Overlaps in AVIP/IPAR, the Italian Treebank of Spontaneous Speech

This paper will present work carried out on the 60,000 words Italian Spontaneous Speech Corpus which collapses together two corpora, the first one called AVIP, under national project API - the Italian version of MapTask, the other one called IPAR. We will present in particular the parser, to produce syntactic structures of overlapped temporally aligned turns. The paper will argue in favour of a joint and thus temporally aligned representation of overlapping material to capture all linguistic information made available by the local context. We argue that this is different from simply aligning on a time scale linguistic annotations on separate layers of representation. In fact, we are interested in collapsing overlapping linguistic material within a single representation in order to capture pragmatic inferences. This will result in a syntactically branching node we call OVL which contains both the overlapper's and the overlappee's material (linguistic or non-linguistic). The paper will also comment on data from our Written Text Treebank called VIT and compare them with the Oral Treebank.