Main/embedded clause asymmetries in the Scandinavian languages (MECAS)
April 14-15, 2011
The background of this workshop is best described by a couple of papers written by Christer Platzack in the 80s, where he demonstrates for Swedish that even though it is possible to formulate characteristic features or rules that help us distinguish finite embedded clauses from main clauses, there are also clauses that share properties of both and therefore are not as easily classified (see also Lars-Gunnar Andersson 1975). By way of illustration, consider the five descriptive characteristics frequently used in grammar classes to help students identify embedded clauses in Swedish (Platzack 1987):
(i) It cannot appear alone,
(ii) sentential adverbs appear before the finite verb,
(iii) the auxiliary ha (have) can be omitted,
(iv) it is introduced by a subordinate conjunction (complementizer),
(v) the conjunction is immediately followed by the subject.
How do we classify e.g. exlamative clauses of the kind "Att han inte kom!" (that he not came) according to these criteria? These appear alone, thus qualify as main clauses by (i) but pattern with embedded clauses with regard to the remaining properties (ii)-(v). Embedded verb second clauses constitute another case in point; "Han sa att han hade inte läst boken" (he said that he had not read the book). These appear to be main clauses with respect to property (ii) but embedded clauses with respect to properties (iv) and (v). Classification with regard to property (i) depends on whether the subjunction should be ignored in this test and classification with respect to property (iii) poses yet another problem; if the auxiliary is omitted, we have no finite verb to determine the presence or absence of property (ii). To complicate the matter a little further, the subordinate conjunction can in fact be dropped. If we move the object of the embedded clause to the position immediately following the subjunction; "Han sa att boken hade han inte läst" (he said that the book had he not read), the clause counts as a main clause by (i) (if we choose to ignore the subjunction for this purpose), (ii), (iii), and (iv) but the clause still has to be introduced by a subordinate conjunction (v), which tells us it is an embedded clause. Some of the problems, as Platzack suggests, are immediately solved when we take the hierarchical structure into consideration. Others, however, remain.
All of the above listed properties -- in particular those that pertain to the verb second word order -- have gained attention within research on the Scandinavian lanaguages recently, increasing the empirical base. The goal of this workshop is to bring together scholars working on these and related areas within the Scandinavian languages from syntactic, semantic, and pragmatic perspectives and from synchronic as well as diachronic veiw points as an effort to deepen our understanding of the principles that govern clausal embedding more generally.