Eugene Chung. 2018. Lexico-Syntactic Structures of Spatial Expressions. Language Information, Volume 26. 79-102. This work aims to model a lexico-syntactic structure for various spatial configurations. In English, spatial expressions are similar to tr...
Eugene Chung. 2018. Lexico-Syntactic Structures of Spatial Expressions. Language Information, Volume 26. 79-102. This work aims to model a lexico-syntactic structure for various spatial configurations. In English, spatial expressions are similar to transitive verbs in the pattern where prepositions take arguments to describe spatial relationships. The meaning expressed by a spatial preposition indicates how its arguments physically relate to each other in space. The spatial preposition has two arguments, a Figure and a Ground. This study adapts Pustejovsky's (1991, 1995) argument structure and co-composition from the Generative Lexicon Theory in order to represent multiple configurations. Various features are employed to describe spatial relations and spatial entities. Arrangement features and physical relationship features are for spatial relations. Region features, dimensionality, and orientation features are utilized to provide fine-grained specification of spatial entities. The values for arrangements are CONTACT, ADJACENCY, OVERLAP, INCLUSION, SURROUNDING. Region feature takes one of SURFACE, BOUNDARY, TOP, BOTTOM, INTERIOR, EXTERIOR. For the dimensionality, we have 1DIM, 2DIM, 3DIM values. HORIZONTAL and VERTICAL are the values for the orientation feature. A locus structure is introduced to encode various spatial features that are related with words about spatial entities and spatial relations. A spatial preposition and its arguments construct spatial expressions. They compose phrasal level meanings through co-specification with the locus values in the loci structure. The proposed framework can be utilized to represent the formal rules of encoding spatial expressions.