It has been suggested that Danish is a language particularly prone to spoken language reductions in spontaneous speech. Previous studies have shown that reduction phenomena, in Danish and other languages, are rule-governed by e.g. phonological context, word frequency and stress patterns. This paper analyses two reduction phenomena, those occurring in the endings -ede and -te in a genre of spoken Danish which is particularly resistant to reductions, viz. radio news readings. Its first aim is to establish the reduction rules of formal spoken Danish and compare these with the rules of more informal spoken Danish, e.g. sociolinguistic interviews. Reduction of -te is found to follow the same general rules as in spontaneous speech, although reductions are far less frequent in news readings. Reduction of -ede is found to follow rules different from those of spontaneous speech. The second aim is to investigate whether the reduction rules have changed over the 70 years which the data span. It is found that the rules, and thus the style, have indeed changed. The modern rules appear to be simpler and include less complex interaction effects.

The purpose of this article is to show that Norwegian has complex predicates, in which two verbs are reanalyzed as one predicate in a monoclausal structure, comparable to complex predicates that have been proposed for other languages. The central evidence comes from the construction called the long passive, in which the subject of the first verb is typically the patient of the second verb. Norwegian long passives often have passive morphology on both verbs, and I consider this a case of verbal feature agreement. The article also discusses evidence for complex predicates from active sentences.

This paper discusses criteria for distinguishing attributive (also known as adjectival) clauses and relative clauses, and argues that traditional approaches to these subclause categories and their interrelation lack consistency as well as empirical justification, from a modern Scandinavian perspective. Relative clauses are traditionally and in current Scandinavian reference grammars treated as a formally defined clause type with the attributive function as their prototypical, but not their only, constituent function. In this paper it is argued that the traditional criteria for identifying relative clauses, especially in Danish, Norwegian and Swedish, are essentially not formal but functional criteria, more or less identical to those applied to identify attributive clauses. Indeed, the modern Scandinavian languages, it is argued, hardly offer the empirical basis for a coherent category of relative clauses to be consistently distinguished on formal criteria. Rather, to account for the heterogeneous inventory of subclauses traditionally associated with relative clauses, a general outline of the functional category of attributive clauses in Danish is proposed, based on authentic spoken and written Danish discourse. Traditional definitions of attributive clauses are discussed and refined, and are shown to apply to a different and wider range of Danish subclause forms than has traditionally been recognised.

In a recent NJL article (Enger 2013), Hans-Olav Enger argues against some analyses of gender and ‘pancake sentences’, in particular against Josefsson (2009). In this short contribution, I will discuss what I take to be misunderstandings in Enger (2013). In addition I will discuss some data not included in Enger's (2013) analysis, which I will show to be crucial for the comparison between the different analyses proposed.