史學研究的推進，有賴於對新史料的發掘與對舊史料之價值的再認識，唐令復 原這一持續百年的研究課題亦莫能外。《天聖令》殘卷的問世，揭開了唐令復原的 新一頁，並進一步推動學者對於既有史料的深入檢討。迄今為止，相關研究皆圍繞 某一文獻，檢討其之於唐令復原的史料意義。由此可進一步思考以下問題：兩種以 上史料是否會因其所錄制度之時間節點不同等而互為矛盾？此時該如何取捨？檢討 所得的復原原則僅僅具備「高度蓋然性」，能否保證「萬無一失」？因此，本文的 檢證將不再圍繞單一文獻，而訴諸令式分辨這一線索展開。 唐代令、式在規範屬性上有相近之處，學界歷來以令為原則、式為細則予以分 辨，也有論者以規範對象作為劃分依據，提出令的規定涉及官吏與庶民、而式僅僅 規範官吏的觀點。又由於繼受唐代制度的日本令、式相對保存完好，論者又多以 「唐令─日令」、「唐式─日式」這種單一、封閉的法律移植路徑為據，反推唐代 令、式之別。其實，這些原則、標準僅是令、式關係的一個側面，以此作為復原唐 令的決定性「理據」，具有相當風險。 唐代律、令、格、式等法源互為配合，內中不乏因規範內容相類而導致文字相 近、相同的現象，而文字相近、相同並不意味著條文完全一致。因此，圍繞同一條 文，數種史料或標為「令」或記為「式」，其中不但要考慮記載訛誤的可能性、因 史料無斷句而出現的誤判，以及「令」、「式」並非實指法源而泛稱法律等因素， 還應結合令、式發展的階段不盡同步這一歷史過程，引入令、式之間條文轉化的動 態視角。 既然史家或抄手可能出現疏漏，導致史料記載的訛誤，立法者也概莫能外。若 是唐代法典並非「邏輯」嚴密、體系井然的法律文本，那麼令式分辨、唐令復原、 唐式輯佚將面臨更加嚴峻的考驗。當然，這一「邏輯」究竟是唐人之法理還是近人 之思維，則有待於對唐代立法技術的進一步探究。

Advancement in the studies of history relies on both discovery of new historical materials and rethinking of the old ones. Despite of being carried out for over a century already, the restoration of ling (statutes) in the Tang dynasty is not an exception. With the discovery of the remains of Tiansheng ling, the restoration of ling in the Tang dynasty enters a new phase and scholars are incited to further analyze existing materials in more depth. However, so far, these researches have focused on studying one specific historical material at a time. Several questions may therefore arise. Could it be possible for two historical materials from different period of time be contradictory? If they were, which then would we deem as the legitimate source? Also, can we be certain that our restoration is absolutely correct? This paper attempts to pay attention to the discrimination between ling and shi (ordinances), while rethinking about these questions in above. Ling and shi in the Tang dynasty were much similar in character in that both were without provisions of punishment. A common perception among the academics is that ling was the principle and shi was the detailed rule. Some scholars also think that ling bound both the officials and the populace, but shi bound the officials only. Moreover, some other scholars discriminate ling from shi with reference to ancient legal sources from Japan, as the Japanese at that time adopted the legal institutions from the Tang. Actually all these assumptions only reflect one aspect of the relationship between ling and shi, so it is quite dangerous to consider any of them as the sole determining proof. Lü (code), ling, ge (regulations) and shi existed simultaneously during the Tang dynasty. Although sometimes provisions from each of the four legal sources on the same matter showed resemblance, there were still notable differences among them. Even if a provision was recorded as ling in some materials and as shi in others, it shouldn’t be concluded casually that ling and shi had this provision in common, as the records could possibly be mistaken or we might make wrong pauses in reading these ancient and unpunctuated materials. We should note as well that one provision in a ling could possibly change into a provision in a shi because developments of these two legal sources sometimes didn’t synchronize. Since it was also possible for ancient historians or the copyists to make human mistakes, the legislators could very likely make errors accordingly. In short, the research on the discrimination between different legal sources in the Tang dynasty and the reconstruction of these legal sources therefore require extra attention and care.