Verb-particles: particle placement in passives

Björn Lundquist

UiT The Arctic University of Norway

1. Introduction

As discussed in the chapter on particle placement in active clauses (Lundquist 2014a), there is some interesting variation within Scandinavia with respect to particle placement in active clauses. In participial passive clauses, the verb particle always prefixes to the participle in Swedish, while it follows the participle in Danish and Icelandic, see e.g. Johannisson (1939) and Svenonuis (2005) for details. In Standard Norwegian and Faroese, the particle most often follows the participle, but in certain contexts, the particle can also prefix to the participle, just like in Swedish. There are also dialects of Norwegian where particles more freely prefix to participles, most notably the dialect spoken in Romsdal, as discussed in detail in Sandøy (1979). The following examples show the two possible particle placements in Scandinavian: (1) from Danish has the post-participial particle, and (2) from the Romsdal dialect has an incorporated particle (from Sandøy 1979):

(1)
Han blev smidt ut af baren. (Dan.)
he bli.past thrown out of bar.def
'He was thrown out of the bar.'
(2)
So blei det opphengt til tørk. (Rom.)
so bli.past it up.hanged to dry.
'Then it was hanged up to dry.'

In the Scandinavian languages, we find two types of passives: a periphrastic passive, which consists of a (passive) past participle and an auxiliary (usually bli, as in the examples above), and a synthetic passive in which the passive marker -s or -st is suffixed to the verb (following other inflectional material). With respect to particle placement, synthetic passives generally behave just like the corresponding active sentences: particles do not incorporate into synthetic passives in Scandinavian.[1] However, the Romsdal dialect is an exception, and here particle prefixation can take place both in passive participles and infinitival synthetic passives (Helge Sandøy, p.c.).[2]

(3)
Datorn ska självklart stängas av efter användning. (Swe.)
computer.def shall evidently switch.inf.pass off after use
'The computer should of course be switched off after use.'
(4)
Etter bruk skal datamaskina sjølvsagt tåslåast. (Rom.)
after use shall computer.def evidently off.switch.inf.pass
'The computer should of course be switched off after use.'

It is noteworthy that Swedish, which has obligatory particle shift in active clauses, also has a prefixed particle in passives, while Danish, which disallows particle shift in active sentences, always seem to have the particle at the end of the verb phrase. Also the Romsdal dialect, which to a large extent has incorporated particles in passives, has a high proportion of shifted particles in active sentences, which suggest that there is a correlation between shifted particles in active clauses and incorporated particles in periphrastic passives. As discussed in Svenonius (2005), participle agreement also influences the placement of the particle: the particle is more likely to be incorporated when the participle shows overt number and gender agreement. Agreement on participles is discussed in a separate chapter in NALS (Larsson 2014). In the discussion below, agreement is largely ignored. The nature of the passive construction also influences particle placement, and when the participle is more adjectival, the particle is more likely to be prefixed, as we will see below.

In Sweden, Finland and Norway, particle placement in event-denoting/verbal passives (with the auxiliary bli) and result- or state-denoting passives (with the auxiliary vara) were tested. Particle placement in event-denoting passives was also tested in Faroese.

2. Results

2.1 Nordic Syntactic Database (NSD)

The following sentences tested the availability of particle prefixation in event-denoting passives:

(5) a)
Radion blev utsliten (ur bilen). (767) (Swe.)
radio.def bli.past out.pull.part from car.def
'The radio was pulled out of the car.'
b)
Tei gomli húsini blivu niðurrivin. (767) (Far.)
the old house.pl.def bli.past down.tear.part
'The old houses were torn down.'

The results are shown in Map 1 below:

Map 1: Prefixed particle in event-denoting passives  (#767)
Map 1: Prefixed particle in event-denoting passives
(#767:Radion blev utsliten (ur bilen)./ Tei gomli húsini blivu niðurrivin.'The radio was pulled out of the car.'/'The old houses were torn down.' )
(White = high score, grey = medium score, black = low score)

 

Map 1 shows that prefixation is possible in event-denoting passives all over the Swedish speaking area (with some exceptions), on the Faroe Islands, and in central and western Norway. These results should be compared to the results for the sentences with post-participial particles (6 a-b below), as shown in Map 2.

(6) a)
Radion blev sliten ut (ur bilen). (766) (Swe.)
radio.def bli.past pull.part out from car.def
'The radio was pulled out of the car.'
b)
Tei gomli húsini blivu rivin niður. (766) (Far.)
the old house.pl.def bli.past tear.part down
'The old houses were torn down.'

Map 2. Post-participial particle in event-denoting passives (#766)
Map 2: Post-participial particle in event-denoting passives
(#766: Radion blev sliten ut (ur bilen)./Tei gomli húsini blivu rivin niður. 'The radio was pulled out of the car.'/'The old houses were torn down.').
(White = high score, grey = medium score, black = low score)

Sentence (6) gets high and intermediate scores in several locations in Sweden and Finland. In Sweden, the high scores are mainly found in the southern and the western parts. Sentence (6) receives higher scores than sentence (5) (prefixed particle) at two locations in Sweden, Våxtorp in Halland, and Särna in Dalarna, but otherwise the prefixed participle is either preferred or the only version available. The post-participial particle is in general accepted in Norway and in the Faroe Islands. Only in two locations in Norway (Gausdal in Oppland and Oppdal in Sør-Trøndelag) did Sentence (5) get higher scores than sentence (6).

The following two sentences were used to test particle placement in resultative passives, i.e. passives with the auxiliary vara rather than bli.

(7)
Radion är utsliten (ur bilen). (765) (Swe.)
radio.def is.pres out.pull.part from car.def
'The radio was pulled out of the car.'

(8)
Radion är sliten ut (ur bilen). (764) (Swe.)
radio.def is.pres pull.part out from car.def
'The radio was pulled out of the car.'

The results are shown in Map 3 and Map 4 below:

Map 3. Prefixed particle in resultative passives (#765) Map 3: Prefixed particle in resultative passives
(#765: Radion är utsliten (ur bilen). 'The radio was pulled out of the car.')
(White = high score, grey = medium score, black = low score)


Post-participial particle in resultative passives (#764)
Map 4: Post-participial particle in resultative passives
(#764: Radion är sliten ut (ur bilen). ' The radio was pulled out of the car.')
(White = high score, grey = medium score, black = low score)

The difference between Swedish and Norwegian comes out most clearly in the judgements of (8); see Map 4 which shows that free particles are not accepted in Swedish (with very few exceptions), while they are accepted throughout Norway (again, with very few exceptions). Maps (1) and (3) show that prefixed particles are accepted in quite a large part of Norway. When comparing maps (2) and (4), we also see that free particles are sometimes judged as more acceptable in eventive passives than in stative passives in Sweden, and never the other way around.

3. Discussion

When comparing the results from the test sentences above with the data on particle placement in active sentences (Lundquist 2014a), we can observe that dialects that prefix particles to participles tend to have particle shift in active sentences to a large degree. The places in Norway that show the highest acceptance of prefixed particles in event-denoting passives (3), are largely the same locations where we find a high number of particles preceding objects, even pronominal objects, in the Nordic Dialect Corpus (see section 3 in Lundquist 2014a). However, we see that free (post-participial) particles are more acceptable in the passive paradigm in Swedish, than are particles following the object in active clauses. This indicates that there are grammars that allow for free particles in passive clauses, but allow the order particle-object in active clauses. We can then tentatively conclude that all varieties of Scandinavian that allow for particle shift in active clauses also allow for prefixed particles in eventive passives, but that varieties that obligatorily have particles preceding the object in active clauses, sometimes allow for free particles in participial passives.

As pointed out in the introduction above, Danish never allows for particles prefixed to passive participles, unless the particle can be prefixed to the active verb as well. Danish thus looks like the Norwegian dialects spoken in the northernmost and southernmost parts of Norway. However, as far as I am aware, in all Norwegian varieties, a particle can be prefixed when the participle is used attributively. The particle can never follow the participle in this case:

(9) a)
ei utreven side på. gulvet. (Nor.)
a out.tear.part page lay on floor.def
b)
*ei reven ut side gulvet. (Nor.)
 a tear.part out page lay on floor.def
*'A torn-out page lay on the floor'

Note that no Scandinavian variety allows the "English" word order (which is not necessarily fully acceptable in English either at least not in less adjectival contexts), with a particle following an attributive participle as shown in the translation line in the example above. However, Danish seems to disallow both prefixed particles and post-participial particles in attributive position, unless the parti cle-participle combination is highly lexicalized, or if the particle can prefix even in active contexts. As shown in (10b-c), the particle inde ‘in’ can neither prefix to nor follow the attributive participle. (10a) gives the regular passive form, and in (10d) the Norwegian attributive version is given (Danish examples from Tanya Karoli Christensen, p.c., and Norwegian examples from Christine Bjerkan Østbø, p.c.):

(10) a)
manden blei låst inde i kælderen. (Dan.)
man.def bli.past lockpart in in basement.def
'The man was locked up in the basement'
b)
*den ind(e)låste mand (Dan.)
 the in.lock.part man
*'the locked up man'
c)
*den låst inde mand (Dan.)
 the lock.part in man
*'the locked up man'
d)
den inn(e).låste mannen (Nor.)
the in.lock.part man.def
'the locked up man'

On the other hand, Danish have many more verbs that can take a prefixed particle in the active forms than Swedish and Norwegian, and these verbs can form participles that can be used attributively. For example, the following Danish prefixed verbs have to take a free particle in Swedish in the active paradigm: udskrive ‘out-write’ (‘print out’), udfylde ‘fill out’, indesprærre ‘in-shut’ (‘lock up’), opgive ‘give up’. Particle prefixation is still lexically restricted in Danish though, and free particles are probably more common than bound particles (see Herslund 1984 for discussion).

References

Herslund, Michael. 1984. 'Particles, prefixes, and preposition stranding,' Topics in Danish Syntax (Nydanske Studier & Almen Kommunikationsteori 14), Akademisk Forlag, København.

Johannisson, Ture. 1939. Verbal och postverbal partikelkomposition i de germanska språken. Diss. Lindstedt, Lund.

Lundquist, Björn. 2014a. 'Verb particles: active verbs.' Nordic Atlas of Linguistic Structures (NALS) http://www.tekstlab.uio.no/nals#/chapter/4.

Larsson, Ida. 2014. 'Expletives and agreement in passives.' Nordic Atlas of Linguistic Structures (NALS) http://www.tekstlab.uio.no/nals#/chapter/58.

Sandøy, Helge. 1976. Laust samansette verb i vestnordisk: ein samanliknande leddstillingsanalyse for islandsk, færøysk og romsdalsmål. Universitetet i Oslo, Oslo.

Svenonius, Peter. 2005. 'The verb-particle alternation in the Scandinavian languages'. Handout, Leikanger, available at: http://norms.uit.no/thematic/workshops/ScanVPrtHOLeikanger.pdf.


Web sites:

Nordic Atlas of Language Structures (NALS) Journal: http://www.tekstlab.uio.no/nals

Nordic Dialect Corpus: http://www.tekstlab.uio.no/nota/scandiasyn/index.html

Nordic Syntax Database: http://www.tekstlab.uio.no/nota/scandiasyn/index.html



[1] Note that some verbs have optional particle incorporation even in other tenses and voices, e.g. avfyra - fyra av ‘off-fire’ -‘fire off’. For these verbs, the particle can of course be incorporated in the passive as well. As will be returned to in the discussion part, this phenomenon is probably more common in Danish than the other Scandinavian varieties.

[2] Note that the synthetic passive is only available in the infinitive in the Romsdal dialect and many other Norwegian dialects.