|
Verb movement in Scandinavian
In the spring and fall semesters of 2006 a group at the University of Tromsø has been working specifically on the syntax of verb movement across Scandinavian varieties. The semianar has now been replaced by a seminar/reading group on the Left Periphery
Go to this page for a list of references to the papers that we have discussed in the reading group.
Fall 2006
Thursday Oct 5th: Verb Movement
Topic: Types of embedded clauses allowing embedded V2
We'll be reading and discussing some relevant background literature, in particular Wechsler (1991). Here is otherwise a list of relevant papers provided by Marit Julien, Lund.
Thursday Oct 12th: Verb Movement
Topic: Types of embedded clauses allowing embedded V2
Empirical fieldwork within the group
(Wednesday Oct 18th/Thursday 19th: No meetings)
Thursday Oct 26th: Verb Movement
Topic: Types of subjects in embedded V2 clauses
Is it easier to get certain types of subjects with embedded V2 than others; e.g. specificity effects, definiteness effects, given vs. new info effects, etc.
This time, we'll do the empirical session within the group first, and then if anything interesting shows up, we'll look at this at the next meeting too.
(Wednesday Nov 1st: Scinfinitivals)
Thursday Nov 9th: Verb Movement
Topic: Types of subjects in embedded V2 clauses
(Wednesday Nov 15th: Scinfinitivals)
Thursday Nov 23rd: Verb Movement
Topic: The possibility of extraction out of embedded V2 clauses.
Empirical fieldwork within the group / discussion of relevant papers.
(Wednesday Nov 29th: Scinfinitivals)
Thursday Dec 7th: Verb Movement
Topic: The possibility of extraction out of embedded V2 clauses.
Empirical fieldwork within the group / discussion of relevant papers.
Spring 2006
Mánudagur 22. maí
May 11, Anna-Lena, Þorbjörg, Kristine and I gave a talk in the CASTL colloquium series called Spring Cleaning of Verb Movement in Norwegian and Icelandic.
(Revised) abstract:
The general picture of the distribution of verb movement in Scandinavian has long been that Icelandic displays verb movement to the inflectional domain independently of V2, whereas Norwegian does not. In this talk we will present data that support the opposite picture; Icelandic does not display independent verb movement to the inflectional domain, whereas Northern Norwegian dialects do. We will argue that both languages display remnant VP-movement to relatively high positions. Differences between the two languages will be shown to follow from two parameters:
(i) VP-movement targets high vs. low position
(ii) VP-movement is less vs. more remnant.
The handout can be downloaded here
-Gunnar Hrafn
Þriðjudagur 21. mars
A summary from the report after the meeting where we discussed Angantýsson (2001) and Kristine's Northern Norwegian data.
Embedded V2
Although the Scandinavian languages do not have generalised embedded V2, V2 (V-to-C) is available in certain embedded contexts. In Icelandic, this (roughly) seems to be the case for all embedded clauses introduced by 'that,' whereas in the Mainland Scandinavian languages (and English) this is (roughly) restricted to 'that'-clauses embedded under 'bridge' verbs. An indiciation that this is really embedded V2 is that non-subjects can be topicalised in such clauses, in which case we get subject-verb inversion.
Embedded V-to-I
However, there are other types of embedded clauses in which neither Icelandic nor the Mainland Scandinavian languages allow embedded V2 (i.e. non-subjects cannot be topicalised), e.g. relative clauses and embedded wh-questions. Usually, we find verb movement across medial adverbs and negation in these embedded clauses in Icelandic, whereas Standard Mainland Scandinavian never shows such movement of the verb.
Optional V-to- I in Icelandic and Northern Norwegian
Angantýsson (2001), however, has shown that verb movement seems to be optional in certain types of embedded clauses in Icelandic. More specifically, this verb movement is optional in exactly those cases where embedded V2 is not available (e.g. relative clauses and embedded wh-questions). Thus, the movement that is optional here, is of the V-to-I type (rather than V-to-C). This means that when embedded V2 is possible, the verb has to move across sentence medial adverbs and negation, whereas such movement is not necessary if embedded V2 is not an option. (For Icelandic, at least for some speakers, the necessity of verb movement seems to correlate with the presence of the complementizer að 'that'.)
As Bentzen (2005) shows, Northern Norwegian (NN) has optional verb movement at least across lower adverbs (but never across negation) in non-V2 embedded contexts (e.g. relative clauses and embedded wh-questions). As in Icelandic, V2 is not available in NN here, so again the optional verb movement is of the V-to-I type (rather than V-to-C). In embedded V2 contexts, the verb has to move (V-to-C) when a non-subject is topicalised (as in Icelandic). In subject-initial embedded V2 contexts, verb movement past medial adverbs and negation is optional in NN. Note that verb movement past negation is an option here, contrary to what is found in 'true' non-V2 contexts in NN.
Summing up
Two types of movements (at least!) are involved:
Embedded V2 (V-to-C):
* Obligatory in both Icelandic and NN in embedded V2 contexts (e.g. 'that'-clauses) when a non-subject is topicalised.
Embedded V-to-I (movement past sentence medial adverbs and negation):
* Obligatory in Icelandic embedded V2 contexts ('that'-clauses)
* Optional in NN embedded V2 contexts ('that'-clauses)
* Optional in Icelandic embedded non-V2 contexts (e.g. relative clauses and embedded wh-questions)
* Optional in NN embedded non-V2 contexts (e.g. relative clauses and embedded wh-questions); BUT only available past sentence medial adverbs, not past negation.
In both Icelandic and in NN the 'exceptional' cases are the non-V2 embedded contexts (e.g. relative clauses and embedded wh-questions):
(i) Icelandic - generally assumed to have obligatory embedded V-to-I movement - unexpectedly has optional V-to-I movement (past sentence medial adverbs and negation) here.
(ii) NN - generally assumed to not allow embedded V-to-I movement - has optional V-to-I movement here, BUT this movement is restricted to only apply past sentence medial adverbs, not past negation.
-Gunnar Hrafn and Kristine
Fimmtudagur 16. mars
Kristine Bentzen and I are working on a report from the meeting where we discussed the differences/similarities between Icelandic and Northern Norwegian when it comes to verb movement in embedded clauses. As soon as we finish it, it will be available here.
-Gunnar Hrafn
|
