East Asian wooden architecture has long established an order organized around the concept of the kan (bay), through which structure and space are articulated. Within this tradition, the preference for odd-numbered bays has been regarded as a self-evid...
East Asian wooden architecture has long established an order organized around the concept of the kan (bay), through which structure and space are articulated. Within this tradition, the preference for odd-numbered bays has been regarded as a self-evident premise. However, rapidly accumulating excavation data from the Gyeongju area reveals that during the transitional period before the odd-bay paradigm was firmly established, diverse column grids and planar experiments coexisted to a far greater extent than previously assumed. This study focuses on the 2-bay × 2-bay plan—previously dismissed as an exceptional anomaly or a static primitive form—to empirically reconstruct its historical trajectory and architectural significance, ranging from its emergence and variation to its eventual decline.
The primary scope of analysis is the Silla Royal Capital, where archaeological data is most densely concentrated, covering the 4th through 9th centuries. The study employs a quantitative methodology based on a database of drawings and stratigraphic information from excavation reports. To enhance empirical validity, this is supplemented by a review of literary sources such as the Samguk Sagi and Tang Liudian, alongside field surveys. Furthermore, by cross-referencing contemporaneous data from China, Japan, and major Korean regions (Goguryeo, Baekje, and Silla), the study avoids reducing findings to isolated regional phenomena and establishes comparative historical coordinates.
For quantitative analysis, a complete survey of major excavation reports published between 2002 and 2023 was conducted, establishing a primary dataset of 246 building sites with legible bay configurations. To overcome limitations in time-series comparison caused by varying dating systems across reports, chronology was standardized into three phases: Period ① (5th–early 7th c.), Period ② (late 7th–8th c.), and Period ③ (9th–early 10th c.). Moreover, based on the hypothesis that minute structural differences within the 2-bay × 2-bay plan reflect deliberate choices in structural technology and spatial management, the plan was further classified into subtypes based on the presence of a toe (extended bay) and the arrangement of inner pillars.
Results indicate that the 2-bay × 2-bay plan was not a specialized type limited to specific hierarchies or functions, but a versatile form found widely from central residential buildings to authoritative architectural complexes. The fact that this plan accounts for approximately 10% (26 sites) of the total sample and forms a significant axis of plans with a depth of two or more bays (alongside the 3-bay × 2-bay plan) suggests it was a design vocabulary with established rules and utility, not a random exception. Simultaneously, its competition and coexistence with 3-bay × 2-bay and 3-bay × 1-bay forms—showing shifting patterns of dominance and scale—demonstrate that the establishment of the odd-bay paradigm was not an instantaneous selection but the result of prolonged competition among multiple architectural solutions. Additionally, distinct choices regarding inner pillars reveal that different spatial principles, such as securing spatial depth or reinforcing frontality, operated within the same square grid. In sum, the 2-bay × 2-bay plan was less a single prototype and more a “variable standard” capable of flexible adjustment according to situation and demand.
Based on these findings, the historical trajectory of the 2-bay × 2-bay plan is reinterpreted in three stages. In the embedded-pillar phase, it functioned as a “prototypical form,” offering a minimal unit solution with structural stability and economic efficiency. Even as advanced technologies like foundation stones and tiled roofs spread around the Unification of the Three Kingdoms, the form did not disappear but persisted as a “modified form” by adjusting its hierarchy and function. Furthermore, as the 3-bay façade norm became consolidated following Unification, the plan transformed into a “reconstructed form” from the 9th century onward, adapting to the new spatial order through methods such as adding a toe or expanding its scale. Thus, the 2-bay × 2-bay plan should be understood not as a fleeting transitional remnant that vanished before the odd-bay era, but as a dynamic form that continuously renewed its role and morphology in response to changes in technology, hierarchy, and norms.
The significance of this study lies in empirically restoring the trajectory of even-numbered bay experiments, represented by the 2-bay × 2-bay plan, thereby illuminating the settlement process of the odd-bay paradigm from its converse side. The narrative of a “norm” is incomplete on its own; its formative process is fully revealed only when read alongside the logic and life cycles of competing alternatives. Simultaneously, while acknowledging the disparity in data density between the linear data of the Silla Royal Capital and the scattered data across East Asia, this study provides a foundational framework for expanding the comparative horizon in future research.