The Chinese spatial terms shàng (’上’) and lǐ (’裡’) are two of the most frequently used localizers, depicting the spatial relationship of support and containment between the figure and the ground. Throughout the process of lexical development, shàng and lǐ have evolved from precise localization to generic localization, and this change has resulted in an expanded range of meanings for both. This evolution allows both localizers to be interchangeably applied in describing specific spatial scenes. Adopting a corpus-based approach, we retrieved and analyzed spatial scenes that can be encoded by both shàng and lǐ. These spatial scenes are further classified into six types based on geometric features of the ground and the functional relationship between the figure and the ground. The study aims to analyze the use types and meanings of shàng and lǐ in these six types of spatial scenes. The research reveals that the generic position interpretations of the localizers and the speaker’s cognitive construal motivate the interchangeability of shàng and lǐ. Additionally, the precise position interpretations of the localizers and the idiomatic structure of locative phrases are identified as crucial factors influencing the selection tendencies of the two localizers.