Architectural plans are design diagrams that describe building layout where space is planned according to design requirements. Style in architecture is generally characterized as common features appearing in a particular class of building design. This research seeks to address how to recognize architectural design style from a 2D plan diagram. We explore this question in a computational encoder-analyzer (E-A) model for 2D plans, where a characterization of 2D style is based on qualitative spatial representation and information theoretic measures. In a preliminary study of a prominent architect's plans, we demonstrate the effectiveness of our approach. We conclude by discussing practical applications of automated plan recognition and classification in design support tools.