Quantifying Spatial Patterns of Rockery–Plant Arrangements in Suzhou Classical Gardens: A Coding Framework and Perceptual Evaluation
Year: 2026
Type: preprint
Abstract: Traditional cultural landscapes preserve not only physical forms but also implicit design knowledge developed through long-term human practices. However, such knowledge is often transmitted through textual interpretation and expert experience, making it difficult to describe, compare, and analyze systematically. This study examines how the spatial knowledge embedded in Suzhou classical gardens can be transformed into measurable evidence through the analysis of rockery–plant configurations. A structured coding framework was developed to represent three interconnected dimensions of garden composition: rockery features, plant features, and their relative spatial relationships. The framework was applied to 23 micro-scale courtyard landscapes from the Lingering Garden and Lion Grove Garden in Suzhou. Combining field documentation, eye-tracking-based identification of visual focal elements, and perceptual evaluation, this study investigated recurring spatial configurations and their relationship with contemporary visual interpretation. The analysis revealed recurrent patterns in rockery–plant arrangements, including the layered relationship of foreground plants, middle-layer rockeries, and background trees. Taihu stone represented the dominant rockery material, while low shrubs and taller trees frequently contributed to different vertical layers of courtyard composition. Perceptual results indicated that these recurring configurations showed certain correspondence with contemporary visual responses, although no single arrangement demonstrated universal preference. The findings demonstrate how traditional landscape knowledge can be translated from tacit design practices into structured spatial evidence. This approach provides a possible pathway for connecting heritage interpretation, human perception studies, and contemporary landscape research.
Source: Research Square
Authors: yakun chang, Danqing Mo, Danyang Zhuang, Yingnan LI
Institution: Jiangsu University
Language: English
Subfield: Architecture
Field: Engineering
Domain: Physical Sciences
Open Access status: green
Awards: G4372786304, G8137603610
OpenAlex ID: W7226036792
