王念孫（1744-1832)為清代考據學之重要學者，其《廣雅疏證》一書，素來為學者所重。本文從資料取證的角度論《廣雅疏證》的校勘方法，分兩個部分論之。如《疏證》取證資料宏富，注重《廣雅》訓詁之傳承脈絡，常云「《廣雅》多本某書」、或「某書本於《廣雅》」，並由此作校勘之判斷依據。《疏證》有言《方言》、《說文》為《廣雅》所本者，而《玉篇》、《廣韻》、《集韻》則本於《廣雅》，但這非絕對之情形，另存有王氏校勘的主觀思考成份。本文旨在探討王氏自云所側重之幾種主要資料，以見其校勘《廣雅》之方法。其次，探討《疏證》以「諸書無訓」為校勘之立論基礎，並依其校勘形式分作兩點析論，分別為「諸書無訓 者，字之譌誤」、「諸書無訓者，上下條誤合為一」。

Wang Niansun (1744-1832) is a well-known scholar in the Qing Dynasty. This article discusses the method of collation of Guangya Shuzheng from the perspective of data forensics. It is divided into two parts. The first is to discuss the “exclusion of the book”. It is the basis for the collation, and it is divided into two points according to its collation. Secondly, "Shenzhen" emphasizes the inheritance of the "Guang Ya" exegesis, and thus serves as the basis for the judgment of the collation. But this is not the absolute case, but the subjective thinking of Wang's collation. This article aims to explore the material that Wang Niansun has focused on and how to collate his method of Guangya.