This article studies changes in the Swedish double object construction (DOC), focusing on fourteen verb-specific constructions (see Croft 2001). The study uses two frequency measures – lexical variation and text frequency. The former measures the type frequency of nouns in the direct object slot of these verb-specific constructions and can be taken as a rough measure of productivity, while the latter deals with the token frequency of the verb-specific constructions as well as of the verbs in general. Most of the verb-specific constructions undergo a decrease in text frequency, mirroring the general decrease in text frequency of the DOC. Four of them show a decrease in lexical variation, indicating a connection between a decrease in lexical variation and a decrease in text frequency.

This thesis examines the use of the Swedish double object construction (the DOC) and compares this with the use of three semantically equivalent prepositional object constructions (POCs): the till-POC, the åt-POC and the för-POC. The thesis has a diachronic perspective, investigating changes in the use of these four constructions between 1800 and 2016. The theoretical framework employed is usage-based construction grammar. The overall aim of the thesis is to shed light on the use of the DOC and the three POCs in Swedish during the selected period and to incorporate these findings into the general theoretical discussion in usage-based construction grammar on double object constructions and dative alternations. In addition, the thesis aims to determine the theoretical status of the double object construction in present-day Swedish. Ditransitives are an extensively studied topic in construction grammar. However, ditransitives in Swedish have not previously been the subject of any major study. The present study is corpus-based, drawing on prose fiction data from 1800–1999 and blog data from 2016. The data were collected from the National Language Bank of Sweden, using the search interface Korp (Borin et al. 2012).

The study shows that the DOC has decreased in text frequency. The DOC has also undergone a decrease in the lexical richness of the verb slot, which means that the construction is used with fewer verbs in present-day Swedish compared to nineteenth-century Swedish. Furthermore, the semantic range of the DOC has decreased. An analysis of a number of selected verb-specific and verb-class-specific constructions also revealed that some verbs undergo a decrease in the lexical richness of the direct object slot, when used in the DOC. The relations between the DOC and the three POCs have also changed considerably since the early nineteenth century. In present-day Swedish, verbs that are attested in the till-POC and the för-POC make up a larger proportion of the verbs found in the DOC compared to the nineteenth century. Thus, the overlap between the constructions has increased. This is the consequence of an increased use of both the till-POC and the för-POC, combined with the decreased use of the DOC. Both the till-POC and the för-POC are used with a wider range of verbs in present-day Swedish compared to nineteenth-century Swedish. The proportion of åt-POC verbs attested in the DOC remains stable over time. However, the semantic range of the åt-POC has decreased during the investigated period. The analysis of selected verb-specific and verb-class-specific constructions revealed that the till-POC appears to be taking over functions from the åt-POC.

The results are discussed in the light of different models drawn from usage-based construction grammar, including semantic maps, taxonomic networks and horizontal links. I argue that the DOC and the åt-POC have become less schematic, whereas the till-POC has become more schematic. The för-POC, I argue, has retained its level of schematicity but appears to be used more liberally in present-day Swedish. The fact that there is an increased overlap between, on the one hand, the DOC and the till-POC and, on the other, the DOC and the för-POC, indicates strengthened horizontal links between the DOC and these two POCs. Finally, I argue that the decreased use of the DOC might to some extent be the consequence of a change in stylistic norms in the prose fiction genre.

This article presents an attempt to interpret the Viking Age runic inscription Vg 199 Norra Lundby based partially on a new reading. The inscription has hitherto been considered to lack linguistic meaning. The interpretation runs as follows: [Stæinn] kveðsk hēr[si] s[tand]a længi. Kveðsk Hialms Gunna hald[a …] … …ṃiþ : ka… … Gott orð bar. ‘The stone proclaims that it will long stand here. (It) proclaims that it will preserve Hialmʀ’s Gunni … …ṃiþ : ka… . (He) had a good reputation.’ Since a relatively large part of the stone is missing, some of the runic sequences associated with the lacunae cannot be inter­preted with certainty. A possible but purely speculative interpretation is none­theless offered. It is in any case beyond doubt that the inscription is in fact linguistically meaning­ful. Vg 199 bears a strong resemblance to the inscription on the Jut­landic stone DR 131 Års, and just like that one, it is most likely formu­lated in verse. Remarkably, Vg 199 lacks a proper memorial formula. This suggests that the accompanying memorial formulation was inscribed on a separate stone.