• Dutch1
  • Frisian
  • Saterfrisian
  • Afrikaans
Show all
3.1.4.Directional adpositional phrases
quickinfo

Directional PPs express that the located object traverses a certain path. A path can be defined as an ordered set of vectors, each of which is associated with a certain position on the time line. The path denoted by van A naar B'from A to B' can then be represented as in Figure 7, taken from Section 1.3.1.1, sub V, which can be read as a cartoon.

Figure 7: Van A naar B 'from a to B' (repeated)

The following subsections discuss some potential cases of modification of directional PPs. It will be shown that the modification possibilities are restricted to modifiers that express to what extent the implied path has been covered.

readmore
[+]  I.  Modifiers of orientation and distance

The fact that the directional PPs also involve vectors may give rise to the expectation that directional PPs can also be modified by means of adverbial phrases of orientation and distance. This, however, does not seem to be borne out. Consider the examples in (56).

56
a. # Jan liep recht/schuin naar de barkruk.
  Jan walked  straight/diagonally  to the bar stool
b. * Jan liep ver/vlak/pal naar de barkruk.
  Jan walked  far/close  to the bar stool

The examples in (56a) are acceptable but not on the intended reading in which recht'straight' and schuin'diagonally' modify the path that Jan is traversing; these examples are only acceptable if recht and schuin are interpreted as supplementives predicated of Jan that express something about Janʼs posture. This shows that recht and schuin are comparable to elements like rechtop'upright', which can never be used as a modifier of a PP. The examples in (56b) are clearly unacceptable.
      The discussion above does not entail that directional PPs cannot be modified. Perhaps the PP via een omweg'via a detour', which is in a paradigm with the adjective rechtstreeks'directly', can be considered such a modifier. However, the fact that these examples can be paraphrased by means of the ... en doet dat + AP/PP'... and does that AP/PP' strongly suggests that we are dealing with VP adverbs; cf. adverb tests.

57
a. Jan rijdt rechtstreeks/via een omweg naar Groningen.
  Jan drives  directly/via a detour  to Groningen
b. Jan rijdt naar Groningen en hij doet dat rechtstreeks/via een omweg.
  Jan drives  to Groningen  and  he does that  directly/via a detour

      The elements midden, achter, voor, boven and onder, discussed in Section 3.1.3, cannot readily be used in combination with a directional adpositional phrase either. The examples in (58) are all unacceptable under the intended reading, where the located object is situated with respect to the reference object. The number sign indicates that (58b) is acceptable if boven is interpreted as “upstairs”; this reading, on which boven functions as an adverbial phrase that refers to the place where the event of throwing the picture into the cupboard took place, is irrelevant here.

58
a. Jan sprong (*midden) de plas in.
  Jan jumped     middle  the puddle  into
b. Jan gooide de foto (#boven/*onder) de kast in.
  Jan threw  the picture    above/under  the cupboard  into
  'Jan threw the picture in the top/lower drawer of the cupboard.'
c. Jan sprong (*achter/*voor) de auto op.
  Jan jumped  behind/in.front.of  the car  onto
[+]  II.  Modifiers of the implied path

Modification of directional adpositional phrases seems possible with modifiers that express to which extent the implied path is covered. Modifiers of this type come in two kinds: nominal measure phrases and adjectives. The adjectives helemaal'completely' and gedeeltelijk'partly' deserve some special attention.

[+]  A.  Nominal measure phrases and adjectival modifiers

Clear cases of modification of directional adpositional phrases are given in (59); note that the reference object can either precede or follow the modifier. Given that the reference object can also be placed in front of VP adverbs like snel'quickly' (Jan wandelde de berg snel op'Jan ascended the mountain quickly'), it seems plausible that the order in which the object precedes the modifier is derived by means of leftward movement of the noun phrase.

59
a. Jan wandelde <het bos> 3 km <het bos> in.
  Jan walked  the wood  3 km  into
  'Jan walked three kilometers into the woods.'
b. Jan wandelde <de berg> 3 km <de berg> op.
  Jan walked  the mountain  3 km  onto
  'Jan walked three kilometers up the mountain.'

In (60), we give similar examples involving an adjectival modifier; the choice of the modifier again depends on the nature of the reference object.

60
a. Jan wandelde <het bos> diep <het bos> in.
  Jan walked  the wood  deep  into
  'Jan walked into deep into the woods.'
b. Jan klom hoog de boom in.
  Jan climbed  high  the tree  into
  'Jan climbed high up into the tree.'

      The modifications in (59) and (60) do not involve the orientation or the magnitude of the vectors involved, but the implied path: it is expressed that a subpart of the implied path has been covered. The examples in (59), for example, presuppose a trajectory that goes into the wood/up the mountain and it is claimed that the located object has covered 3 kilometers of this trajectory. The difference between the directional constructions in (59) and the non-directional constructions in (61) is that in the latter Jan may return to his starting position after he has finished walking 3 kilometers, whereas in the former Jan must be located in the woods/up the mountain.

61
a. Jan wandelde 3 km in het bos.
  Jan walked  3 km in the wood
  'Jan walked three kilometers in the woods.'
b. Jan wandelde 3 km op de berg.
  Jan walked  3 km on the mountain
  'Jan walked three kilometers on the mountain.'

      The adjectives in the non-directional constructions in (62) differ from the directional ones in (60) in that they just specify the place where the event is taking place: (62a) expresses that the activity of walking took place deep in the woods, and (62b) that the activity of climbing took place high in the tree.

62
a. Jan wandelde diep in het bos.
  Jan walked  deep  in the wood
  'Jan was walking in the depth of the woods.'
b. Jan klom hoog in de boom.
  Jan climbed  high  in the tree
  'Jan was climbing high in the tree.'

      Modification of the type in (59) and (60) is only possible if the path is not intrinsically bounded, that is, if the starting and the endpoint of the path are not fixed. In (59a), this condition is met; even though the path must be situated somewhere in the woods, it is left implicit where the starting and endpoint of the path are situated; any position external to the woods is an appropriate starting point and any position internal to the woods is an appropriate endpoint of the implied path. Something similar holds for the path denoted by the postpositional phrase in (59b). That the boundedness of the paths is relevant can also be demonstrated by means of the contrast between the use of the directional PPs headed by the prepositions van'from', naar'to' and tot'until' in (63a), and the use of the directional PPs headed by the phrasal preposition in de richting van'towards' in (63b). The intuition is that the length of the path in (63a) is determined by some implied anchoring point, for instance, the position of the speaker. This means that the length of the implied path is contextually fixed, and therefore cannot be modified. In (63b), on the other hand, the starting and endpoint of the implied path are not given (neither explicitly nor implicitly), so that the length of the implied path is not contextually determined and modification is possible.

63
a. * Jan reed twee kilometer van/naar/tot Groningen.
  Jan drove  two kilometer  from/to/until Groningen
b. Jan reed twee kilometer in de richting van Groningen.
  Jan drove  two kilometers  towards Groningen

Other cases of directional phrases denoting paths that are inherently bounded and that therefore cannot be modified, are given in (64). In (64a-c), for example, the length of the implied paths is largely determined by the dimensions of the reference object; the starting and endpoint of the implied paths are bounded by two positions adjacent to and at opposite sides of the field/tunnel/house. Note that it is not clear whether the degraded example in (64d) can be accounted for in a similar way.

64
a. * Jan liep twintig meter het veld over.
  Jan walked  twenty meters  the field  across
b. * Jan liep twee kilometer de tunnel door.
  Jan walked  two kilometers  the tunnel  through
c. ?? Jan liep twee meter het huis voorbij.
  Jan walked  two meter  the house  past
d. ?? Jan liep drie kilometer het kanaal langs.
  Jan walked  three kilometers  the canal  along

      Finally, it can be noted that both the nominal and the adjectival modifiers can be extracted from the adpositional phrase by means of wh-movement. This is demonstrated in (65) by means of the interrogative counterparts of the examples in (59) and (60).

65
a. Hoeveel kilometer wandelde Jan het bos in?
  how.many kilometers  walked  Jan the wood  into
a'. Hoeveel kilometer wandelde Jan de berg op?
  how.many kilometers  walked  Jan the mountain  onto
b. Hoe diep wandelde Jan het bos in?
  how deep  walked  Jan the woods  into
b'. Hoe hoog klom Jan de boom in?
  how high  climbed  Jan the tree  into
[+]  B.  Helemaal'completely' and gedeeltelijk'partly'

The boundedness of the implied path is also relevant in the case of modification by helemaal/gedeeltelijk'completely/partly', which indicates whether the implied path is fully or partly covered. If the path is not inherently bounded, as in (66a), the use of these modifiers does not make sense and therefore results in unacceptability. Given the discussion of (59b) in the previous subsection, the grammaticality of (66b) may come as a surprise, but the difference in acceptability of (59b) and (66b) goes hand in hand with a difference in interpretation: in (59b) the endpoint of the implied path is left implicit in the sense that it can be situated anywhere on the mountain, whereas in (66b) the endpoint must be the top of the mountain. In other words, in the former case op is interpreted as “onto”, whereas in the latter case it is interpreted as “on top of”. If the “on top of” reading is not possible, as in (66b'), the judgments are as expected. Note that, in contrast to what is the case in (59b), the modifier in (66b) must follow the reference object (under neutral intonation of the sentence).

66
a. * Jan wandelde <het bos> helemaal/gedeeltelijk <het bos> in.
  Jan walked  the wood  completely/partly  into
b. Jan wandelde <de berg> helemaal/gedeeltelijk <*de berg> op.
  Jan walked  the mountain  completely/partly  onto
b'. * De supporter rende <het veld> helemaal/gedeeltelijk <het veld> op.
  the fan  ran    the field  completely/partly  onto

The pattern we find in (67) is more or less what we expect: the PPs in (67a) denote bounded paths and modification by helemaal is correctly predicted to be possible, although it is surprising that the use of gedeeltelijk gives rise to a somewhat marked result; the PPs in (67b) denote unbounded paths, and modification by helemaal/gedeeltelijk is correctly predicted to be impossible.

67
a. Jan reed helemaal/?gedeeltelijk van/naar/tot Groningen.
  Jan drove  completely/partly  from/to/until Groningen
b. * Jan reed helemaal/gedeeltelijk in de richting van Groningen.
  Jan drove  completely/partly  towards Groningen

The examples in (68) all involve inherently bounded paths, and the use of helemaal/gedeeltelijk is possible, as predicted. Note that in these cases, the modifier also follows the reference object (under neutral intonation of the sentence).

68
a. Jan liep <het veld> helemaal/gedeeltelijk <*het veld> over.
  Jan walked    the field  completely/partly  across
b. Jan liep <de tunnel> helemaal/gedeeltelijk <*de tunnel> door.
  Jan walked    the tunnel  completely/partly  through
c. Jan liep <het huis> helemaal/gedeeltelijk <*het huis> voorbij.
  Jan walked   the house  completely/partly  past
d. Jan liep <het kanaal> helemaal/gedeeltelijk <*het kanaal> langs.
  Jan walked    the canal  completely/partly  along

      For completeness' sake, it can be noted that the modification possibilities in (66) and (68) correlate nicely with the (im)possibility of using the adjective heel'whole' in (69) as an attributive modifier or predeterminer of the noun phrase expressing the reference object. There seems, however, to be a subtle difference in meaning between the two sets of constructions: whereas the examples in (66) and (68) suggest that the path proceeds along a more or less straight line, the examples in (69) suggest that the path proceeds in a more disordered fashion. This perhaps also accounts for the contrast between (69c&d) and (69e&f); while it is certainly possible to cross a field or proceed through a tunnel in a disorderly fashion, it seems less likely to pass a house of follow a canal in this way.

69
a. * Jan liep <heel> het <hele> bos in.
  Jan walked   whole  the   whole  wood  into
b. Jan liep <*heel> de <?hele> berg op.
  Jan walked    whole  the   whole  mountain  onto
c. Jan liep <heel> het <hele> veld over.
  Jan walked   whole  the   whole  field  across
d. Jan liep <heel> de <hele> tunnel door.
  Jan walked   whole  the   whole  tunnel  through
e. ? Jan liep <heel> het <hele> huis voorbij.
  Jan walked   whole  the   whole  house  past
f. ? Jan liep <heel> het <hele> kanaal langs.
  Jan walked   whole  the   whole  canal  along

The intuition that the implied paths in (69) are of a more disordered nature may be related to the fact that heel'whole' forces a distributive reading if used as an attributive modifier of the head of a reference object in a locational construction. Consider the examples in (70).

70
a. Er wonen mensen in het kasteel.
a'. Er wonen mensen in het hele kasteel.
  there  live  people  in the whole  castle
  'The (whole) castle is inhabited by people.'
b. Er liggen palen langs de weg.
b'. Er liggen palen langs de hele weg.
  there  lie  poles  along the whole  road
  'Poles are lying (all) along the road.'

Whereas (70a) is compatible with one family living in the castle (or with only a part of the castle being inhibited), example (70a') implies that the castle is divided into separate housing units; people are more or less evenly distributed in the castle. Similarly, (70b) could be used to express that one pile of poles is lying at the side of the road, whereas (70b') implies that the poles are placed along the road at certain intervals. In the same vein, it may be the case that the examples in (69) express that the implied path is more or less “evenly distributed” on the reference object; see Section N7.2 for more extensive discussion of the attributive modifier heel.

[+]  III.  A seeming case of modification

At first sight, the type of modification in (59a&b) seems of a similar sort to that found in the examples involving locational PPs in (48). This impression is getting even stronger when we consider the directional counterparts of the examples in (48) in (71).

71
a. Jan sloeg de spijker 3 cm de muur in.
  Jan hit the nail  3 cm  the wall  into
  'Jan hit the nail 3 cm into the wall.'
a'. Jan sloeg de spijker recht/schuin de muur in.
  Jan hit  the nail  straight/diagonally  the wall  into
  'Jan hit the nail straight/diagonally into the wall.'
b. Jan trok de spijker 3 cm de muur uit.
  Jan pulled  the nail  3 cm  the wall  out.of
  'Jan pulled the nail 3 cm out of the wall.'
b'. Jan trok de spijker recht/schuin de muur uit.
  Jan pulled  the nail straight/diagonally  the wall  out.of
  'Jan pulled the nail straight/diagonally out of the wall.'

As in the case of the prepositional phrases headed by in/uit, the nominal measure phrases of the postpositional phrases in the primeless examples of (71a&b) indicate to what extent the located object penetrates/protrudes from the wall; cf. Figure 6A&B. Similarly, the modifiers recht and schuin in the primed examples indicate in what way the nail penetrates/protrudes from the wall; cf. the discussion of Figure 6A'&B'. So, if we are dealing with modification involving distance and orientation, the adpositional phrases headed by in and uit must denote inward oriented vectors.
      It is, however, doubtful that we are really dealing in (71) with adpositional phrases denoting inwardly oriented vectors. Consider the examples in (72), subsection IIB, has shown that the modifiers in (72a) indeed seem to specify the part of the path denoted by the directional adpositional phrase covered by Jan, which is also clear from the fact that they cannot be used in the construction in (72b), where the PP acts as a locational (adverbial) phrase.

72
a. Jan wandelde de berg gedeeltelijk/helemaal/voor de helft op.
  Jan walked  the mountain  partly/entirely/halfway  onto
  'Jan walked partly/completely/halfway to the top of the mountain.'
b. * Jan wandelde gedeeltelijk/helemaal/voor de helft op de berg.
  Jan walked  partly/entirely/halfway  on the mountain

In (73a), on the other hand, the modifiers do not require the presence of a directional phrase; if we replace the directional postpositional phrase by a locational prepositional one, as in (73b), the result is still fully acceptable. This is due to the fact that the examples in (73) do not involve modification of the path covered, but modification of the part of the located object that has penetrated the reference object; in other words, we are not dealing with modification of the adpositional phrase but with modification or predication of the located object.

73
a. Jan sloeg de spijker gedeeltelijk/helemaal/voor de helft de muur in.
  Jan hit  the nail  partly/entirely/halfway  the wall  into
b. Jan sloeg de spijker gedeeltelijk/helemaal/voor de helft in de muur.
  Jan hit  the nail  partly/entirely/halfway  in the wall

A weaker argument against assuming that (73a) involves modification of the implied path is that it does not alternate with example (74), in which the adjective heel'whole' is used as a predeterminer or an attributive modifier of the noun phrase referring to the reference object; example (69) has shown that this is often possible if heel functions as a modifier of the implied path.

74
* Jan sloeg de spijker <heel> de <hele> muur in.
  Jan hit  the nail   whole  the   whole  wall  into
References:
    report errorprintcite