• Dutch
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
4.2.1.2.Non-spatial constructions
quickinfo

Although adpositional complementive phrases are normally spatial in nature, they can sometimes also receive a non-spatial interpretation. The following subsections will consecutively discuss (idiomatic) copular, resultative and vinden-constructions.

readmore
[+]  I.  The copular construction (idioms)

Non-spatial adpositional complementives are rare. The largest group by far consists of predicative prepositional phrases that denote a mental state, and therefore typically take a +humansubject. Some examples are given in (38); see Section 1.3.3, sub I, and A8.4, sub III for more examples. These PPs are mostly fixed, idiomatic expressions.

38
a. Marie is in de wolken/'r sas/'r knollentuin.
  Marie is in the clouds/her sas/her vegetable.garden
  'Marie is on cloud nine.'
b. Jan is van streek/over zʼn toeren.
  Jan is van streek/over his toeren
  'Jan is distressed.'

Although they seem less numerous, there are also more or less fixed expressions that are predicated of -human entities. Two examples are given in (39a&b). Perhaps we can also consider the possessive van-PP in (39c) as a complementive.

39
a. Die regeling is al drie jaar van kracht.
  that regulation  is already three years  in force
  'That regulation has been in effect for three years already.'
b. De voorzieningen zijn nog niet op peil.
  the facilities  are  yet not  on standard
  'The facilities are still not up to the required standard.'
c. Dit boek is van Jan.
  this book  is of Jan
  'This book belongs to Jan.'

Below, we will investigate how these non-spatial complementives behave with respect to topicalization, scrambling, PP-over-V and R-extraction.

[+]  A.  Topicalization

The complementives in (38) and (39) can be topicalized, provided that the PP is assigned contrastive accent, as in (40).

40
a. In zʼn sas is Jan.
  in his sas  is Jan
b. Van kracht is die regeling al drie jaar.
  in force  is that regulation  already three years
c. Van Jan is dat boek.
  of Jan  is that book
[+]  B.  Scrambling

The examples in (41) show that scrambling of the complementives in (38) and (39) is prohibited. The primed examples do not improve if the moved PP is assigned contrastive stress.

41
a. dat Jan al de hele dag in zʼn sas is.
  that  Jan already the whole day  in his sas  is
a'. * dat Jan in zʼn sas al de hele dag is.
b. dat de regeling al drie jaar van kracht is.
  that  the regulation  already three years  in force  is
b'. * dat de regeling van kracht al drie jaar is.
c. dat dit boek waarschijnlijk van Jan is.
  that  this book  probably  of Jan  is
c'. * dat dit boek van Jan waarschijnlijk is.
[+]  C.  PP-over-V

PP-over-V of the complementives in (38) and (39) gives rise to a degraded result, just as in the case of the spatial complementives.

42
a. dat Jan in zʼn sas is.
  that  Jan is in his sas
a'. * dat Jan is in zʼn sas.
b. dat die regeling al drie jaar van kracht is.
  that  that regulation  already  three years  in force  is
b'. * dat die regeling al drie jaar is van kracht.
c. dat dit boek van Jan is.
  that  this book  of Jan  is
c'. * dat dit boek is van Jan.
[+]  D.  R-extraction

Non-spatial complementives generally do not allow R-extraction, which may be due to the idiomatic nature of these constructions, that is, to the fact that the complements of the prepositions normally do not refer to an entities in the domain of discourse. If this suggestion is on the right track, example (43a) is unacceptable because the complement of the PP, sas, has no denotation at all and example (43a') is only acceptable if the complement of the PP is interpreted literally, that is, if knollentuin refers to some kind of garden. Something similar holds for example (43b), since the noun kracht has no denotation in the domain of discourse. The possessive construction in (43c) is possible due to the fact that the noun jongen does have a denotation.

43
a. * de sas waar Jan in is
  the sas  where  Jan in is
a'. # de knollentuin waar Jan in is
  the vegetable.garden  where  Jan in is
b. * de kracht waar de regeling van is
  the force  where  the regulation  in  is
c. de jongen waar dit boek van is
  the boy  where  this book  of  is
  'the boy to whom this book belongs'
[+]  II.  The resultative construction

This subsection is concerned with non-spatial adpositional resultative constructions like those given in (44). Such constructions generally involve a prepositional phrase headed by the preposition tot (literally “until”) followed by a bare (possibly modified) noun phrase and express that, as the result of some action of the subject of the clause, the referent of the accusative argument in the clause obtains the state of having the property denoted by the nominal complement of the preposition, that is, that it obtains the state of being a knight, a chairman, or a teacher, respectively.

44
a. De koningin slaat Els morgen tot ridder in de orde van Oranje-Nassau.
  the queen  hits  Els  tomorrow  tot knight in the order of Orange-Nassau
  'The queen will knight Els tomorrow.'
b. We benoemen Marie morgen tot voorzitter van het AVT-bestuur.
  we  appoint  Marie tomorrow  tot chairman of the AVT-board
  'Tomorrow, we appoint Marie chairman of the AVT-board.'
c. We leiden onze studenten binnen een jaar tot leraar op.
  we  train  our students  within a year  tot teacher  prt.
  'We train our students as teachers within a year.'

In (45a), we give some more examples that involve the preposition tot. The resultative construction is not restricted to PPs headed by tot: in (45b&c), we give other potential examples involving als, and in/te. The discussion below will focus on the examples in (45a).

45
a. tot ridder slaan'to knight', tot vijand maken'to make into an enemy', aanstellen tot (penningmeester)'to appoint (treasurer)', bekeren tot (christen)'to convert to Christianity', benoemen tot (secretaris)'to appoint to (secretary)', bevorderen tot (generaal)'promote to general', promoveren tot doctor'to take the doctor degree', tot moes stampen'to hit to pulp', tot schuim kloppen'to whip to foam'tot ridder slaan'to knight', tot vijand maken'to make into an enemy', aanstellen tot (penningmeester)'to appoint (treasurer)', bekeren tot (christen)'to convert to Christianity', benoemen tot (secretaris)'to appoint to (secretary)', bevorderen tot (generaal)'promote to general', promoveren tot doctor'to take the doctor degree', tot moes stampen'to hit to pulp', tot schuim kloppen'to whip to foam'
b. aanstellen als'to appoint as', benoemen als'to appoint as', kiezen als'to elect as', erkennen als'to recognize as'
c. in scherven/te pletter vallen'to smash to smithereens', in mootjes hakken'to cut into pieces'
[+]  A.  Topicalization

As in the case of the spatial complementives, topicalization gives rise to a fully acceptable result, provided that the complement of the PP is assigned emphatic accent.

46
a. Tot ridder slaat de koningin Els morgen.
  tot knight  hits  the queen Els tomorrow
b. Tot voorzitter benoemen we Marie morgen.
  tot chairman  appoint  we Marie tomorrow
c. Tot leraar leiden we onze studenten binnen een jaar op.
  tot teacher  train  we our students  within a year  prt.
[+]  B.  Scrambling

Scrambling is categorically blocked and, again, the non-spatial resultatives behave like the spatial ones in this respect.

47
a. dat de koningin Els morgen tot ridder slaat.
  that  the queen  Els tomorrow  tot knight  hits
a'. * dat de koningin Els tot ridder morgen slaat.
b. dat we Marie morgen tot voorzitter benoemen.
  that  we Marie tomorrow  tot chairman  appoint
b'. * dat we Marie tot voorzitter morgen benoemen.
c. dat we onze studenten binnen een jaar tot leraar opleiden.
  that  we our students  within a year  tot teacher  prt.-train
c'. * dat we onze studenten tot leraar binnen een jaar opleiden.
[+]  C.  PP-over-V

The question as to whether non-spatial adpositional resultatives allow PP-over-V is complex. The (a)-examples in (48) suggest that PP-over-V is prohibited, but the (b)- and (c)-examples show that in the majority of cases PP-over-V is allowed.

48
a. dat de koningin Els morgen tot ridder slaat.
  that  the queen  Els tomorrow  tot knight  hits
a'. * dat de koningin Els morgen slaat tot ridder.
b. dat we Marie morgen tot voorzitter benoemen.
  that  we Marie tomorrow  tot chairman  appoint
b'. dat we Marie morgen benoemen tot voorzitter.
c. dat we onze studenten binnen een jaar tot leraar opleiden.
  that  we our students  within a year  tot teacher  prt.-train
c'. dat we onze studenten binnen een jaar opleiden tot leraar.

This difference in acceptability of PP-over-V seems to be related to the fact that the verb in the (a)-examples is simple, whereas it is prefixed or combined with a verbal particle in the (b)- and (c)-examples. The examples in (49), which were given earlier in (25) and (26), show that in the case of locational constructions also, the addition of a particle like neer'down' lifts the prohibition on PP-over-V of the PP; (49b) is only acceptable if the particle neer is present, and thus suggests that the difference between (48a') and (48c') can also be accounted for by the fact that only the latter involves a verbal particle. The acceptability of (48b') shows that a prefixed verb like benoemen essentially behaves like a particle verb in this respect. For completeness' sake, (49c) shows that the locational PP resembles the PPs in (48) in that it resists scrambling.

49
a. dat Jan het boek gisteren op de tafel (neer) legde.
  that  Jan the book  yesterday  on the table  down  put
  'that Jan put the book (down) on the table yesterday.'
b. dat Jan het boek gisteren *?(neer) legde op de tafel.
c. dat Jan het boek op de tafel gisteren *?(neer) legde.

      Interestingly, the verb need not be complex from a synchronic point of view given that benoemen in (48b) is not productively related to the verb noemen'to mention/call'; the mere fact that be- is still recognizable as a prefix is apparently sufficient to license PP-over-V. Examples (50) and (51) give two more resultative verbs that allow PP-over-V but are (synchronically speaking) not derived by means of prefixation. For veranderen'to change', this is quite clear since the verb anderen is not part of the present-day vocabulary (according to the Woordenboek der Nederlandsche Taal), it used to occur with the same meaning as non-causative veranderen in syntactic frames like (50a)). The doubly-primed examples show that, like the other PPs discussed above, the PP in een schildpad'into a turtle' in (50) cannot undergo scrambling.

50
a. dat de heks gisteren in een schildpad veranderde.
  that  the witch  yesterday  into a turtle  changed
a'. dat de heks gisteren veranderde in een schildpad.
a''. * dat de heks in een schildpad gisteren veranderde.
b. dat de heks de kabouter gisteren in een schildpad veranderde.
  that the witch  the goblin  yesterday  into a turtle  changed
  'that the witch turned the goblin into a turtle.'
b'. dat de heks de kabouter gisteren veranderde in een schildpad.
b''. * dat de heks de kabouter in een schildpad gisteren veranderde.

Semantically, the verb herleiden (tot)'to reduce (to)' is only remotely related to leiden'to direct' and it is therefore doubtful whether present-day speakers are still able to relate the two verbs. Scrambling of the PP is again excluded.

51
a. dat ze het probleem gelukkig tot een misverstand konden herleiden.
  that  they  the problem  fortunately  to a misconception could reduce
  'that they could fortunately reduce the problem to a misconception.'
b. dat ze het probleem gelukkig konden herleiden tot een misverstand.
c. * dat ze het probleem tot een misverstand gelukkig konden herleiden.
[+]  D.  R-extraction

The examples in (52) show that R-extraction leads to a degraded or at least marked result. This may be due to the fact that the nouns denote properties and do not refer to specific entities in the domain of discourse.

52
a. * dat de koningin hem er toe sloeg.
  that  the queen  him  there  toe  hit
a'. * de waardigheid waar de koningin hem toe sloeg
  the dignity  where  the queen  him  toe  hit
b. *? dat we Marie er toe benoemen.
  that  we Marie there  toe  appoint
b'. ? de functie waar we Marie toe benoemen
  the function  where  we Marie toe  appoint
c. *? dat we onze studenten er toe opleiden.
  that  we our students  there  toe  prt.-train
c'. ? het beroep waar we onze studenten toe opleiden
  the profession  where  we our students  toe  prt.-train

That we are dealing with a semantic and not with a syntactic constraint on R-extraction can be supported by the fact illustrated in (53) that the syntactically similar locational resultative construction with a particle verb in (49) does allow it.

53
a. dat Jan het boek er gisteren op neer legde.
  that  Jan the book there  yesterday  on  down  put
  'that Jan put the book down on it yesterday.'
b. de tafel waar Jan het boek gisteren op neer legde
  the table  where  Jan the book  yesterday  on  down  put
  'the table on which Jan put the book down yesterday'

Examples such as (51), in which the complement of the preposition is preceded by an article and hence more likely to refer to some entity, also seems to give rise to a better result, at least in the relative clause construction in (54b).

54
a. ? dat ze het probleem er gelukkig toe konden herleiden.
  that  they  the problem  there  fortunately  toe  could reduce
  'that they fortunately were able to reduce the problem to it.'
b. het misverstand waar ze het probleem toe konden herleiden
  the misconception  where  they  the problem  toe  could reduce
  'The misconception that they were fortunately able to reduce the problem to.'
[+]  III.  The vinden-construction

The name vinden-construction is in fact a misnomer for the constructions discussed in this subsection, because they do not involve the verb vinden'to consider', but the verbs in (55). However, like true vinden- constructions such as Ik vind Marie aardig'I consider Marie nice', constructions headed by the verbs in (55) express a subjective evaluation by the referent of the subject of the clause. The verbs in the (a)-examples are followed by a phrase headed by als and those in the (b)-examples by a phrase headed by voor.

55
a. behandelen als'to treat as', beschouwen als'to consider'
b. aanzien voor'to mistake for', houden voor'to look upon as', uitmaken voor'to call (names)', verslijten voor'to take for'

The phrases headed by als and voor take a (possibly modified) adjective, a bare noun or a noun preceded by the indefinite determiner een'a' as their complement. The complement is used to denote some property attributed to the referent of the accusative argument of the clause by the referent of the subject of the clause. The number of verbs that can be used in this construction is quite limited. Some full-fledged examples are given in (56).

56
a. Ik beschouwde hem als heel knap/(een) held.
  considered  him  als  very bright/a hero
  'I considered him very bright/a hero.'
b. Ze versleten Peter voor dom.
  they  took  Peter voor stupid
  'They took Peter for stupid.'
c. Ze zien Peter voor gek aan.
  they  see  Peter voor foolish  prt.
  'They see Peter as foolish.'
c'. Ze zagen Peter voor iemand anders/een dief aan.
  they  saw  Peter voor  someone else/a thief  prt.
  'They mistake Peter for someone else/a thief.'
d. Ze maakten Marie voor leugenaar uit.
  they  made  Marie  voor  liar  prt.
  'They called Marie a liar.'

It is important to note that there are various unsolved problems with the constructions in (56). First, it is not clear whether als should be seen as a preposition; cf. Section 1.4, sub I. Neither is it clear what determines whether the noun can or cannot be preceded by the indefinite article een'a': the article optionally precedes the noun held in (56a), it must precede the noun dief in (56c'), and it cannot precede the noun leugenaar in (56d). The verb verslijten in (56b) seems to require that the property denoted by the complement of voor be negatively valued: *Ze versleten Peter voor slim (intended reading: “They took Peter for a smart person”). We will ignore all these questions and leave them for future research.
      Just as in the case of the non-spatial resultative constructions, the “vinden”-constructions are headed by prefixed verbs, as in (56a&b), or particle verbs, as in (56c&d). Note that from a synchronic point of view the prefixed verbs are not derived by means of a productive process: beschouwen is not perceived as being derived from schouwen'to look (at)', which is archaic anyway, and neither is verslijten perceived as being derived from slijten'to wear (out)'; the productively derived form verslijten means “to wear out”. The following subsections will discuss the behavior of the “vinden”-constructions in (56) with respect to topicalization, scrambling, PP-over-V and R-extraction.

[+]  A.  Topicalization

The examples in (57) show that, if the adpositional phrase is assigned emphatic accent, topicalization generally gives rise to a reasonably acceptable result in the “vinden”-construction.

57
a. ? Als heel knap/(een)held beschouwde ik hem.
  als  very bright/a hero  considered  him
b. Voor dom versleten ze Peter.
  voor stupid  took  they  Peter
c. Voor gek zien ze Peter aan.
  voor foolish  see  they  Peter prt.
c'. Voor iemand anders zagen ze Peter aan.
  voor  someone else/a thief  saw  they  Peter prt.
d. Voor leugenaar maakten ze Marie uit.
  voor  liar  made  they Marie prt.
[+]  B.  Scrambling

Scrambling leads to an unacceptable result in the “vinden”-construction.

58
a. Ik beschouwde hem altijd al als heel knap/held.
  considered  him  always  already  als  very bright/hero
a'. * Ik beschouwde hem als heel knap/held altijd al.
b. Ze versleten Peter vaak voor dom.
  they  took  Peter often  voor stupid
b'. * Ze versleten Peter voor dom vaak.
c. Ze zagen Peter natuurlijk voor gek/iemand anders aan.
  they  saw  Peter of.course  voor  foolish/someone else  prt.
c'. * Ze zagen Peter voor gek/iemand anders natuurlijk aan.
d. Ze maakten Marie gisteren voor leugenaar uit.
  they  made  Marie  yesterday  voor liar  prt.
d'. * Ze maakten Marie voor leugenaar gisteren uit.
[+]  C.  PP-over-V

Since we are dealing with prefixed verbs and particle verbs, our discussion of PP-over-V in resultative constructions in Subsection II would lead us to expect that PP-over-V is acceptable. This expectation is only partly borne out; the examples in (59) show that PP-over-V is marked in the “vinden”-construction if the complement of the preposition is adjectival in nature.

59
a. dat ik hem altijd al als held/heel knap beschouwde.
  that  him  always  already  als  hero/very bright  considered
a'. dat ik hem altijd al beschouwde als held/?heel knap.
b. dat ze Peter altijd al voor dom versleten.
  that  they  Peter always  already  voor stupid  took
b'. ?? dat ze Peter altijd al versleten voor dom.
c. dat ze Peter voor iemand anders/gek aanzagen.
  that  they  Peter voor  someone else/foolish  prt.-saw