modality, Barbiers (1995) suggests that the verb kunnen'can' in examples such as (654a) is able to do so, and indeed examples like these can be construed as a speech act of granting permission. It is, however, not so clear whether we are really dealing with directed deontic modality, since Palmer (2001:77) notes that examples such as (654b) may simply express that there is nothing to stop Jan from leaving and suggests that we are dealing with dynamic/dispositional (ability) modality here. A similar interpretation may be possible for (654a) if we assume that the speaker indicates by using this sentence that in his view that all preconditions for Jan's leaving are fulfilled.

654
a. Jan kan vertrekken.
  Jan may  leave
  'Jan is able to leave (the speaker lifts any prohibition).'
b. Jan kan ontsnappen.
  Jan can  escape
  'Jan is able to escape (there is no external impediment).'

The discussion of the examples in (654) shows that it is not always easy to determine the type of modality that we are dealing with, but we will see in the next subsection that there may be reason for assuming that kunnen can indeed be used as a directed deontic modal.

[+]  3.  Non-Directed deontic modality

This subsection discusses a type of modal construction that was first discussed in Klooster (1986) and called non-directed modality in Barbiers (1995). In order to introduce this type of event modality, we will begin with a brief digression on passivization of clauses expressing event modality. First, consider example (655a), which expresses dynamic/dispositional modality: the agent het meisje'the girl' has the wish to stroke the cat. The passive counterpart of this example in (655b) likewise expresses dynamic/dispositional modality, although now it is the patient de kat'the cat' that wants to be stroked. In both cases, however, the moving force is internal to the nominative subject of the construction as a whole.

655
a. Het meisje wil de kat aaien.
subject-oriented modality
  the girl  wants  the cat  stroke
  'The girl wants to stroke the cat.'
b. De kat wil door het meisje geaaid worden.
subject-oriented modality
  the cat wants  by the girl stroked  be
  'The cat wants to be stroked by the girl.'

That we are dealing with regular dynamic/dispositional modality in the passive construction in (655b) is also supported by the fact illustrated in (656) that the modal verb remains dyadic under passivization; pronominalization of the infinitival clauses does not affect the nominative subject regardless of the voice of the embedded clause. This shows that the nominative subject cannot originate in the embedded infinitival clause but must be selected by the modal verb itself; we are dealing with control structures in both the active and the passive case.

656
a. Het meisje wil dat.
modal is dyadic
  the girl  wants  that
b. De kat wil dat.
modal is dyadic
  the cat  wants  that

That the nominative subject of the construction as a whole is selected by willen is also supported by the examples in (657), which show that the subject of the passive construction must have volition. If not, the construction is semantically incoherent.

657
a. Jan wil het hek verven.
  Jan wants  the gate  paint
  'Jan wants to pain the gate.'
b. $ Het hek wil geverfd worden.
  the gate  wants  painted  be

      Things are quite different in the case of directed deontic constructions. This is immediately clear from the fact that examples such as (658a) can readily be passivized, with the result that the nominative subject of the passive construction is an inanimate entity without control over the proposition expressed by the infinitival clause; the fact that (658b) is nevertheless semantically coherent shows that the obligation expressed by the modal verb moeten cannot be directed towards the subject of the clause but must be directed towards the implicit agent of the infinitival verb verven'to paint'.

658
a. Jan moet het hek verven van zijn vader.
subject-oriented modality
  Jan must  the gate  paint  of his father
  'Jan must paint the gate (his father asked him to do that).'
b. Het hek moet geverfd worden van zijn vader.
no subject-oriented modality
  the gate must painted  be  of his father
  'The gate must be painted (his father requested it).'

Barbiers (1995) refers to examples in (658a) and (658b) as, respectively, directed and non-directed deontic modality, where (non-)directed should be interpreted as "(not) directed towards the subject of the sentence". The examples in (659) show that the active and passive constructions in (658) differ not only with respect to the directional force of the modal, but also as regards pronominalization of the embedded infinitival clause; whereas the nominative subject of the active construction is not affected by pronominalization, the nominative subject of the passive construction is.

659
a. Jan moet dat van zijn vader.
modal is dyadic
  Jan must  that  of his father
  'Jan must do that (his father asked him to do that).'
b. Dat moet van zijn vader.
modal is monadic
  that  must  of his father
  'That gate must be done (his father requested it).'

As is already indicated by the comments in straight brackets in (656) and (659), dynamic/dispositional and deontic constructions differ with respect to the origin of the nominative subject of the construction as a whole. In dynamic/dispositional constructions the subject originates as an argument of the modal verb, regardless of whether the embedded infinitival clause is in the active or the passive voice; the schematic representations in (660a&a') show that we are dealing with control structures in both cases. In deontic constructions, on the other hand, the origin of the nominative subject depends on the voice of the embedded infinitival clause: if the infinitival clause is active, the subject is an argument of the modal verb, but if it is passive, the subject originates as an internal argument of the infinitive; the schematic representations in (660b&b') show that we are dealing with a control structure in the former and with a subject raising structure in the latter case. In short, non-directed deontic modality is special in that it patterns with epistemic modality in requiring a subject raising analysis.

660
a. NPi Vdispositional [... PROi ... Vinf]
dynamic/dispositional
a'. NPi Vdispositional [... PROi ... Vpart worden]
dynamic/dispositional
b. NPi Vdeontic [... PROi Vinf]
directed deontic
b'. NPi Vdeontic [... ti ... Vpart worden]
non-directed deontic

      The discussion above provides us with a test to answer the question raised at the end of the previous subsection as to whether the verb kunnen can be used to express directed deontic modality. If so, we expect that example (661a) can be passivized, and that the resulting construction need not involve subject-oriented modality. This seems to be borne out, as example (661b) must be interpreted in such a way that the ability is ascribed to the implicit agent of the infinitival verb, and not to the inanimate subject dat boek'that book'. That (661a) and (661b) are, respectively, directed and non-directed deontic is also supported by the fact that pronominalization of the infinitival verbs affects the nominative subject of the construction as a whole in the latter case only; this shows that kunnen is dyadic in (661a) but monadic in (661b).

661
a. Jan kan dat boek nu ophalen.
subject-oriented modality
  Jan can  that book  now  prt.-fetch
  'Jan may fetch that book now (there is nothing to prevent it).'
a'. Jan kan dat nu /#Dat kan nu.
verb is dyadic
  Jan can  that  now    that  can  now
b. Dat boek kan nu opgehaald worden.
no subject-oriented modality
  that book  can  now  prt.-fetched  be
  'That book can now be fetched (there is nothing to prevent it).'
b'. Dat kan nu /*Dat boek kan dat nu.
verb is monadic
  that  can  now     that book  can  that  now

The fact that the moving force in non-directed deontic constructions is directed towards some entity other than the nominative subject also means that this type of modality differs from the other types of event modality in that the nominative subject need not be able to control the eventuality expressed by the infinitival clause. The examples in (662) show that, as a result, the infinitival clause can be a copular construction, or headed by an unaccusative/undative verb; all examples are taken from the internet.

662
a. Gebruik van geweld moet proportioneel zijn.
copular
  use of violence  must  proportional  be
  'Use of force must be proportional.'
a'. Die boete mag van mij wel wat hoger zijn.
  that fine  may  of me  prt  a.bit  higher  be
  'As far as Iʼm concerned, that fine can be a bit higher.'
b. Ingevroren vlees moet langzaam ontdooien.
unaccusative
  frozen meat  must  slowly  defrost
  'Frozen meat must be defrosted slowly.'
b'. Stoofvlees mag langzaam sudderen (zonder dat u ernaar om kijkt).
  stew  may  slowly  simmer  without  that you  to.it  after  look
  'Stew may simmer slowly (without you having to look after it).'
c. Het interieur moet nog een verfje krijgen.
undative
  the interior  must  still  a layer.of.paint  get
  'The interior must still be painted.'
c'. De muziek mag nooit de overhand krijgen (of de kijker irriteren).
  the music  may  never  the upper.hand  get   or  the viewer  annoy
  'The music should never get the upper hand (or annoy the viewer).'
[+]  C.  A binary feature analysis of modal verbs

The previous subsections discussed several types of modality that can be expressed by means of modal verbs taking a bare infinitival complement. Putting aside the option of analyzing perception verbs as verbs expressing evidential (sensory) modality, we concluded that there are four basic verb types, expressing, respectively, epistemic, dynamic/dispositional, directed deontic and non-directed deontic modality. Table (663) aims at providing a classification of these four types of modality by referring to the type of moving force involved. The feature ±external indicates whether or not there is an external moving force; if not, the moving force may be internal or be absent. The feature ±subject-oriented, which is adopted from Barbiers (1995), indicates whether the moving force is directed towards the nominative subject of the construction as a whole; if not, the moving force may be directed towards some other (implicit) entity or be absent.

663
Moving force and modality
[–subject-oriented] [+subject-oriented]
[–external] epistemic dynamic/dispositional
[+external] non-directed deontic directed deontic

The semantic classification in (663) is supported by syntactic/semantic evidence. First, the previous subsections have already shown that the feature ±subject-oriented affects the adicity of the modal verb, and thus determines whether we are dealing with control or subject raising constructions: epistemic and non-directed modal verbs are monadic and trigger subject raising; dynamic/dispositional and directed deontic modal verbs are dyadic and involve control. Second, the feature ±external reflects the fact that the two types of deontic modal verb license an adverbial van- or volgens-phrase which indicates the source of the moving force; such phrases are not possible (with the same meaning) in the case of epistemic and dynamic/dispositional modal verbs. Finally, the union of the +subject-oriented and +external modal verbs also forms a natural class in the sense that they normally involve a polarity transition (Barbiers 1995): the truth value of the proposition expressed by the infinitival clause can potentially be changed from false to true. This is illustrated in the examples in (664) which all involve the adverbial phrase of time nu'now'. The epistemic constructions in (664a) do not involve a polarity transition; they express the speaker's evaluation of the likelihood that the proposition expressed by the infinitival clause is true at the moment of speech. The remaining examples, on the other hand, all strongly suggest a truth transition: the proposition expressed by the infinitival clause is false at the moment of speech, but can be made true in the non-actualized part of the present-tense interval.

664
a. Jan moet/kan/zal nu het boek wel lezen.
epistemic
  Jan must/may/will  now  the book prt  read
  'Jan must/may/will read the book now.'
b. Jan moet/kan/wil het boek nu lezen.
dynamic/dispositional
  Jan  must/is.able/wants  the book  now  read
  'Jan must/can/wants to read the book now.'
c. Jan moet/mag het boek nu lezen.
directed deontic
  Jan  must/is.allowed  the book  now  read
  'Jan must/is allowed to read the book now.'
d. Dat boek moet/mag nu gelezen worden.
non-directed deontic
  that book  must/is.allowed  now  read  be
  'The book must/may be read now.'

This difference also accounts for the contrast between the examples in (665); the adverbial phrase gisteren'yesterday' situates the eventuality expressed by the infinitival clause in the actualized part of the present-tense interval, which causes the perfect-tense construction in (665a) to be interpreted as epistemically only. The adverbial phrase morgen'tomorrow' in (665b) situates the eventuality expressed by the infinitival clause in the non-actualized part of the present-tense interval, which causes the example to be four-ways ambiguous (where the preferred reading depends on contextual factors).

665
a. Jan moet dat boek gisteren hebben gelezen.
epistemic
  Jan must  that book  yesterday  have  read
  'Jan must have read that book yesterday.'
b. Jan moet dat boek morgen hebben gelezen.
four-ways ambiguous
  Jan must  that book  tomorrow  have  read
  'Jan must have read that book by tomorrow.'

For completeness' sake, note that Barbiers characterized the different modal types by appealing directly to the binary feature ±polarity transition. This seems less suited given that non-epistemic modal verbs only imply polarity transitions if the embedded verb is non-stative; cf. Erb (2001:81ff.). A speaker can readily express his assessment of Marie's dancing skills by means of (666a) at the very moment that he is watching her dancing. Similarly, the context of (666b) makes clear that the speaker is already waiting at the moment he utters the sentence Ik moet hier wachten. An example such as (666c) can readily be used when the speaker gives the addressee information about the switches of a specific machine. The fact that the occurrence of a polarity transition also depends on the infinitive makes it less suitable as a defining property of the basic modal types we have distinguished.

666
a. Marie kan goed dansen.
dynamic/dispositional
  Marie  can  well  dance
  'Marie dances well.'
b. Waarom sta je daar? Ik moet hier wachten.
directed deontic
  why  stand  you  there  must  here  wait
  'What are you standing here for? Iʼm supposed to wait here.'
c. Deze schakelaar moet altijd zo staan.
non-directed deontic
  this switch  must  always  like.that  stand
  'This switch must always be in this position.'
References:
  • Barbiers, Sjef1995The syntax of interpretationThe Hague, Holland Academic GraphicsUniversity of Leiden/HILThesis
  • Barbiers, Sjef1995The syntax of interpretationThe Hague, Holland Academic GraphicsUniversity of Leiden/HILThesis
  • Barbiers, Sjef1995The syntax of interpretationThe Hague, Holland Academic GraphicsUniversity of Leiden/HILThesis
  • Barbiers, Sjef1995The syntax of interpretationThe Hague, Holland Academic GraphicsUniversity of Leiden/HILThesis
  • Barbiers, Sjef1995The syntax of interpretationThe Hague, Holland Academic GraphicsUniversity of Leiden/HILThesis
  • Barbiers, Sjef1995The syntax of interpretationThe Hague, Holland Academic GraphicsUniversity of Leiden/HILThesis
  • Barbiers, Sjef1995The syntax of interpretationThe Hague, Holland Academic GraphicsUniversity of Leiden/HILThesis
  • Erb, Marie Christine2001Finite auxiliaries in GermanTilburgUniversity of TilburgThesis
  • Erb, Marie Christine2001Finite auxiliaries in GermanTilburgUniversity of TilburgThesis
  • Haeseryn, Walter, Romijn, Kirsten, Geerts, Guido, Rooij, Jaap de & Toorn, Maarten C. van den1997Algemene Nederlandse spraakkunstGroningenNijhoff
  • Klooster, Wim1986Problemen met complementenTabu16122-132
  • Klooster, Wim1986Problemen met complementenTabu16122-132
  • Klooster, Wim1986Problemen met complementenTabu16122-132
  • Klooster, Wim1986Problemen met complementenTabu16122-132
  • Palmer, F.R2001Mood and ModalityCambridge University Press
  • Palmer, F.R2001Mood and ModalityCambridge University Press
  • Palmer, F.R2001Mood and ModalityCambridge University Press
  • Palmer, F.R2001Mood and ModalityCambridge University Press
  • Palmer, F.R2001Mood and ModalityCambridge University Press
  • Palmer, F.R2001Mood and ModalityCambridge University Press
  • Palmer, F.R2001Mood and ModalityCambridge University Press
  • Vliegen, Maurice2011Evidentiality. Dutch <i>seem </i>and <i>appear </i>verbs: <i>blijken</i>, <i>lijken</i>, <i>schijnen</i>Nouwen, Rick & Elenbaas, Marion (eds.)Linguistics in the Netherlands 2011Amsterdam/Philadelphia125-137
report errorprintcite