• Dutch
  • Frisian
  • Afrikaans
Show full table of contents
13.4. Weak proform shift
quickinfo

Weak (phonetically reduced) proforms normally occur in the left periphery of the middle field of the clause, with the exception of weak subject pronouns, which may also occur in clause-initial position; cf. Section 9.3. We can distinguish the three groups of weak elements in (173), all of which have strong counterparts with the exception of expletive and partitive er.

Example 173
Weak proforms
a. Referential personal pronouns; ie/ ze'he/she', ʼ m/ ʼr'him/her', etc.
b. Reflexive personal pronouns: me'myself', je'yourself', zich'him/herself', etc.
c. the R-word er: expletive, locational, prepositional and quantitative

The set of elements in (173) closely resembles the set of clitics found in French: see the lemma French personal pronouns at Wikipedia for a brief review. We will see that the relative order of the weak proforms also exhibits a number of similarities with the French clitics, which may justify the claim that the Dutch weak proforms are clitics as well; see Huybregts (1991), Zwart (1993/1996) as well as Haegeman (1993a/1993b) on W-Flemish. It should be noted, however, that the Dutch proforms differ from French clitics in that they do not need a verbal host: while the French clitics always cluster around a main or an auxiliary verb, the Dutch proforms do not require this. In order to not bias the discussion beforehand, we will refer to the movement that places weak proforms in the left periphery of the middle field as weak proform shift. Subsection I starts with a discussion of the weak referential personal pronouns, Subsection II discusses the weak (simplex) reflexive pronouns, and Subsection III concludes with the various uses of the weak R-word er.

readmore
[+]  I.  Referential personal pronouns

Table 2 shows the classification of referential personal pronouns, which is more extensively discussed in Section N5.2.1. The discussion in this subsection focuses on the distribution of the weakforms.

Table 2: Referential personal pronouns
  singular plural
  subject object subject object
  strong weak strong weak strong weak strong weak
1st person ik ’k mij me wij we ons
2nd person regular jij je jou je jullie jullie
  polite u u u u
3rd person masculine hij -ie hem ’m zij ze henacc
hundat
ze
  feminine zij ze haar (d)’r        
  neuter ?het ’t *?het ’t        

[+]  A.  Weak pronominal subjects of (in)transitive verbs

In embedded clauses weak subject pronouns are right-adjacent to the complementizer (if present), and immediately precede or follow the finite verb in second position in main clauses; cf. Paardekooper (1961). This is illustrated in (174) by means of the 3rd person singular feminine pronoun ze'her'.

Example 174
a. dat ze waarschijnlijk morgen komt.
embedded clause
  that  she  probably  tomorrow  comes
  'that sheʼs probably coming tomorrow.'
b. Ze komt waarschijnlijk morgen.
subject-initial main clause
  she  comes  probably  tomorrow
  'Sheʼs probably coming tomorrow.'
b'. Waarschijnlijk komt ze morgen.
other main clauses
  probably  comes  she  tomorrow
  'Probably sheʼs coming tomorrow.'

The examples in (175) show that subject pronouns can occur in positions more to the right only if they are strong and carry contrastive focus accent. The question mark in example (175b) is used to indicate that even then strengthening of the pronoun by means of a focus particle is often preferred.

Example 175
a. * dat waarschijnlijk ze morgen komt.
  that  probably  she  tomorrow  comes
b. dat waarschijnlijk ?(zelfs) zij morgen komt.
  that  probably    even she  tomorrow  comes
  'that even she is probably coming tomorrow.'

The examples in (176) show that the singular third person masculine subject pronoun ie'he' is exceptional in that it cannot occur in clause-initial position: it is a truly enclitic pronoun in that it obligatorily follows the complementizer or the finite verb in second position.

Example 176
a. dat-ie waarschijnlijk morgen komt.
embedded clause
  that-he  probably  tomorrow  comes
  'that heʼs probably coming tomorrow.'
b. Hij/*Ie komt waarschijnlijk morgen.
subject-initial main clause
  he/he  comes  probably  tomorrow
  'Heʼs probably coming tomorrow.'
b'. Waarschijnlijk komt-ie morgen.
other main clauses
  probably  comes-he  tomorrow
  'Probably heʼs coming tomorrow.'

Example (177) shows that weak subject pronouns differ conspicuously from weak object pronouns in that the latter cannot occur in sentence-initial position.

Example 177
a. Gisteren heeft Jan het boek/ʼt gelezen.
  yesterday  has  Jan the book/it  read
  'Yesterday Jan read the book/it.'
b. Het boek/*ʼt heeft Jan gisteren gelezen.
  the book/it  has  Jan yesterday  read

This fact motivated the claim in (178) that subject-initial sentences are not CPs but TPs, as this hypothesis makes it possible to maintain the generalization that weak pronouns cannot be topicalized, that is, wh-moved into the specifier of CP; we refer the reader to Section 9.3 for detailed discussion.

Example 178
a. Subject-initial sentence
b. Other main clauses
[+]  B.  Weak pronominal objects

This subsection discusses weak proform shift of object pronouns. In succession, we will address the placement of weak object pronouns with respect to subjects, the relative order of weak direct and indirect object pronouns, and the relative order of weak object pronouns with respect to accusative subjects of AcI-constructions.

[+]  1.  Order of subject and weak object pronouns in main clauses

Example (177) above has already shown that weak object pronouns cannot occur in sentence-initial position but must occupy a position in the middle field of the clause. The examples in (179) further show that they immediately follow the subject if it is not in sentence-initial position; cf. Huybregts (1991). This does not only hold if the subject is in the regular subject position, as in the primeless examples, but also if it is contrastively focused and can be assumed to be located in the specifier of FocP lower in the clause, as in the primed examples; cf. Section 13.3.2.

Example 179
a. dat <*ʼt> Jan/ie <ʼt> waarschijnlijk <*ʼt> niet gelezen heeft
  that       it  Jan/he  probably  not  read  has
  'that Jan/he probably hasnʼt read it.'
a'. dat <*ʼt> waarschijnlijk zelfs Jan <ʼt> niet gelezen heeft.
  that       it  probably  even Jan  not  read  has
  'that even Jan probably hasnʼt read it.'
b. dat <*ʼm> Marie/ze <ʼm> waarschijnlijk <*ʼm> goede raad wil geven.
  that     him  Marie/she  probably  good advice  wants  give
  'that Marie/she probably wants to give him good advice.'
b'. dat <*ʼm> waarschijnlijk zelfs Marie <ʼm> goede raad wil geven.
  that     him  probably  even Marie  good advice  wants  give
  'that even Marie probably wants to give him good advice.'

In subject-initial main clauses, weak object pronouns immediately follow the finite verb in second position. This is illustrated in (180) by showing that modal adverbs cannot precede the object pronoun but this holds for other constituents as well.

Example 180
a. Jan heeft <ʼt> waarschijnlijk <*ʼt> niet gelezen.
  Jan has    it  probably  not  read
  'Jan probably hasnʼt read it.'
b. Marie wil <ʼm> waarschijnlijk <*ʼm> goede raad geven.
  Marie wants    him  probably  good advice  give
  'Marie probably wants to give him good advice.'
[+]  2.  Order of direct and indirect object

The previous subsection has shown that weak proform shift cannot affect the unmarked order of the subject and the objects. This is different when it comes to the relative order of direct and indirect objects: while direct objects normally follow nominal indirect objects under a neutral intonation pattern, weak pronominal direct objects normally precede indirect objects. Example (181b) shows that this holds regardless of whether the indirect object is non-pronominal or pronominal. It should further be noted that it also holds if the two object pronouns have the same form: the first object pronoun in dat Peter ʼm ʼm aanbood'that Peter offered it to him' is construed as the direct object.

Example 181
a. dat Peter <*de auto> Marie <de auto> aanbood.
  that  Peter      the car  Marie     the car  prt.-offered
  'that Peter offered Marie the car.'
b. dat Peter <ʼm> Marie/ʼr <??ʼm> aanbood.
  that  Peter    him  Marie/her  prt.-offered
  'that Peter offered it to Marie/her.'

Weak objects pronouns are always adjacent to each other, which may be due to the fact illustrated in the previous subsection that they must both be adjacent to the finite verb or the subject if it is not in clause-initial position; the only new thing is that this restriction does not hold for the individual pronouns but for the full cluster. It should also be noted that Haegeman (1993a) observes for W-Flemish that inversion of the indirect and direct object requires the indirect object to be scrambled. Example (182) shows that the same seems to hold in Dutch, although it should be noted that the degraded order improves if the indirect object is assigned contrastive accent.

Example 182
dat Jan ʼt <Marie> waarschijnlijk <*?Marie> gegeven heeft.
  that  Jan  it    Marie  probably  given  has
'that Jan has probably given it to Marie.'

      This reversal of the direct and the indirect objects is possible only with reduced direct objects. It is not easy, however, to demonstrate reversal for strong referential personal pronouns because they cannot be used to refer to inanimate entities. The examples in (183) therefore illustrate this reversal by means of the demonstrative die'that one'; the judgments only hold under a non-contrastive intonation pattern.

Example 183
a. dat Peter <??die> Marie <die> aanbood.
  that  Peter      dem  Marie  prt.-offered
  'that Peter offered Marie that one.'
b. dat Peter <*die> ʼr <die> aanbood.
  that  Peter     dem  her  prt.-offered
  'that Peter offered her that one.'

The fact that object pronouns can be inverted while non-pronominal nominal arguments cannot has given rise to the hypothesis that they do not occupy the same position in the middle field of the clause, since only weak pronouns undergo weak proform shift; cf. Zwart (1996). If we assume in addition that weak proform shift is similar to clitic movement in languages like French, this hypothesis can be supported by the fact that third person direct and indirect object clitics in French appear in the same order as in Dutch: Jean leDO luiIO donnera'Jean will give it to him/her'. The fact discussed earlier that weak object pronouns cluster provides additional support to the hypothesis that they are clitic-like.

[+]  3.  Order of subject and object in AcI-constructions

Subjects and direct objects of infinitival complement clauses in AcI-constructions are indistinguishable as far as their morphological form is concerned: this holds not only for referential noun phrases but also for their pronominalized counterparts, as both appear as object pronouns. Nevertheless, the examples in (184a&b) show that weak proform shift of an embedded object can optionally cross the subject of the infinitival clause; cf. Zwart (1996). Example (184c) shows that this is in fact the preferred option if the subject is also realized as a weak pronoun. The acceptability of inversion shows that the restriction established above, namely that weak proform shift of objects cannot affect the unmarked order of subjects and objects, only holds if the subject is assigned nominative case.

Example 184
a. Jan zag/liet <*het boektheme> Marieagent < het boektheme> lezen.
  Jan saw/let      the book  Marie  read
  'Jan saw/let Marie read the book.'
b. Jan zag/liet <ʼttheme> Marieagent <ʼttheme> lezen.
  Jan saw/let      it  Marie  read
c. Jan zag/liet <ʼttheme> ʼragent <??ʼttheme> lezen.
  Jan saw/let      it  her  read

The examples in (185) show that weak proform shift of an embedded direct object may also cross the subject if the infinitival clause is ditransitive: the direct object pronoun must cross the indirect object and optionally crosses the embedded subject.

Example 185
a. Jan zag/liet <*het boektheme> Elsagent Petergoal <het boektheme> aanbieden.
  Jan saw/let      the book  Els  Peter  prt. offer
  'Jan saw/let Els offer Peter the book.'
b. Jan zag/liet <ʼttheme> Elsagent <ʼttheme> Petergoal <??ttheme> aanbieden.
  Jan saw/let     it  Els  Peter  prt.-offer

It seems, however that weak embedded indirect object pronouns cannot cross the subject of the infinitival clause: according to us, (186b) can only be interpreted with the pronoun as an agent and Els as a goal. It seems plausible that the deviance of (186b) is related to the fact that the agent and the goal are both +human.

Example 186
a. Jan zag/liet <*Petergoal> Elsagent <Petergoal> het boektheme aanbieden.
  Jan saw/let      Peter  Els the book  prt. offer
  'Jan saw/let Els offer Peter the book.'
b. * Jan zag/liet <ʼmgoal> Elsagent <ʼmgoal> het boektheme aanbieden.
  Jan saw/let     him  Els the book  prt. offer
  'Jan saw/let Els offer him the book.'

Something similar holds for cases in which both the direct and the indirect object surface as weak pronouns: examples such as (187b), which are given as fully acceptable in Zwart (1993/1996), are only acceptable to us if the pronoun ʼm is interpreted as agent and Els as goal. The unacceptability of (187c) deserves special mention as it is unexpected in the light of the fact that (184c) is fully acceptable; the fact that weak object pronouns must be adjacent to each other again provides support to the hypothesis that they are clitic-like in that they obligatorily cluster.

Example 187
a. Jan zag/liet Elsagent ʼttheme ʼmgoal aanbieden.
  Jan saw/let  Els  it  him  prt. offer
  'Jan saw/let Els offer it to him.'
b. * Jan zag/liet ʼttheme ʼmgoal Elsagent aanbieden.
  Jan saw/let  it  him  Els prt. offer
c. * Jan zag/liet ʼttheme Elsagent ʼmgoal aanbieden.
  Jan saw/let  it  Els  him  prt. offer

Example (188a) shows that if all the arguments of a ditransitive infinitival clause surface as weak pronouns they must occur in the order agent > theme > goal. It should be pointed out, however, that some speakers find a sequence of three weak pronouns difficult to pronounce and may therefore prefer the version in (188b) with a prepositional indirect object; in such cases the theme again preferably precedes the agent.

Example 188
a. Jan zag/liet ʼragent ʼttheme ʼmgoal aanbieden.
  Jan saw/let  her it  him  prt. offer
  'Jan saw/let her offer him the book.'
b. Jan zag/liet <ʼttheme> ʼragent <ʼttheme> aan ʼmgoal aanbieden.
  Jan saw/let    it her her  to him  prt. offer
  'Jan saw/let her offer it to him.'

      Example (189) suggests that weak proform shift is able to feed binding: while non-pronominal direct objects cannot bind a reciprocal indirect object, shifted direct object pronouns can.

Example 189
dat Marie <zetheme> elkaargoal <*de jongenstheme> voorgesteld heeft.
  that  Marie  them  each.other      the boys  prt.-introduced  has
'that Marie has introduced them to each other.'

Since feeding of binding is generally seen as a hallmark of A-movement, this may also suggest that weak proform shift is A-movement. It should be noted, however, that weak proform shift may be preceded by nominal argument shift and that it may be the case that this is responsible for feeding binding; cf. Haegeman (1993a/1993b). We provisionally assume that weak proform shift of arguments is A'-movement because Subsection III will show that weak proforms that do not function as arguments may undergo a similar shift.

[+]  C.  Pronominal subjects of passive and unaccusative constructions

Section 2.1.2 has shown that derived subjects can either precede or follow an indirect object; this is illustrated again in (190a) by means of the passivized counterpart of the ditransitive construction dat Jan Peter/ʼm de baan aanbood'that Jan offered Peter/him the job'. Example (190b) shows that the weak subject pronoun must precede the indirect object (which is not very surprising because strong subject pronouns are obligatorily moved into the regular subject position by nominal argument object shift; see 13.2, sub IB). The examples in (191) show the same for the dyadic unaccusative (nom-dat) verb bevallen'to please'.

Example 190
a. dat <de baan> Peter/ʼm <de baan> aangeboden werd.
  that  the job  Peter/him  prt.-offered  was
  'that the job was offered to Peter/him.'
b. dat <ie> Peter/ʼm <*ie> aangeboden werd.
  that   he  Peter/him     he  prt.-offered  was
  'that it was offered to Peter/him.'
Example 191
a. dat <de film> Peter/ʼm <de film> bevallen is.
  that  the movie  Peter/him  pleased  is
  'that the movie has pleased Peter/him.'
b. dat <ie> Peter/ʼm <*ie> bevallen is.
  that    he  Peter/him     he  pleased  is
  'that it has pleased Peter/him.'
[+]  4.  Conclusion

The previous subsections have shown that weak object pronouns cannot be moved across nominative subjects. At first sight, this would suggest that weak proform shift cannot affect the unmarked order of nominal arguments (agent > goal > theme), but this turns out not to be correct, as is clear from the fact that weak direct object pronouns preferably precede nominal indirect objects, and that they can also be moved across an embedded subject in an AcI-construction. That weak proform shift can affect the unmarked order of nominal argument shows that weak pronouns can occupy positions in the clause that are not accessible to their non-pronominal counterparts, which in turn gives credence to the hypothesis that they are clitic-like. This hypothesis is further supported by the fact that weak object pronouns obligatorily cluster together.
      Weak subject and object pronouns exhibit various properties often attributed to clitics. The examples in (192) show, for example, that they cannot be used as independent utterances and cannot be topicalized or coordinated; see Haegeman (1993b) for relevant discussion. Zwart (1996) notes, however, that these properties also hold for the English reduced pronouns, which behave syntactically as regular pronouns, and concludes from this that they are not defining characteristics of clitics but simply follow from the fact that weak pronouns cannot be accented.

Example 192
a. Wie heb je gezien? Hem/*ʼm.
  who  have  you  seen  him/him
b. Hem/*ʼm heb ik niet gezien.
  him/him  have  not  seen
  'Him, I havenʼt seen.'
c. [hem en haar]/ *[ʼm en ʼr]
  him  and  her     him  and  her

A potential problem for the claim that Dutch weak pronouns are clitics is that they differ from run-of-the-mill clitics in that they are not hosted by a verb. A related problem is that they can occur in PPs: bij ʼm'with him'; cf. Haegeman (1993b). The hypothesis that Dutch weak pronouns are clitics thus requires there to be some (phonetically empty) functional head that they can cliticize to. Currently, there does not seem to be a generally accepted analysis available but the tentative proposals in Haegeman (1993a/1993b) and Zwart (1993/1996) do agree on the fact that the prospective functional head(s) have nominal (case or agreement) features. We leave this claim for future research.

[+]  II.  The simplex reflexive zich

Section N5.2.1.5 has shown that Dutch has two types of reflexive pronouns: simplex reflexive pronouns such as third person zich and complex ones such as third person zichzelf'him/herself/themselves'. Simplex reflexive pronouns differ from complex ones in that they must precede modal adverbs such as waarschijnlijk'probably'; cf. Huybregts (1991). We refer the reader to N5.2.1.5 for a more extensive discussion of these two forms.

Example 193
a. Marie heeft <zichzelf> waarschijnlijk <zichzelf> aan Jan voorgesteld.
  Marie has    herself  probably  to Jan  prt.-introduced
  'Marie has probably introduced herself to Jan'
b. Marie heeft <zich> waarschijnlijk <*zich> voorgesteld aan Jan.
  Marie has    refl  probably  prt.-introduced  to Jan
  'Marie has probably introduced herself to Jan.'

Simplex reflexive pronouns behave like object pronouns in that they cannot precede subject pronouns. We illustrate this in (194) by means of a number of strong singular referential personal pronouns; the judgments do not change if we replace the strong subject pronouns by their weak counterparts.

Example 194
a. dat <*me> ik <me> nog niet heb voorgesteld.
  that    refl  yet  not  have  prt.-introduced
  'that I havenʼt introduced myself yet.'
b. dat <*je> jij <je> nog niet hebt voorgesteld.
  that   refl  you  yet  not  have  prt.-introduced
  'that you havenʼt introduced yourself yet.'
c. dat <*zich> zij <zich> nog niet heeft voorgesteld.
  that     refl  she  yet  not  has  prt.-introduced
  'that she hasnʼt introduced herself yet.'

Simplex reflexive pronouns are special, however, in that they normally precede non-specific indefinite and negative subject pronouns, which we illustrate in (195) by means of expletive- there constructions; cf. Haeseryn et al. (1997:1314). This means that they differ from object pronouns, which can never be moved across the subject of their clause but instead push the subject up into the regular subject position: see Section 13.2, sub IC1, for discussion.

Example 195
a. dat er <zich> drie vaten bier <*zich> in de kelder bevinden.
  that  there    refl  three barrels [of] beer  in the cellar are.located
  'There are three barrels of beer in the cellar.'
b. dat er <zich> een meisje <*zich> in de kelder opgehangen heeft.
  that  there    refl  a girl  in the cellar  prt.-hanged  has
  'that a girl has hanged herself in the cellar.'

With respect to specific indefinite and generic subject pronouns simplex reflexive pronouns again behave like object pronouns in that they follow them; cf. Haeseryn et al. (1997:1314).

Example 196
a. dat <*zich> een vriendin van hem <zich> in de kelder opgehangen heeft.
  that    refl  a friend of him  in the cellar  prt.-hanged  has
  'that a lady friend of his has hanged herself in the cellar.'
b. dat <*zich> een puber <zich> nu eenmaal zo gedraagt.
  that    refl  an adolescent  prt  prt  like.that  behaves
  'that an adolescent will behave like that.'

The ordering with respect to definite subjects seems to be relatively free, as is clear from example (197b). The placement of the subject in this example seems to be determined by the information structure of the clause: it follows the reflexive if it is part of the new information of the clause, while it precedes the reflexive if it is part of the presupposition; cf. Haeseryn et al. (1997:1315).

Example 197
a. dat er zich hier een drama heeft afgespeeld.
  that  there  refl  here  a tragedy  has  prt.-played
  'that a tragedy took place here.'
b. dat <dat drama> zich hier <dat drama> afgespeeld heeft.
  that that tragedy refl  here  prt.-played  has
  'that that tragedy took place here.'

This is consistent with the observation in Haeseryn et al. that the order reflexive–subject is found especially with inherently reflexive predicates that denote a process of appearing or coming into existence. Some examples are given in (198).

Example 198
a. In de verte verhieven zich de Alpen.
  in the distance  rose  refl  the Alps
  'In the distance rose the Alps.'
b. Er dienen zich twee problemen aan.
  there  present  refl  two problems  prt.
  'Two problems present themselves.'
c. Er tekende zich een kleine meerderheid af.
  there  silhouetted  refl  a small majority  prt.
  'A small majority became apparent.'

The ordering vis-a-vis negative subjects also has a semantic effect: while (199) expresses that there are no registrations at all, example (199b) does not necessarily imply this but may also be used to express that no individual from a contextually defined set has registered; cf. Haeseryn et al. (1997:1315).

Example 199
a. dat zich nog niemand heeft aangemeld.
  that  refl  yet  nobody  has  prt.-registered
  'that nobody has registered yet.'
b. dat niemand zich nog heeft aangemeld.
  that   nobody  refl  yet  has  prt.-registered
  'that nobody has registered yet.'

In some cases it is virtually impossible for the reflexive pronoun to precede the subject; in (200) the simplex reflexive must follow the negative subject even though this means that it cannot be shifted across the modal adverb; cf. (193b). The acceptability contrast indicated is confirmed by the fact that a Google search (7/14/2015) on the string [ zich n iemand herinnert] resulted in no more than five relevant examples from the 19th century, while the alternative order resulted in 48 hits. It is not yet clear what determines precisely whether the order reflexive pronoun–subject is possible or not, although it is conspicuous that all examples given in Haeseryn et al. (1997) involve intransitive inherently reflexive verbs.

Example 200
a. dat (waarschijnlijk) niemand zich die man herinnert.
  that   probably  nobody  refl  that man  remembers
  'that probably nobody remembers that man.'
b. ?? dat zich (waarschijnlijk) niemand die man herinnert.
  that  refl   probably  nobody  that man  remembers

In transitive constructions, the relative order of weak object and simplex reflexive pronouns seems to be relatively free, although there is a clear preference for the former to precede the latter. We checked this for the pronoun het'it', which is virtually always weak in speech, by doing a Google search (7/14/2015) on the search strings [ het zich (niet) herinner t] and [ zich het (niet) herinner t].

Example 201
a. dat Jan ʼt zich (niet) herinnert.
207 hits
  that  Jan  it  refl  not  remembers
  'that Jan remembers it/that Jan doesnʼt remember it.'
b. dat Jan zich ʼt (niet) herinnert.
36 hits
  that  Jan  refl  it   not  remembers
  'that Jan remembers it/that Jan doesnʼt remember it.'

For completeness’ sake, it should be noted that the preferred Dutch order differs from the one found in French, where the reflexive clitic precedes the object clitic: cf. Il se le rappelle'He remembers it'.

[+]  III.  The weak R-word er

The phonetically weak R-word er has the four distinctive functions illustrated in (202). Expletive er normally introduces some indefinite subject (cf. N8.1.4) but also occurs in impersonal passives (cf. 3.2.1.2), locational er refers to some contextually defined location, prepositional er represents the nominal part of a pronominalized PP (cf. P5), and quantitative er is associated with an interpretative gap [e] in a quantified noun phrase (cf. N6.3). Sometimes a single occurrence of er expresses more than one function, but this will be ignored here; see Section P5.5 for extensive discussion.

Example 202
a. dat <er> waarschijnlijk <*er> iemand ziek is.
expletive
  that  there  probably  someone  ill  is
  'that there is probably someone ill.'
b. dat Jan <er> waarschijnlijk <*er> geweest is.
locational
  that  Jan  there  probably  been  is
  'that Jan has probably been there.'
c. dat Jan <er> waarschijnlijk <?er> over wil praten.
prepositional
  that  Jan there  probably  about  wants  talk
  'that Jan probably wants to talk about it.'
d. dat Jan <eri> waarschijnlijk <*eri> [twee/veel [ei]] heeft.
quantitative
  that  Jan there  probably   two/many  has
  'that Jan probably has two/many of them.'

This subsection will focus on the distribution of the various types within the clause. The examples in (202) already show that all types resemble weak pronouns in that they normally precede modal adverbs such as waarschijnlijk'probably'. Details concerning their placement will be discussed in separate subsections.

[+]  A.  Expletive er

The distribution of expletive er is identical to that of (weak) subject pronouns: in main clauses it immediately precedes or follows the finite verb and in embedded clauses it immediately follows the complementizer (if overtly realized). It is therefore not surprising that it is often assumed that expletive er is located in the regular subject position, that is, the specifier of TP. Putting aside cases in which expletive er occupies the sentence-initial position, this correctly predicts that it is always the leftmost element in the middle field of the clause.

Example 203
a. Er komt morgen waarschijnlijk een vriend van hem op visite.
  there  comes  tomorrow  probably  a friend of his  on visit
  'There is probably a friend of his coming to visit us tomorrow.'
a'. Morgen komt er waarschijnlijk een vriend van hem op visite.
  tomorrow  comes  there  probably  a friend of his  on visit
  'Tomorrow there is probably a friend of his coming to visit us.'
b. dat er morgen waarschijnlijk een vriend van hem op visite komt.
  that  there  tomorrow  probably  a friend of his  on visit  comes
  'that there is probably a friend of his coming to visit us tomorrow.'
[+]  B.  Locational er

Locational er differs from other locational proforms in that it must precede the modal adverbs. The (a)-examples in (204) illustrate this for an adverbial phrase, and the (b)-examples for a complementive. Observe that the locational R-word daar can also be moved across the modal adverb; we return to this in Subsection C.

Example 204
a. dat Jan waarschijnlijk in de speeltuin speelt.
adverbial
  that  Jan probably  in the playground  plays
  'that Jan is probably playing in the playground.'
a'. dat Jan <daar/er> waarschijnlijk <daar/*er> speelt.
  that  Jan there/there  probably  plays
  'that Jan is probably playing there.'
b. dat Jan waarschijnlijk in de speeltuin geweest is.
complementive
  that  Jan probably  in the playground  been  is
  'that Jan has probably been in the playground.'
b'. dat Jan <daar/er> waarschijnlijk <daar/*er> geweest is.
  that  Jan there/there  probably  been  is
  'that Jan has probably been there.'

The examples in (205) show that location er resembles the French locative clitic y in that it follows weak object pronouns: cf. Je les y ai vus'I have seen them there'.

Example 205
a. dat ik ze er gezien heb.
  that  them  there  seen have
  'that I have seen them there.'
b. * dat ik er ze gezien heb.
  that  there  them  seen  have
[+]  C.  Prepositional er

Pronominal PPs functioning as an argument of the verb can be split; movement of heavier R-words such as daar is optional, while movement of the weak form er is greatly preferred. The two parts of the pronominal PP are in italics.

Example 206
a. dat Jan waarschijnlijk over dat probleem wil praten.
  that  Jan probably  about that problem  wants  talk
  'that Jan probably wants to talk about that problem.'
b. dat Jan <daar> waarschijnlijk [PP <daar> over] wil praten.
  that  Jan     there  probably  about  wants  talk
  'that Jan probably wants to talk about that.'
c. dat Jan <er> waarschijnlijk [PP <?er> over] wil praten.
  that  Jan there  probably  about  wants  talk
  'that Jan probably wants to talk about it.'

The fact that daar and er can both be moved leftward, which was also observed in the previous subsection for locational proforms, can perhaps be taken as evidence against the claim that er is clitic-like by assuming that the ability to undergo leftward movement is simply a more general property of R-words. Indeed, it has been suggested that there is a designated [+R]-position in the functional domain of the clause that serves as a landing site for R-words; cf. Van Riemsdijk (1978). The examples in (207) show, however, that it is possible to shift two R-words in a single clause as long as they are not both weak or both strong.

Example 207
a. dat Jan er hier waarschijnlijk niet over wil praten.
  that  Jan  there  here  probably  not  about  wants talk
  'that Jan probably doesnʼt want to talk about it here.'
b. * dat Jan er er waarschijnlijk niet over wil praten
  that  Jan  there  there  probably  not  about  wants talk
  'that Jan probably doesnʼt want to talk about it here.'
c. ?? dat Jan daar hier waarschijnlijk niet over wil praten.
  that  Jan  there  here  probably  not  about  wants  talk
  'that Jan probably doesnʼt want to talk about it here.'

Huybregts (1991) concluded from this that there are actually two [+R]-positions, one of which is accessible to weak R-words only. If correct, this shows that it is possible to identify a designated position for the weak R-word er after all, as required by the hypothesis that er is clitic-like. We will not digress on this here, but refer to reader to Section P5.5 for a detailed discussion of Huybregts’ proposal.
      The examples in (208) show that while prepositional er is able to precede non-pronominal objects, it must follow weak object pronouns.

Example 208
a. Jan heeft zijn kinderen tegen ongewenste invloeden beschermd.
  Jan has  his children  against  undesirable influences  protected
  'Jan has protected his children against undesirable influences.'
a'. Jan heeft < er> zijn kinderen < er> tegen beschermd.
  Jan has  there his children  against  protected
  'Jan has protected his children against them.'
a''. Jan heeft <* er> ze < er> tegen beschermd.
  Jan has  there  them   against protected
b. Marie heeft Peter tot diefstal gedwongen.
  Marie has  Peter  to  theft  forced
  'Marie has forced Peter to steal.'
b'. Marie heeft < er> Peter <er> toe gedwongen.
  Marie has  there  Peter  to  forced
  'Marie has forced Peter to do it.'
b''. Marie heeft <*er> ʼm <er> toe gedwongen.
  Marie has  there  him  to  forced
[+]  D.  Quantitative er

Quantitative er is associated with an interpretative gap within a quantified nominal argument which can be filled in on the basis of contextual information. While Peter is looking for a pan, the speaker may tell him how to obtain one by means of the utterances in (209a&b). Example (209c) likewise implies that there is a contextually defined set of individuals (say, students) who are given a book.

Example 209
a. Er staan eri waarschijnlijk [NP twee [ei]] in de keuken.
subject
  there  stand  there  probably  two  in the kitchen
  'There are probably two [pans] in the kitchen.'
b. Jan heeft eri waarschijnlijk [NP drie [ei]] op tafel gezet.
direct object
  Jan has  there  probably  three  on table  put
  'Jan has put three [pans] on the table.'
c. Jan gaf eri waarschijnlijk [NP één [ei]] een boek.
indirect object
  Jan gave  there  probably  one  a book
  'Jan probably gave one [student] a book.'

The examples in (209) show that quantitative er is obligatorily placed in front of the modal adverb and follows the finite verb in subject-initial clauses. If the subject is located in the middle field. as in (210), quantitative er follows the subject even if the subject follows a modal adverb.

Example 210
a. dat Jan eri waarschijnlijk [NP één [ei]] heeft.
  that  Jan there  probably  one  has
  'that Jan probably has one.'
b. dat waarschijnlijk niemand eri [NP één [ei]] heeft.
  that  probably  nobody  there  one  has
  'that probably nobody has one.'

When we consider the relative order of quantitative er and weak object pronouns, at least three cases should be distinguished; this will be discussed in the following subsections.

[+]  1.  The associate noun phrase is a subject

If the associate of quantitative er is a subject, a weak direct object pronoun must follow the associate, and even then the result is somewhat marked, which we indicate here by means of a question mark. This is shown in (211b), on the basis of the clause dat vier studenten het boek gelezen hebben'that four students have read the book.'

Example 211
a. dat eri [NP vier [ei]] het boek gelezen hebben.
  that  there  four  the book  read  have
b. dat <*ʼt> eri <*ʼt> [NP vier [ei]] < ?ʼt> gelezen hebben.
  that  it  there  four  read  have

Example (212b) shows that the same holds for weak indirect object pronouns, on the basis of the clause dat twee studenten Peter het boek aangeboden hebben'that two students have offered Peter the book'. For completeness’ sake, the (c)-examples show that the direct and indirect pronouns must appear after the associate; although the primeless (c)-example is somewhat marked itself, the contrast with the primed ones is quite sharp.

Example 212
a. dat eri [NP twee [ei]] Peter het boek aangeboden hebben.
  that  there  two  Peter  the book  prt.-offered  have
b. dat eri <*ʼm> [NP twee [ei]] < ?ʼm> het boek aangeboden hebben.
  that  there     him  two  the book  prt.-offered  have
c. ? dat eri [NP twee [ei]] ʼt ʼm aangeboden hebben.
  that  there  two  it  him  prt.-offered  have
c'. * dat eri ʼt [NP twee [ei]] ʼm aangeboden hebben.
  that  there  it  two  him  prt.-offered  have
c''. * dat eri ʼt ʼm [NP twee [ei]] aangeboden hebben.
  that  there  it  him two  prt.-offered  have
[+]  2.  The associate noun phrase is a direct object

Example (213a) shows that quantitative er may either precede or follow the indirect object. This is not entirely optional, however, as the (b)-examples bear out that the choice is partly determined by the surface position of the indirect object. Example (213b) shows that if the indirect object surfaces after the modal verb, the shift of quantitative er is indeed optional, although it should be noted that the shift must cross the modal adverb. Example (213c) shows that if the indirect object has undergone nominal argument shift, weak proform shift must apply as well although it may end up either preceding or following the indirect object.

Example 213
a. Marie heeft <eri> Jan <eri> [NP één [ei]] gegeven.
  Marie has  there  Jan  one  given
  'Marie has given Jan one.'
b. Marie heeft <eri> waarschijnlijk <*er> Jan <eri> [NP één [ei]] gegeven.
  Marie has  there  probably  Jan  one  given
c. Marie heeft <eri> Jan <eri> waarschijnlijk <*eri> [NP één [ei]] gegeven.
  Marie has  there  Jan  probably  one  given

While the examples in (213) show that quantitative er may either precede or follow a non-pronominal indirect object, there may be a preference for it to follow weak indirect object pronouns, although Haeseryn et al. (1997:1321) take both orders to be fully acceptable; note that the /d/ in (214b) is a linking sound that is inserted to break the sequence of two schwa’s. If this preference is indeed significant, we should conclude that quantitative er behaves similarly in this respect to the French partitive clitic en: cf. Je lui en ai donné une'I have given him one'.

Example 214
a. Jan heeft <?eri> ʼm <eri> [NP één [ei]] gegeven.
  Jan has  there  him  one  given
  'Jan has given him one.'
b. Ik heb <?eri> ze <(d)eri> [NP een paar [ei]] gegeven.
  have   there  them  a couple  given
  'Iʼve given them a couple.'
[+]  3.  The associate noun phrase is an indirect object

It is hard to construct cases with a weak indirect object pronoun. It seems that the pronoun preferably precedes quantitative er. We illustrate this in (215) for the sentence dat ik twee studenten het boek heb aangeboden'that I have offered the book to two students'.

Example 215
a. dat ik eri [NP twee [ei]] het boek heb aangeboden
  that  there  two  the book  have  prt.-offered
b. ?? dat ik eri [NP twee [ei]] ʼt heb aangeboden.
  that  there  two  it  have  prt.-offered
b'. * dat ik eri ʼt [NP twee [ei]] heb aangeboden.
  that  there  it  two  have  prt.-offered
b''. ? dat ik ʼt eri [NP twee [ei]] heb aangeboden.
  that  it  there  two  have  prt.-offered
[+]  IV.  Conclusion

The previous subsections have shown that there are grounds for assuming that weak proforms are clitic-like. The first and foremost reason is that weak proforms are like clitics in that they cluster together. Furthermore, there are certain similarities in the relative order of weak proforms and, e.g., French clitics. This holds especially for weak object pronouns. First, weak proform shift inverts the order of third person indirect and direct objects, just like clitic placement in French. Second, weak object pronouns precede most other weak proforms, as do the object clitics in French. The only difference involves the reflexive forms: reflexive clitics precede object clitics while simplex reflexive zich tends to follow the weak object pronouns. Another reason not yet mentioned is that weak proform shift is clause-bound: it is never possible to move a weak proform out of its minimal finite clause (cf. Huybregts 1991). A conspicuous difference between clitics and weak proforms is that the former normally attach to a verbal host while the latter do not: with the exception of the simplex reflexive zich the Dutch proforms must follow the (nominative) subject. It should also be noted that the location of the subject is immaterial:

Example 216
a. dat Jan ʼt waarschijnlijk gekocht heeft.
  that  Jan it  probably  bought  has
  'that Jan has probably bought it.'
b. dat <*ʼt> waarschijnlijk Jan <ʼt> gekocht heeft
  that    it  probably  Jan  bought  has
  'that Jan has probably bought it.'
b'. dat <*ʼt> waarschijnlijk niemand <ʼt> gekocht heeft
  that    it  probably  nobody  bought  has
  'that probably nobody has bought it.'

If we adopt the conclusions from Section 13.2 and 13.3.1 that the subjects in the examples in (216) occupy different positions, we must conclude that there is no fixed target position for weak proform shift either, which may be a potential problem for claiming that weak proform shift and clitic placement are virtually the same operation. We leave this issue to future research.

References:
  • Haegeman, Liliane1993The morphology and distribution of object clitics in West FlemishStudia Linguistica4757-94
  • Haegeman, Liliane1993Some Speculations on Argument Shift, Clitics and Crossing in West-FlemishAbraham, Werner & Bayer, Josef (eds.)Dialektsyntax (Linguistische Berichte Sonderheft 6OpladenWestdeutscher Verlag131-160
  • Haegeman, Liliane1993The morphology and distribution of object clitics in West FlemishStudia Linguistica4757-94
  • Haegeman, Liliane1993The morphology and distribution of object clitics in West FlemishStudia Linguistica4757-94
  • Haegeman, Liliane1993Some Speculations on Argument Shift, Clitics and Crossing in West-FlemishAbraham, Werner & Bayer, Josef (eds.)Dialektsyntax (Linguistische Berichte Sonderheft 6OpladenWestdeutscher Verlag131-160
  • Haegeman, Liliane1993Some Speculations on Argument Shift, Clitics and Crossing in West-FlemishAbraham, Werner & Bayer, Josef (eds.)Dialektsyntax (Linguistische Berichte Sonderheft 6OpladenWestdeutscher Verlag131-160
  • Haegeman, Liliane1993Some Speculations on Argument Shift, Clitics and Crossing in West-FlemishAbraham, Werner & Bayer, Josef (eds.)Dialektsyntax (Linguistische Berichte Sonderheft 6OpladenWestdeutscher Verlag131-160
  • Haegeman, Liliane1993The morphology and distribution of object clitics in West FlemishStudia Linguistica4757-94
  • Haegeman, Liliane1993Some Speculations on Argument Shift, Clitics and Crossing in West-FlemishAbraham, Werner & Bayer, Josef (eds.)Dialektsyntax (Linguistische Berichte Sonderheft 6OpladenWestdeutscher Verlag131-160
  • Haeseryn, Walter, Romijn, Kirsten, Geerts, Guido, Rooij, Jaap de & Toorn, Maarten C. van den1997Algemene Nederlandse spraakkunstGroningenNijhoff
  • Haeseryn, Walter, Romijn, Kirsten, Geerts, Guido, Rooij, Jaap de & Toorn, Maarten C. van den1997Algemene Nederlandse spraakkunstGroningenNijhoff
  • Haeseryn, Walter, Romijn, Kirsten, Geerts, Guido, Rooij, Jaap de & Toorn, Maarten C. van den1997Algemene Nederlandse spraakkunstGroningenNijhoff
  • Haeseryn, Walter, Romijn, Kirsten, Geerts, Guido, Rooij, Jaap de & Toorn, Maarten C. van den1997Algemene Nederlandse spraakkunstGroningenNijhoff
  • Haeseryn, Walter, Romijn, Kirsten, Geerts, Guido, Rooij, Jaap de & Toorn, Maarten C. van den1997Algemene Nederlandse spraakkunstGroningenNijhoff
  • Haeseryn, Walter, Romijn, Kirsten, Geerts, Guido, Rooij, Jaap de & Toorn, Maarten C. van den1997Algemene Nederlandse spraakkunstGroningenNijhoff
  • Huybregts, Riny1991CliticsModel, Jan (ed.)Grammatische analyseDordrechtICG Printing
  • Huybregts, Riny1991CliticsModel, Jan (ed.)Grammatische analyseDordrechtICG Printing
  • Huybregts, Riny1991CliticsModel, Jan (ed.)Grammatische analyseDordrechtICG Printing
  • Huybregts, Riny1991CliticsModel, Jan (ed.)Grammatische analyseDordrechtICG Printing
  • Huybregts, Riny1991CliticsModel, Jan (ed.)Grammatische analyseDordrechtICG Printing
  • Paardekooper, P.C1961Persoonsvorm en voegwoordDe Nieuwe Taalgids54296-301
  • Riemsdijk, Henk C. van1978A case study in syntactic markedness: the binding nature of prepositional phrasesPeter de Ridder Press
  • Zwart, Jan-Wouter1993Dutch syntax. A minimalist approachGroningenUniversity of GroningenThesis
  • Zwart, Jan-Wouter1993Dutch syntax. A minimalist approachGroningenUniversity of GroningenThesis
  • Zwart, Jan-Wouter1993Dutch syntax. A minimalist approachGroningenUniversity of GroningenThesis
  • Zwart, Jan-Wouter1996Clitics, Scrambling, and Head Movement in DutchHalpern, Aaron L. & Zwicky, Arnold M. (eds.)Approaching second. Second position clitics and related phenomenaStanfordCSLI579-611
  • Zwart, Jan-Wouter1996Clitics, Scrambling, and Head Movement in DutchHalpern, Aaron L. & Zwicky, Arnold M. (eds.)Approaching second. Second position clitics and related phenomenaStanfordCSLI579-611
  • Zwart, Jan-Wouter1996Clitics, Scrambling, and Head Movement in DutchHalpern, Aaron L. & Zwicky, Arnold M. (eds.)Approaching second. Second position clitics and related phenomenaStanfordCSLI579-611
  • Zwart, Jan-Wouter1996Clitics, Scrambling, and Head Movement in DutchHalpern, Aaron L. & Zwicky, Arnold M. (eds.)Approaching second. Second position clitics and related phenomenaStanfordCSLI579-611
  • Zwart, Jan-Wouter1996Clitics, Scrambling, and Head Movement in DutchHalpern, Aaron L. & Zwicky, Arnold M. (eds.)Approaching second. Second position clitics and related phenomenaStanfordCSLI579-611
  • Zwart, Jan-Wouter1996Clitics, Scrambling, and Head Movement in DutchHalpern, Aaron L. & Zwicky, Arnold M. (eds.)Approaching second. Second position clitics and related phenomenaStanfordCSLI579-611
Suggestions for further reading ▼
phonology
  • Dutch
  • Frisian
  • Afrikaans
Show more ▼
morphology
  • Dutch
  • Frisian
  • Afrikaans
Show more ▼
syntax
  • Dutch
  • Frisian
  • Afrikaans
Show more ▼
cite
print
This topic is the result of an automatic conversion from Word and may therefore contain errors.
A free Open Access publication of the corresponding volumes of the Syntax of Dutch is available at OAPEN.org.