• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
2.3.1.Movement of the PP-complement
quickinfo

This section shows how discontinuous APs may arise by movement of PP-complements. We start by discussing PP-over-V, which results in placement of the PP after the verb(s) in clause-final position. After this we will discuss several processes that place the PP-complement in a position preceding the adjective. This section is concluded by a brief discussion of PP-complements of pseudo-participles and deverbal adjectives, which exhibit somewhat deviant behavior.

readmore
[+]  I.  PP-over-V

When we consider the relative order of PPs and main verbs in clause-final position, it turns out that many PPs may occur on both sides of the verb as a result of PP-over-V. This is illustrated in (51): (51b) involves PP-over-V of the adverbial adjunct of place op het station; (51c) involves PP-over-V of the PP-complement op zijn vader of the main verb, and in (51d) both PPs follow the main verb.

51
a. Jan heeft op het station op zijn vader gewacht.
  Jan has  at the station  for his father  waited
  'Jan has waited for his father at the station.'
b. Jan heeft op zijn vader gewacht op het station.
c. Jan heeft op het station gewacht op zijn vader.
d. Jan heeft gewacht op zijn vader op het station.

      Now consider the examples in (52), which involve an adjective that takes a PP-complement. Example (52b) shows that this PP-complement may also undergo PP-over-V, which results in a structure in which the AP and its PP-complement are no longer adjacent; note that we indicated the original position of the moved PP by means of the tracet.

52
a. Jan is nooit [AP trots op zijn vader] geweest.
  Jan is never  proud  of his father  been
  'Jan has never been proud of his father.'
b. Jan is nooit [AP trots ti] geweest [op zijn vader]i

Given the structure in (52b), it does not come as a surprise that the adjective can be topicalized in isolation; the structure in (53a) involves movement of the same constituent as in (50b), namely AP (the indices, i and j in this case, keep track of what moves where). Observe that the adjective normally cannot be topicalized if the PP occurs between the adverb nooit and the participle geweest, as in (53b); this is only possible if the frequency adverb nooit receives focus accent. This suggests that PP-over-V must apply in order to make topicalization of the adjective in isolation possible (although there is an alternative option that may license this, which will be discussed in Subsection IIB).

53
a. [AP Trots ti]j is Jan nooit tj geweest [op zijn vader]i.
b. ?? Trots is Jan nooit op zijn vader geweest.

      Example (50c), repeated here as (54a), may have a structure similar to (53a). The main difference would be that PP-over-V cannot be observed because the verb is not in clause-final position, but occupies the second position of the main clause as a result of Verb-Second. However, since the finite verb occupies the clause-final position in embedded clauses, this predicts that when movement of the adjective takes place in an embedded clause, the PP may show up after the finite verb. That this is indeed borne out is shown in (54b), which contains an embedded interrogative (or exclamative) clause.

54
a. Trots is Jan op zijn vader.
  proud  is Jan of his father
b. (Je weet niet) [hoe trots Jan is op zijn vader].
  you know not   how proud  Jan is of his father
[+]  II.  PP-preposing

Subsection I has shown that discontinuity may arise as a result of PP-over-V, and this subsection shows that it may also be the result of PP-preposing. Two cases can be distinguished: leftward movement of the PP into the initial position of the clause (topicalization and wh-movement) and leftward movement into some clause-internal position (focus and negation movement).

[+]  A.  Topicalization and wh-movement

Another source of discontinuous APs is PP-preposing. One case involves movement of the PP into clause-initial position. Consider again example (51a). The primeless examples in (55) show that both the adverbial PP op het station and the complement PP op zijn vader can be moved into clause-initial position by topicalization. The primed examples show that the same result can be obtained by means of wh-movement if the nominal complement of the preposition is questioned.

55
a. Op het stationi heeft Jan ti gewacht.
  at the station  has  Jan  waited
a'. Op welk stationi heeft Jan ti gewacht?
  at which station  has  Jan  waited
b. Op zijn vaderi heeft Jan ti gewacht.
  for his father  has  Jan  waited
b'. Op wiei heeft Jan ti gewacht?
  for whom  has  Jan  waited

The examples in (56) show that PP-complements of adjectives can undergo the same processes. This is another source for the discontinuity of the AP.

56
a. Jan is nooit [AP trots op zijn vader] geweest.
  Jan is  never  proud  of his father  been
  'Jan has never been proud of his father.'
b. [Op zijn vader]i is Jan nooit [AP trots ti] geweest.
   of his father  is Jan never  proud  been
c. [Op wie]i is Jan nooit [AP trots ti] geweest?
  of whom  is Jan never  proud  been
[+]  B.  Short leftward movement

Leftward movement of PP-complements need not always involve movement into clause-initial position, but may also target some clause-internal position. This kind of movement will be referred to as short leftward movement. At least two types of short leftward movement can be distinguished: focus and negation movement.

[+]  1.  Focus movement

The (a)-examples in (57) illustrate short leftward movement of the PP-complement of the verb praten'to talk' across the adverbial phrase niet langer'no longer'. Short leftward movement of PPs normally results in a word order that is perceived as marked, and is only possible if the nominal complement of the preposition is able to bear accent; if the nominal complement is a weak pronoun, as in the (b)-examples, short leftward movement of the PP is excluded. Although this goes against a popular belief (which finds its origin in Neeleman 1994b and Vikner 1994/2006), we will assume that short leftward movement of PPs is an instance of focus movement; see Broekhuis (2008:67ff.) for more extensive discussion.

57
a. Jan wil niet langer op zijn vader wachten.
  Jan wants  no longer  for his father  wait
  'Jan doesnʼt want to wait for his father any longer.'
a'. Jan wil [op zijn vader]i niet langer ti wachten.
b. Jan wil niet langer op ’m wachten.
  Jan wants  no longer  for him  wait
  'Jan doesnʼt want to wait for him any longer.'
b'. * Jan wil [op ’m]i niet langer ti wachten.

Example (58b) shows that focus movement is also possible with PP-complements of adjectives. Example (58c) further shows that this movement is only possible if the nominal complement of the preposition is able to bear accent; if the complement is a weak pronoun, short leftward movement of the PP is excluded.

58
a. Jan is altijd al [AP trots op zijn vader/’m] geweest.
  Jan has  always  proud  of his father/him  been
  'Jan has always been proud of his father.'
b. Jan is [op zijn vader]i altijd al [AP trots ti] geweest.
c. * Jan is [op ’m]i altijd al [AP trots ti] geweest.

      Since the adverbial phrase of frequency altijd al in (58) modifies the clause and focus movement of the PP crosses this modifier, we can safely assume that the landing site of focus movement is an AP-external position. This is also confirmed by the fact that the adjective can be topicalized and wh-moved in isolation, albeit that topicalization may require that the adjective be contrastively stressed.

59
a. [AP Trots ti]j is Jan [op zijn vader]itj geweest.
b. [AP Hoe trots ti]j is Jan [op zijn vader]itj geweest?

This does not automatically preclude, however, that there may be an additional AP-internal landing site. If this were the case, we would expect that the PP could also follow the adverbial phrase and that the preposed PP could be pied-piped under AP-topicalization. Since the resulting structures in (60b&c) are highly marked, these expectations do not seem to be borne out.

60
a. Jan is altijd al [AP trots op zijn vader] geweest.
  Jan has  always  proud  of his father  been
b. ?? Jan is altijd al [AP [op zijn vader]i trots ti] geweest.
c. ?? [AP [Op zijn vader]i trots ti]j is Jan altijd al tj geweest.

It should be noted, however, that (60b) improves considerably if the adverbial phrase altijd al is assigned emphatic accent, as in (61a). Still, given that AP-topicalization cannot pied-pipe the PP in this case either, we have to maintain that the landing site of the preposed PP is AP-external, but has simply not crossed the emphatically focused adverbial phrase.

61
a. Jan is altijd al op zijn vaderi trots ti geweest.
b. [AP Trots ti]j is Jan altijd al [op zijn vader]itj geweest.
c. ?? [AP [Op zijn vader]i trots ti] is Jan altijd al geweest.

      The discussion above has shown that as a result of focus movement, many adjectives allow their PP-complement to their left. The examples in (62) simply provide an additional illustration of the resulting word order variation. The (a)-examples show that the adjective and the postadjectival PP form a clausal constituent that may be topicalized as a whole. The adjective and the preadjectival PP, on the other hand, do not form a constituent, which is clear from the fact, illustrated in the (b)-examples, that AP-topicalization cannot pied-pipe the PP.

62
a. dat Els bang voor de hond is.
  that  Els afraid  of the dog  is
a'. [AP Bang voor de hond]j is Els niet tj.
b. dat Els [voor de hond]i bang ti is.
b'. [AP Bang ti]j is Els [voor de hond]i niet tj.
b''. ?? [AP [Voor de hond]i bang ti]j is Els niet tj.

      The examples in (63) show that focus movement of the PP-complement is less felicitous with some of the adjectives in Table 2, that is, those adjectives that show a change of meaning if the PP is omitted/added. The reason for this seems to be that focus movement appears to block the lexicalized meaning of the A+P collocation in favor of a more compositional one: if the PP follows the adjective, the idiomatic meaning “fed up with" is possible; after focus movement, on the other hand, only the compositional meaning “ill as a result of" survives.

63
a. dat Jan ziek van die zuurkool is.
  that  Jan fed.up  with this sauerkraut  is
  'that Jan is fed up with this sauerkraut.'
b. dat Jan van die zuurkool ziek is.
  Not: 'that Jan is fed up with this sauerkraut.'
  Possible: 'that this sauerkraut made Jan ill.'

If the compositional meaning leads to gibberish, as in (64), short leftward movement simply leads to a degraded result.

64
a. dat Jan dol/gek op zijn kinderen is.
  that  Jan fond  of his children  is
a'. ?? dat Jan op zijn kinderen dol/gek is.
b. dat Jan vol van die gebeurtenis is.
  that  Jan full  of that incident  is
b'. ?? dat Jan van die gebeurtenis vol is.

It should be noted, however, that assigning contrastive accent to the adjective or the addition of an accented intensifier may considerably improve the result of focus movement of the PP-complement. This is illustrated by the examples in (65), which seem to be fully acceptable.

65
a. dat Jan op zijn kinderen dol/gek is.
  that  Jan  of his children  fond  is
a'. dat Jan op zijn kinderen hartstikke dol is.
  that  Jan  of his children  extremely  fond  is
b. dat Jan van die gebeurtenis vol is.
  that  Jan of that incident  full  is
b'. dat Jan van die gebeurtenis helemaal vol is
  that  Jan of that incident  completely full is

Needless to say, topicalization of the adjectives in (63) and (64) can only pied-pipe the PP-complement if it follows the adjective, e.g., [gek/dol op zijn kinderen] is Jan versus *[op zijn kinderen (hartstikke) gek/dol] is Jan.
      Topicalization and wh-movement of the PP-complement contrast sharply with focus movement; these movements leave the idiomatic reading intact and, as a result, always give rise to a completely acceptable result. We illustrate this in (66) for the topicalization/wh-movement counterparts of the primed examples of (64).

66
a. Op zijn kinderen is hij dol/gek.
  of his children  is  he  fond
a'. Op wie is hij dol/gek?
  of whom  is  he  fond
b. Van die gebeurtenis is hij vol.
  of that incident is  he  full
b'. Van welke gebeurtenis is hij vol?
  of which incident  is  he  full

      This subsection has shown that PP-complements that precede their selecting adjectives have been moved from their original postadjectival position into some other position in the clause. Section 2.3.1, sub III, will show, however, that there are certain exceptions to this general rule: pseudo-participles and certain deverbal adjectives may take their PP-complement to their left.

[+]  2.  Neg movement

This subsection discusses another case of short leftward movement of PP-complements, which does not involve focus but takes place when the nominal complement of the preposition is negated. We will refer to this movement, which is illustrated in (67), as neg-movement; cf. Haegeman (1991/1995).

67
a. ?? dat Marie tevreden over niets is.
  that  Marie satisfied  about nothing  is
a'. dat Marie [over niets]i tevreden ti is.
b. ?? dat Jan gevoelig voor geen enkel argument is.
  that  Jan susceptible  to no single argument  is
b'. dat Jan [voor geen enkel argument]i gevoelig ti is.
c. ?? dat Els bang voor niemand is.
  that  Els afraid  of no one  is
c'. dat Els [voor niemand]i bang ti is.
d. ?? dat Jan trots op niemand is.
  that  Jan proud  of no one  is
d'. dat Jan [op niemand]i trots ti is.

Neg-movement seems to be obligatory and it has been suggested that it is needed in order to allow negation to take scope over the complete clause, which results in the following meaning for example (67c): “it is not the case that Els is afraid of someone". In fact, it seems that the need for neg-movement also blocks the application of PP-over-V, as will be clear from the degraded status of the examples in (68).

68
a. ?? dat Marie tevreden is over niets.
b. ?? dat Jan gevoelig is voor geen enkel argument.
c. ?? dat Els bang is voor niemand.
d. ?? dat Jan trots is op niemand.

Note in passing that if the nominal complement of the preposition is inanimate, neg-movement can also affect the negative element in isolation by means of extraction of the negative R-pronounnergens from a pronominal PP. So, besides the examples in (67a'&b'), we also find the constructions in (69); we will ignore these alternatives in what will follow, while noting that the leftward movement of the nergens suffices to assign negation scope over the complete clause.

69
a. dat Jan nergens tevreden over is.
  that Jan  nowhere  satisfied about is
  'that Jan isnʼt satisfied about anything.'
b. dat Jan nergens gevoelig voor is.
  that  Jan nowhere  susceptible  to  is
  'that Jan isnʼt susceptible to anything.'

      If neg-movement does not apply, we are dealing with constituent negation. The constituent negation reading does not give rise to a very felicitous result for the examples in (67), but is possible in (70), where the two examples form a minimal pair.

70
a. dat Jan tevreden met niets is.
  that  Jan satisfied  with nothing  is
  'that Jan is satisfied with very little.'
b. dat Jan met niets tevreden is.
  that Jan  with nothing  satisfied is
  'that Jan isnʼt satisfied with anything.'

Example (70a), in which the PP-complement occupies its original postadjectival position, involves constituency negation; this examples literally means that Jan will be happy if he has got nothing, but is normally used in an idiomatic sense to express that Jan has virtually no needs. This interpretation contrasts sharply with the one associated with example (70b), in which neg-movement has applied; this example expresses that Jan will not be happy, no matter what he obtains. For completeness’ sake, note that PP-over-V in dat Jan tevreden is met niets is compatible with the constituent negation reading in (70a), but not with the sentential negation reading in (70b).
      The data in (70) support the claim that neg-movement is needed in order to express sentential negation. Another argument in favor of this claim can be based on the fact illustrated in (71) that the negative polarity verb hoeven requires the presence of a negative adverb niet'not' or some other negated element like niemand'no one' that takes clausal scope.

71
a. Je hoeft *(niet) te komen.
  you  need    not  to come
  'You donʼt have to come.'
b. Je hoeft niemand/*iemand te overtuigen.
  you  need  nobody/someone  to convince
  'You donʼt have to convince anybody.'

If the negated element is part of the PP-complement of an adjective, and the PP stays in its original position, the use of hoeven is completely unacceptable. However, if the PP is moved to the left, as in (72b), the result is perfect. This would be consistent with the fact that sentential negation requires neg-movement. For completeness’ sake, (72c) shows that PP-over-V is also excluded in this context.

72
a. * Je hoeft bang voor niemand te zijn.
  you  need  afraid  of  no one  to be
b. Je hoeft voor niemand bang te zijn.
c. * Je hoeft bang te zijn voor niemand.

      To conclude this subsection, we want to note that West-Flemish provides morphological evidence in favor of the claim that negation can only have clausal scope if the PP-complement has undergone neg-movement. Sentential negation can be morphologically expressed by supplementing the finite verb with the (optional) negative marker en-. This marker is possible if the PP-complement of the adjective has undergone neg-movement, as in (73a), but not if the PP occupies its original position or has undergone PP-over-V, as in (73b).

73
a. da Valère van niemand ketent en-is.
  that  Valère of no one  satisfied  NEG-is
  'that Valère isnʼt pleased with anyone.'
b. * da Valère ketent <van niemand> en-is <van niemand>.
  that  Valère  satisfied     of no one  NEG-is
[+]  III.  An exceptional case: pseudo-participles and deverbal adjectives

Subsection II has shown that PP–A orders are normally the result of leftward movement of the PP-complement. This subsection discusses adjectives that are different in that the original position of their PP-complement may be on their left.

[+]  A.  PP–A orders in clause-initial position

Subsection IIB has shown that the PP–A order normally gives rise to a degraded result if the AP is moved into clause-initial position; cf. (60c), (61c) and (62b''). The examples in (74) show, however, that some adjectives behave differently in this respect. The acceptability of the primed examples could be accounted for by assuming that, at least in some cases, leftward movement of the PP-complement into some AP-internal position is possible after all, but we will argue instead that the PP–A order in clause-initial position is restricted to two morphologically definable classes that in some sense show verbal behavior.

74
a. [AP Geschikt voor deze functie] is hij niet.
  suitable  for this office  is he not
a'. [AP Voor deze functie geschikt] is hij niet.
b. [AP Afhankelijk van zijn toestemming] ben ik niet.
  dependent  on his permission  am  not
b'. [AP Van zijn toestemming afhankelijk] ben ik niet.

Most adjectives that allow the PP–A order in clause-initial position have the appearance of a past/passive participle; cf. Table 6. However, since the adjectives in Table 6A-C do not have a verbal counterpart, they must be considered pseudo-participles. The irregular forms in Table 6D do have a verbal counterpart but these have a completely different meaning: the verb voldoen means “to pay" or “to be sufficient"; the verb begaan means “to commit".

Table 6: Pseudo-participles that take a prepositional complement
form example translation
A. ge- .. -d/t/en gebrand op keen on
gekant tegen opposed to
geschikt voor suitable for
gespitst op especially alert to
gesteld op keen on
ingenomen met delighted with
B. ver- .. -d/t verliefd op in-love with
verrukt over delighted at
verwant aan related to
C. be- .. -d/t bedacht op cautious for
bekend met familiar with
bereid tot willing to
bevreesd voor fearful of
D. irregular forms voldaan over content with
begaan met emotionally involved with

In addition there are a small number of adjectives that are derived from a verb by means of the suffixes -baar and -elijk; the preposition of their PP-complement is identical to the one in the corresponding verbal construction. Three examples are given in (75).

75
Deverbal adjectives
Verbal stem
a. verenigbaar met 'compatible with'
a'. verenigen met 'to reconcile with'
b. vergelijkbaar met 'comparable to'
b'. vergelijken met 'to compare with'
c. afhankelijk van 'dependent on'
c'. afhangen van 'to depend on'

Given that topicalized past participles and infinitives allow their PP-complement both on their left and on their right (cf. (76)), it may not be accidental that the pseudo-participles in Table 6 and the deverbal adjectives in (75) also allow both orders in topicalized position.

76
a. [VP Gewacht op zijn vader] heeft Jan niet.
  waited  for his father  has  Jan not
  'Jan has not waited for his father.'
a'. [VP Op zijn vader gewacht] heeft Jan niet.
b. [VP Wachten op zijn vader] wil Jan niet.
  waiting  for his father  wants  Jan not
  'Jan doesnʼt want to wait for his father.'
b'. [VP Op zijn vader wachten] wil Jan niet.

The following subsections will show that there are indeed reasons for assuming that pseudo-participles and certain deverbal adjectives exhibit verbal behavior. This suggests that the PP–A order with these adjectives can be accounted for by other means than by taking recourse to AP-internal movement; the data in fact suggest that the PP is base-generated to the left of the adjective.

[+]  B.  R-extraction from the PP-complement

If the preposition of a PP-complement of a verb is stranded by means of R-extraction, it always precedes the verb. The stranded preposition of the complement of an adjective, on the other hand, normally follows the adjective. This is illustrated in, respectively, (77) and (78).

77
a. Jan heeft er niet <op> gewacht<*op>.
verb
  Jan has  there  not  for  waited
  'Jan didnʼt wait for it.'
b. Jan heeft er niet <naar> gezocht<*naar>.
verb
  Jan has  there  not  for  searched
  'Jan didnʼt search for it.'
78
a. Jan is er nog steeds <*op> trots <op>.
adjective
  Jan is there  prt  still      of  proud
  'Jan is still proud of it.'
b. Jan is er nog steeds <*voor> bang <voor>.
adjective
  Jan is there  prt  still      of  afraid
  'Jan is still afraid of it.'

The pseudo-participles in Table 6 and the deverbal adjectives in (75), however, are more equivocal in this respect; they allow the stranded preposition on both sides.

79
a. Jan is er niet <voor> geschikt <voor>.
pseudo-participle
  Jan is there  not     for  suitable
  'Jan isnʼt suitable for it.'
b. Jan is er niet <mee> bekend <mee>.
pseudo-participle
  Jan is there  not    with  familiar
  'Jan isnʼt familiar with it.'
c. Jan is er helemaal <van> afhankelijk <van>
deverbal adjective
  Jan is there  completely    on dependent
  'Jan is completely dependent on it.'

The judgments on the degree of acceptability of the examples in (79) with the stranded preposition preceding the adjective may vary somewhat from speaker to speaker, but they are consistently considered much better than those on the corresponding examples in (78). As is illustrated in (80), the stranded preposition is occasionally even rejected in post-adjectival position.

80
a. Jan is er niet <mee> ingenomen <??mee>.
pseudo-participle
  Jan is there  not    with  pleased
  'Jan isnʼt pleased with it.'
b. Jan is er niet <tegen> gekant <*?tegen>.
pseudo-participle
  Jan is there  not    to  opposed
  'Jan isnʼt opposed to it.'

      Often, the position of the stranded preposition is taken to indicate the unmarked position of the PP-complement. The fact that stranded prepositions are situated to the left of the past participle in the examples of (77) is then derived from the general OV-character of Dutch; like nominal complements, PP-complements have an unmarked position immediately to the left of the verb. If this is on the right track, the (78) indicate that PP-complements of adjectives should have an unmarked position immediately to the right of the adjective. The pseudo-participles in Table 6 and the deverbal adjectives in (75) should then be equivocal in this respect: the unmarked position of their PP-complement may be either to their right or to their left. The next subsection will provide more evidence in favor of this conclusion.

[+]  C.  The position of the PP-complement with respect to intensifiers

The introduction to this chapter has shown that intensifiers like erg'very' can be pied-piped under AP-topicalization, and hence that such intensifiers are part of the AP; cf. the discussion of (1). Further, we have seen that focus movement of the PP-complement probably targets a position external to AP. From this, it follows that focus movement places the PP-complement in front of intensifiers such as zeer/erg'very'. The examples in (81) show that this is indeed borne out; the PP-complement cannot intervene between the modifier and the adjective.

81
a. dat Jan zeer trots op zijn kinderen is.
  that  Jan very proud  of his children  is
a'. dat Jan <op zijn kinderen> zeer <*op zijn kinderen> trots is.
b. dat Marie erg tevreden over het resultaat is.
  that  Marie very satisfied  about the result  is
b'. dat Marie <over het resultaat> erg <*over het resultaat> tevreden is.
c. dat Els zeer bang voor de hond is.
  that  Els very afraid  of the dog  is
c'. dat Els <voor de hond> zeer <*voor de hond> bang is.

      The previous subsections suggested that the pseudo-participles in Table 6 and the deverbal adjectives in (75) may take their PP-complement to their left. Since modifiers are more peripheral to the phrase than complements, this correctly predicts that the PP-complement of these adjectives can be placed between the intensifier zeer'very' and the adjective.

82
a. Jan is erg met dat voorstel ingenomen.
  Jan is very  with that proposal  delighted
  'Jan is very delighted with that proposal.'
b. Jan is zeer tegen dat voorstel gekant.
  Jan is very  to that proposal  opposed
  'Jan is strongly opposed to that proposal.'

Of course, the PP-complement may also precede the modifier as the result of focus movement or neg-movement. Note that the negative PP in the primed examples of (83) cannot occupy the position in between the intensifier and the adjective, which is, of course, due to the already established fact that neg-movement is obligatory; cf. the discussion in IIB2.

83
a. Jan is <met dat voorstel> erg <met dat voorstel> ingenomen.
  Jan is  with that proposal  very  delighted
  'Jan is very delighted with that proposal.'
a'. Jan is <met niemand> erg <*met niemand> ingenomen.
  Jan is    with no one  very  delighted
b. Jan is <tegen dat voorstel> zeer <tegen dat voorstel> gekant.
  Jan is    to that proposal  very opposed
  'Jan is strongly opposed to that proposal.'
b'. Jan is <tegen niemand> zeer <*tegen niemand> gekant.
  Jan is    to no one  very  opposed

The claim that the stranded preposition indicates the unmarked position of the PP-complement correctly predicts that it must be situated between the adverbial modifier and the adjective. The data in (84) therefore provide additional support for the claim that the pseudo-participles in Table 6 and the deverbal adjectives in (75) differ from the other adjectives in that they may take their PP-complement to their immediate left.

84
a. Jan is er niet erg mee ingenomen.
  Jan is there  not  very  with  delighted
  'Jan isnʼt very delighted with it.'
a'. * Jan is er niet mee erg ingenomen.
b. Jan is er zeker erg tegen gekant.
  Jan is there  certainly  very  to  opposed
  'Jan is certainly strongly opposed to it.'
b'. * Jan is er zeker tegen erg gekant.
[+]  D.  On the categorial status of pseudo-participles and deverbal adjectives

The fact that the stranded preposition may be placed either before or after pseudo-participles and deverbal adjectives suggests that these adjectives exhibit mixed adjectival and verbal behavior. The primed examples in (85) show that this mixed categorial behavior disappears if the pseudo-participle or deverbal adjective is prefixed by on- (a typical property of adjectives); the stranded preposition can then only occur to the right, which shows that we are dealing with true adjectives.

85
a. Jan is er al jaren <van> afhankelijk <van>.
  Jan is there  for years    on  dependent
  'Jan has been dependent on it for years.'
a'. Jan is er al jaren <*van> onafhankelijk <van>.
  Jan is there  for years      on  independent
  'Jan has not been dependent on it for years.'
b. Jan is er natuurlijk <mee> bekend <mee>.
  Jan is there  of course    with  familiar
  'Of course, Jan is familiar with it.'
b'. Jan is er natuurlijk <*mee> onbekend <mee>.
  Jan is there  of course     with  un-familiar
  'Of course, Jan isnʼt familiar with it.'
[+]  E.  A concluding remark on preposition stranding

The discussion in the preceding subsections more or less followed the traditional claim that stranded prepositions occupy the base position of the prepositional complement. This assumption is not without its problems, of which we will mention one. Consider the examples in (86). If the stranded preposition occupies its base-position, and if topicalization preposes the complete AP, that is, the adjective and its arguments, we would expect the stranded preposition to be obligatorily pied-piped by topicalization of the AP. This means that we would wrongly predict the primeless examples in (86) to be ungrammatical, and the primed examples to be grammatical; it is instead the reverse that is true.

86
a. Trots is Jan er niet op.
  proud  is Jan there  not  of
a'. * [AP Trots op] is Jan er niet.
b. Boos is Jan er niet over.
  angry  is Jan there  not  about
b'. * [AP Boos over] is Jan er niet.

The situation becomes even more mysterious when we consider the topicalized counterparts of the examples in (84) in (87). At first sight, the grammatical examples (87a'-b') seem to have been derived by means of AP-topicalization from the ungrammatical examples in (84a'-b'), whereas application of AP-topicalization to the grammatical examples in (84a-b) results in the ungrammatical examples in (87a-b).

87
a. * Erg mee ingenomen is Jan er niet.
cf. the grammaticality of (84a)
a'. Erg ingenomen is Jan er niet mee.
cf. the ungrammaticality of (84a')
  'Jan isnʼt very delighted with it.'
b. * Erg tegen gekant Jan is er zeker.
cf. the grammaticality of (84b)
b'. Erg gekant is Jan er zeker tegen.
cf. the ungrammaticality of (84b')
  'Jan is certainly very opposed to it.'

We will not discuss this intriguing problem here any further; we leave it to future research to solve it, while noting that we find similar problems with PP-complements of verbs; cf. Den Besten and Webelhuth (1990).

References:
  • Besten, Hans den & Webelhuth, Gert1990StrandingGrewendorf, Günter & Sternefeld, Wolfgang (eds.)Scrambling and barriersAmsterdam/PhiladelphiaJohn Benjamins77-92
  • Broekhuis, Hans2008Derivations and evaluations: object shift in the Germanic languagesStudies in Generative GrammarBerlin/New YorkMouton de Gruyter
  • Haegeman, Liliane1991Negative concordDelfitto, Denis, Everaert, Martin, Evers, Arnold & Stuurman, Frits (eds.)Going Romance, and Beyond. Fifth Symposium on Comparative Grammar45-81
  • Haegeman, Liliane1995The syntax of negationCambridge studies in linguistics 75CambridgeCambridge University Press
  • Neeleman, Ad1994Complex predicatesUtrechtUniversity of UtrechtThesis
  • Vikner, Sten1994Scandinavian object shift and West Germanic scramblingCorver, Norbert & Riemsdijk, Henk van (eds.)Studies on Scrambling. Movement and non-movement approaches to free word-Order phenomenaBerlin/New YorkMouton de Gruyter487-517
  • Vikner, Sten2006Object ShiftEveraert, Martin & Riemsdijk, Henk van (eds.)The Blackwell companion to syntax3Malden, MA/OxfordBlackwell Publishing392-436
report errorprintcite