Objective The legacy of gardens holds a distinct place within Chinese culture, representing a vital facet of Chinese artistic heritage. In contrast to other forms of art such as painting and sculpture, the exploration of Chinese gardens faces a significant challenge due to the limited survival of early gardens predating Qing Dynasty. Consequently, the restoration of these gardens constitutes the foundational and central domain of Chinese garden historiography. This research introduces the concept of “perished garden” comprising two distinct types: Type Ⅰ perished gardens, where all tangible elements have perished, leaving behind only remnants, related drawings and texts, and other artifacts. Examples include the Shining Mountain Residence in Eastern Jin Dynasty (东晋始宁山居), the Wangchuan Villa in Tang Dynasty (唐代辋川别业), the Dule Garden in Song Dynasty (宋代独乐园), and the Zhi Garden in Ming Dynasty (明代止园). Type Ⅱ perished gardens encompass those in which some or all of the physical elements still exist but have undergone extensive reconstruction, resulting in partial obliteration of the original landscape during specific periods. Notably, some elements may persist, albeit with alterations in style, exemplified by the Humble Administrator’s Garden in Suzhou (苏州拙政园), the Jichang Garden in Wuxi (无锡寄畅园), and the Xiequ Garden in Beijing (北京谐趣园).

Methods This research outlines a comprehensive framework for restoration research focusing on perished gardens, and adopts a multifaceted approach rooted in various forms of evidence. Evidentiary sources can be categorized into physical evidence, documentary evidence, and verbal evidence, based on their forms. Additionally, evidence can be classified as direct or indirect, depending on its relevance to specific gardens. This taxonomy encompasses seven distinct categories of evidence: on-site physical evidence, site archaeology, poetic documents, pictorial representations, oral records, related physical artifacts, and gardening theories. While the first five directly pertain to the gardens under research and can independently contribute to their restoration, the latter two constitute indirect evidence, requiring integration with other sources to establish their evidentiary value. This sevenfold evidentiary approach possesses unique characteristics and methodologies, necessitating careful selection and application to construct a coherent chain of evidence supporting the restoration of perished gardens.

Results In the utilization of evidence, three crucial considerations emerge: Firstly, evaluate the availability and probative force of different forms of evidence to establish their respective importance in the restoration research; secondly, recognize potential corroborations or contradictions among evidence sources, identify key linkages between them, and leverage such linkages to reconstruct the historical appearance of gardens based on the formation of an evidence chain or network; thirdly, acknowledge the dynamic nature of evidence. When confronted with uncertainties or contradictions in the restoration process, researchers must broaden the scope and depth of available evidence to resolve these challenges. This shift represents a transition from reliance on a single weight of evidence to development of a multifaceted evidence base. New evidence may potentially revise or enhance existing research findings, offering the prospect of more robust support for prior results.

Conclusion The adoption of an approach based on multiple evidence to the restoration of perished gardens holds promise as a valuable reference for the research on historical gardens, which may contribute to a more comprehensive understanding of the evolution of classical Chinese gardens throughout history.