Colloquium: A thing or two that I learned studying dislocations

November 20, 3:30-5pm, Cobb 201
Luis López Carretero, University of Illinois at Chicago

The following assumptions are generally (almost universally) held to be true of Romance dislocations: They are topics and therefore, they are specific/definite. They surface in Spec,Top, a category sitting between Finite Phrase and Force Phrase. Their position in Spec,Top is motivated by a "Topic Criterion", one of a family of criteria that hold of functional categories in the C-domain. Within the criteria framework wh-movement, focus-movement, topic-movement etc are triggered or licensed each by a different feature – consequently, the A'-dependency construct dissolves, as shown by lack of intervention effects. Topic-movement is also different from wh-movement in that it is only sensitive to strong islands. Dislocate and clitic are initially merged as one constituent forming a "Big DP". In the first part of the talk I will show that none of these assumptions stands to scrutiny. The picture that will emerge will show that dislocations are not that special, movement types are really not fine-grained and A'-dependencies are worth keeping. As a consequence, the proximate and ultimate causes of movement are more deeply dissociated than the criteria approach would make us think.

In the second part of the talk I will address an issue of the grammar of dislocations that has, so far, been neglected: their connectedness with discourse. I will show that the anaphoricity of dislocations should not be confused with English-style givenness and that the connection with the antecedent obeys two conditions: a locality condition and a discourse-subordination condition. I will present and describe these properties raising (but not answering) some questions that might yield fruitful research agendas.