In recent years, the semantics of spatial relational terms has become a topic of intense research interest in linguistics and cognitive science, partly because of an interesting paradox. At first blush, the meanings of spatial relational terms seem simple, clear, and obvious. Closer examination, however, reveals that spatial relational terms are highly variable across languages, exceedingly difficult for second language learners, and relatively slow to be acquired by children given their frequency. In this talk, I will demonstrate that the meanings of spatial relational terms are in fact quite complex, as might be expected from the cross-linguistic variability and from the difficulty they present to learners. After discussing experimental work delving into the factors involved in the use of "in" and "on", I will propose a means by which those factors influencing the use of spatial terms can be combined in a representation of spatial meaning.