《宋史全文》編纂歷程新探——以光宗後各朝文本特色為端緒

New Exploration into the Compilation Process of Song Shi Quanwen: Textual Features of the Entries Beginning from Emperor Guangzong

《宋史全文續資治通鑑》（以下簡稱《宋史全文》）是現代學者據以研究南宋光宗以後歷史的重要文獻，單從書名來看，此書是宋代編纂「通鑑」類編年史書潮流下的成品；然而《宋史全文》對各時期史事的書寫方式與述事詳略明顯有別，恐非一人一時的產物。過往研究在討論《宋史全文》的史料價值或編纂特色時，往往將該書視為一個整體，也並未論及附錄之《增入名儒講義續資治通鑑宋季朝事實》（以下簡稱《宋季朝事實》）的性質。有鑑於此，實有必要針對《宋史全文》與《宋季朝事實》的文本特色進行分析比較，以便推斷其編纂歷程，從而獲致更深入的理解。
本文以較具研究參考價值的光宗後各朝為討論範圍，並以繫日比例、條目長度等指標進行量化分析，亦考察用詞、構句等文字風格。本文指出，《宋史全文》光宗後各朝的文本特色各有異同，光宗朝的紀事時間相對籠統，但各條篇幅較長，較有紀事本末之況味；寧宗、理宗朝的紀事較為細緻，繫日亦較為精確；《宋季朝事實》紀事則短而零碎，亦缺乏明確時間，可能意在呈現亡國前後的事實，而非精確編排時日。文本特色的差異，顯示《宋史全文》應是雜鈔不同書籍而成，《宋季朝事實》也不能遽然視為《宋史全文》「有錄無書」的度宗以後紀事。此外，寧、理兩朝雖然文本特色一致，但兩者接合之處特殊的體例與書籍形式，顯示它們並非在同一時期成書；這說明《宋史全文》經歷多次編纂，才成為今日的形貌，此歷程也具體而微地呼應了宋代史籍的修撰趨勢。最後，對讀《宋史全文》與《宋史》本紀部分，更能體現《宋史全文》紀事的性質。透過上述分析，將有助於理解此書形成的歷史脈絡及其史料價值。

Song shi quanwen xu zizhi tongjian宋史全文續資治通鑑(hereafter Song shi quanwen) is an important work for studying the history of the Southern Song dynasty from the reign of Emperor Guangzong (r. 1189–1194) onwards. With the inclusion of“tongjian”in the title, this book is also a product of the trend of compiling chronological histories in the Song. However, Song shi quanwen exhibits significant differences in its writing style and level of detail regarding historical events from different periods, suggesting that it may not be the product of a single author or a single period. Previous discussions on its historical value and compilation characteristics often treat the book as a whole, not considering the nature of the appended Song ji chao shishi宋季朝事實. It is thus necessary to comparatively analyze the textual characteristics of Song shi quanwen and Song ji chao shishi in order to gain insights into their compilation processes, and obtain a deeper understanding of the two works.
Focusing on the texts beginning from the Guangzong reign in Song shi quanwen, including Song ji chao shishi, which are regarded as having more historical value, the present article conducts a quantitative analysis with indicators such as the frequency of daily records and the length of entries, as well as examining certain textual characteristics such as term usage and sentence structure. The discussion reveals that the texts, which are on different periods, have distinct textual features: to begin, entries on Guangzong are more general yet longer in length, offering a comprehensive account of events; those on the reigns of emperors Ningzong (r. 1194–1224) and Lizong (r. 1224–1264) of the Song provide more detailed records with precise dates, but despite their textual features being largely similar, the formats of the two are distinct, indicating that they were not compiled simultaneously; and, the texts in Song ji chao shishi are short and fragmented, lacking clear time references, features which suggest that it should not be regarded as“recorded but not written”chronicles of the reign of Emperor Duzong (r. 1264–1274) of the Song onwards. Considering these elements, Song shi quanwen was likely compiled from miscellaneous sources and underwent multiple compilations before reaching its current form, a process which echoes trends in the compilation of historical materials in the Song dynasty. Finally, a comparison of Song shi quanwen with the annals in Song shi宋史better reflects the nature of the narrative and contents of the former. This analysis thus works to assist researchers in being able to more accurately grasp the value of its historical data and the historical contexts in which it was compiled.