Double object constructions: active verbs

Björn Lundquist

UiT The Arctic University of Norway

1. Introduction

In the ScanDiaSyn-survey, certain aspects of double object constructions were investigated. For double object verbs in the active diathesis, the focus was on non-selected or "free" indirect objects. More specifically, the question focused on was to which extent non-prototypical ditransitive verbs can take a recipient arguments realized as noun phrases in a position before the direct object. The general syntax of double object constructions was however not investigated in a great detail. I will therefore first give a short overview of possible word orders in double object constructions in North Germanic, before moving on to the results concerning the non-selected indirect objects in section 2.

In Mainland Scandinavian, the indirect object in general precedes the direct object in the verb phrase. The indirect object can also be realized as a prepositional phrase, following the direct object.[1] The standard pattern is shown for Swedish below:

(1) a)
Jag har gett mannen boken. (Swe.)
I have given man.def book.def
'I have given the man the book.'
b)
*Jag har gett boken mannen. (Swe.)
*I have given book.def man.def
*'I have given the man the book.'
c)
Jag har gett boken till mannen. (Swe.)
I have given book.def to man.def
'I have given the book to the man.'
In the ScanDiaSyn survey, the robustness of the pattern was investigated only for Danish, where the following sentence was tested and universally rejected:

(2)
*Ib har jo givet boken Lise. (#1030) (Dan.)
*Ib have jo given book.def Lise.
Int. ‘Ib has given Lisa the book.’

In Swedish, there is a very small number of verbs, possibly just one or two, that allow for inversion of the two objects (see Lundquist 2004 for discussion):

(3) a)
Stevie Wonder tillägnade konserten sin hustru. (Swe.)
Stevie Wonder dedicated concert.def his wife.
'Stevie Wonder dedicated the concert to his wife.'
b)
Stevie Wonder tillägnade sin hustru konserten. (Swe.)
Stevie Wonder dedicated his wife. concert.def.
'Stevie Wonder dedicated the concert to his wife.'
In Icelandic, a fairly large number of double object verbs allow both orders of indirect and direct objects (example from Thráinsson 2007):

(4) a)
Hann gaf konunginum ambáttina. (Ice.)
He gave king.def.dat maidservant.def.acc
‘He gave the king the maidservant.’
b)
Hann gaf ambáttina konunginum. (Ice.)
He gave maidservant.def.acc king.def.dat
‘He gave the maidservant to the king.’

The inverse order (4b) is mainly available for double object verbs that assign dative case to the indirect object, and accusative case to the direct object. The order is stricter for verbs with other case-frames (where indirect objects have to precede direct objects). For verbs like ‘give’, the option of realizing the indirect object in a PP is not available in Icelandic.

The pattern in Faroese is more like that in Mainland Scandinavian (even though Faroese still have quite a lot of overt case marking). The inverted order is not available, while the PP-option apparently is getting more and more common (Zakaris Hansen, p.c.):

(5) a)
Hon gaf Mariu troyggiuna. (Far.)
she gave Mariu sweater.def.acc
‘She gave Maria the sweater.’
b)
Hon gaf troyggiuna till Mariu. (Far.)
she gave sweater.def.acc till Mariu
‘She gave the sweater to Maria.’
c)
*Hon gaf troyggiuna Maria. (Far.)
*She gave sweater.def.acc Mariu
* Int. ‘She gave the sweater to Maria.’
In the Nordic Dialect Corpus, the PP option can be found in Faroese, but not the inverted order (for the verb geva, ‘give’). An example of the PP order is given in (6) below:

(6)
So góvu vit tað bara till ketturnar. (Far.)
so gave we it only to cat.pl.def
‘so we gave it only to the cats’

Thus, none of the North Germanic varieties regularly make use of all three options in (5). (5a) is available in all North Germanic varieties, (5b) is accepted everywhere except Iceland, and (5c) is only available in Icelandic (and in Swedish, for a small number of verbs).

As mentioned above, the possibility of DO-IO order was tested only in Denmark in the ScanDiaSyn-survey. Instead, the focus was on the availability of unselected indirect objects, i.e., to which extent non-prototypical ditransitive verbs can take a recipient argument realized as a noun phrase in a position before the direct object. As we will see, the restrictions on unselected indirect objects are much stricter in the Scandinavian languages than in e.g. English and German. Several issues regarding passive double object verbs were also investigated in the ScanDiaSyn survey, and the results are presented in a separate chapter (Lundquist 2014).

2. Results

2.1 Nordic Syntactic Database (NSD)

In Norway, Sweden, Finland and the Faroe Islands, a sentence containing a verb of creation and an indirect object was tested. The sentence tested in Norway, Sweden and Finland is given in (8a) and the sentence tested in the Faroe Islands is given in (8b):

(8) a)
Han bakte gjesten en kake. (#444) (Nor.)
He baked guest.def a cake.
‘He baked the guest a cake.’
b)
Omman bant gentuni eina troyggju. (#444) (Far.)
Grandma.nom knit.past girl.sg.def.dat a sweater.acc.
‘Grandma knitted the girl a sweater.’

The results are shown in Map 1 below:

Map 1. Unselected indirect object with verb of creation (#444)
Map 1: Unselected indirect object with verb of creation
(#444: Han bakte gjesten en kake.'He baked the guest a cake.'/Omman bant gentuni eina troyggju. 'Grandma knitted the girl a sweater.')
(White = high score, grey = medium score, black = low score).

As can be seen in the Map 1, free indirect objects with verbs of creation is fully acceptable in the Faroe Islands. In Mainland Scandinavia, the tested construction is more accepted in the northern parts, which we will return to in the discussion section.

In Norway, Sweden and Finland, the availability of indirect objects with verbs of fetching was also tested. Note that the indirect object comes in the shape of a pronoun here.[2]

(9)
Han hentet henne en stol. (#447) (Nor.)
he fetch.past her a chair
‘He fetched her a chair.’

The results are shown in Map 2 below:

Map 2. Unselected indirect object with verb of fetching, pronominal indirect object (#447)
Map 2: Unselected indirect object with verb of fetching, pronominal indirect object
(#447: Han hentet henne en stol. 'He fetched her a chair.')
(White = high score, grey = medium score, black = low score).

As we see, this (#447) is much more accepted in than (#444) in Mainland Scandinavia. In Norway, the test sentence is mainly accepted in the northern parts, but at most locations in Sweden. In Sweden and Finland, the sentence with a verb of ballistic motion was tested as well. Here, the indirect object again is in the shape of a pronoun:

(10)
Han kastade henne bollen. (#1403) (Swe.)
hr throw.past her ball.past
‘He threw her the ball.’

The results are shown in Map 3 below:

Map 3. Unselected indirect object with verb of ballistic motion, pronominal indirect object (#1403)
Map 3: Unselected indirect object with verb of ballistic motion, pronominal indirect object
(#1403: Han kastade henne bollen. 'He threw her the ball.')
(White = high score, grey = medium score, black = low score).

As can be seen in the map above, test sentences is rejected in most locations. The interesting exceptions are two adjacent measure points in northern Sweden (Kalix and Överkalix), where the sentence gets overall high scores. The informants at these measure points also accepted other non-selected indirect objects, as can be seen in Map 1 and 2 above.

3. Discussion

In most North Germanic varieties the set of verbs that allows two nominal objects is smaller than in e.g. English; see Barðdal et al. (2011) for an overview. For example, verbs of production, bringing, and ballistic motion (like throw) are usually not accepted with two nominal objects. This has been investigated by Sæther (2001) for the Norwegian-speaking area, and she found some dialectal variation: northern Norwegian speakers allow double object constructions for many more verbs than speakers in southern Norway do. A strong difference was found for the following verbs bake ‘bake’, bestille ‘order’, finne ‘find’, hente ‘fetch’, kjøpe ‘buy’, lage ‘cook’, lese ‘read’, ordne ‘arrange’, skjære ‘cut’, skrive ‘write’, smøre ‘butter’, steke ‘fry’, strikke ‘knit’ and ta ‘take’. The results from the ScanDiaSyn-survey confirm Sæther’s findings, most clearly for the verb hente (447). The creation verb bake is however not as generally accepted as a double object verb, and the dialectal pattern is less obvious. When taking only younger speakers into account, the dialectal difference is clearer, as shown in the Map 4 below, where only the judgments from younger speakers are shown for sentence (#444). Map 4 can be contrasted with Map 5, where the judgement for sentence (#444) from older informants are shown. As can be seen, a ditransitive construal of bake ‘bake’ is accepted by far less older speakers than younger speakers, at least in Norway.

Map 4. Unselected indirect object with verb of ballistic motion, pronominal indirect object (#1403)

 

Map 4: Unselected indirect object with verb of creation, younger speakers
(#444: Han bakte gjesten en kake. 'He baked the guest a cake.')
(White = high score, grey = medium score, black = low score)

Map 5. Particle following NP object (#1423)

 

Map 5: Unselected indirect object with verb of creation, older speakers
(#444: Han bakte gjesten en kake. 'He baked the guest a cake.')
(White = high score, grey = medium score, black = low score)

The variation is less obviously geographically determined in the Swedish-speaking area. It is however clear that the acceptance for hämta ‘fetch’ is generally higher in Sweden than in southern Norway, and that baka ‘bake’ is hardly accepted at all with an indirect object.

It should be noted that the form of the indirect object tends to influence the grammaticality judgments. Most notably, pronominal indirect objects tend to be more acceptable as indirect objects than noun phrases. The difference we find between (#444) and (#447) hence may be triggered by the form of the indirect object rather than by the nature of the verb. The results from Sæther’s (2001) dialect study of double object construction suggests that the form of the indirect object is rather the relevant factor here, since the informants in her study accepted bake ‘bake’ to an almost equal degree as hente ‘fetch’.

As seen in Map 3, the verb kasta ‘throw’ is in general not accepted by Swedish speakers. The only part of Sweden where this sentence gets an overall high score is in the very northeast of Sweden. Note that the same locations that had high scores for the verb baka with two objects also have high scores for kasta with two objects. It is possible that double object constructions are available for more verbs in northern Sweden, just as it is in northern Norway. The results for sentence (#1403) are in accordance with previous reports on the verb kasta, and other verbs of ballistic motion (see e.g. Lundquist and Ramchand 2010, Barðdal et al. 2011). Barðdal et al. (2011) reports that the double object construction is not available with verbs of ballistic motion (like ‘throw’) in any of the Scandinavian languages. However, as reported by Falk (1990), a sentence like (#1403) improves for some Swedish speakers when the direct object is indefinite (‘he threw her a ball’). Note that this is true for verbs of creation (like ‘bake’ and ‘build’) as well, but that we still have reason to believe that Mainland Scandinavian speakers find double object constructions more marked with verbs of ballistic motion than with verbs of production. Further, verbs of ballistic motion never take indirect objects in Icelandic.

One difference between the Mainland North Germanic languages that was pointed out already by Hulthén (1947) is that verbs of communication in Swedish, like berätta ‘tell’ and förklara ‘explain’ cannot take an indirect argument, while they can in Danish and Norwegian.

References

Barðdal, Jóhanna, Kristian E. Kristoffersen & Andreas Sveen. 2011. ‘West Scandinavian Ditransitives as a Family of Constructions: With a Special Attention to the Norwegian V-REFL-NP Construction'. Linguistics 49(1): 53-104.

Brøseth, Heidi. 1997. Dobbelt objekt og tilgrensende kronstruksjoner i norsk. Master’s thesis, Institutt for nordistikk og litteraturvitenskap, Norges Teknisk-Naturvitenskapelige Universitet.

Falk, Cecilia. 1990. ‘On double object constructions'. Working Papers in Scandinavian Syntax 46: 53–100.

Hulthén, Lage. 1947. Jämförande nunordisk syntax II. Elanders Boktryckeri Aktiebolag, Göteborg.

Lundquist, Björn. 2004. Subjektsval vid passivering av ditransitiva verb. Master’s thesis, University of Lund.

Lundquist, Björn. 2014. ‘Double object contstructions: passive verbs.' Nordic Atlas of Linguistic Structures (NALS) http://www.tekstlab.uio.no/nals#/chapter/8.

Lundquist,Björn and Gillian Ramchand. 2012. ‘Verbs of contact in English, Swedish and German.' In Proceedings from CGSW 23, edited by Rhona Alcorn, Peter Ackema, and Caroline Heycock. John Benjamins, Amsterdam.

Sæther, Kari-Anne. 2001. Dialektfoskjeller i realiseringen av indirekte objekt. Master's thesis, University of Tromsø.

Teleman, Ulf, Staffan Hellberg, and Erik Andersson. 1999. Svenska akademiens grammatik. Svenska Akademien, Stockholm.

Thráinsson, Höskuldur. 2008. The Syntax of Icelandic. Cambridge University Press, Cambridge.


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] The Swedish Reference Grammar (Teleman et.al 1999/3:299), gives a list of verb types for which the indirect object cannot be realized as a preposition phrase. They mention verbs with a prefixed/incorporated preposition e.g. tilldela, ‘assign’, and some verbs for which the indirect object is inerpreted as a Malefactive, Addressee or an Affected particicipant. However, most verbs they mention can take prepositional indirect objects in the right context, for example if the indirect object is heavy/long and the direct object is short, as in the following example (modified from the Parole corpus): Stiftelsen tilldelar sex stipendier till doktorander vid Göteborgs universitet ‘The foundation assigns six stipends to Ph.D. students at the University of Gothenburg’. As far as I am aware, the only time you cannot realize the indirect object as a PP, is when the direct object is an infinitival clause, and the indirect object is interpreted as the implicit subject of this clause, as in de befallde honom att gå hem ‘they commanded him to go home’. Examples of this kind might be better analyzed as object control clauses than double object constructions.

[2] In parts of Norway, the sentence han hentet henne et glass vannF ‘he fetched her a glass of water’ (#1389) was tested in addition to sentence (#447). The results were largely the same for this sentence and sentence (#447).