Main/embedded clause asymmetries in the Scandinavian languages (MECAS)

In two papers from the 80s, Christer Platzack demonstrates that even though rules that help distinguish embedded clauses from main clauses can be formulated, there are clauses that share properties of both. In the project, the relevant generalizations reached since the 80s will be gathered and reanalysed from the viewpoint of what we know about grammar today.

Consider the five descriptive characteristics used in grammar classes to help students identify embedded clauses in Swedish:

1. It cannot appear alone.
2. Sentential adverbs appear before the finite verb.
3. The auxiliary ha can be omitted.
4. It is introduced by a subordinate conjunction (complementizer).
5. The conjunction is immediately followed by the subject.

How do we classify exlamative clauses of the kind Att han inte kom!? These appear alone, thus qualify as main clauses by (1) but pattern with embedded clauses w.r.t. (2)-(5). Embedded V2 clauses appear to be main clauses with respect to (2) but embedded clauses with respect to (4) and (5). Classification w.r.t (1) depends on whether the subjunction should be ignored and classification w.r.t. (3) poses yet another problem; if the auxiliary is omitted, we have no finite verb to determine classification w.r.t. property (2). To complicate the matter further, the subordinate conjunction can be dropped. We will gather the important generalisations reached with regard to the above properties and analyze them from the viewpoint of what we know about grammar today. We will investigate the syntax and semantics of embedded clauses from synchronic and diachronic viewpoints and look at important factors of variation in Swedish and Scandinavian more generally. We seek to reach adequate descriptions of (degrees of) embedding and further understanding of the principles that govern cases that appear to fall in between