晚清以來治古文字學者有意識的歸納研治古文字的方法，如羅振玉、王國維、唐蘭、楊樹達、李學勤、裘錫圭諸先生，都提供了許多考釋文字的理性經驗。本文強調先由句法的角度分析甲骨卜辭的重要性，提出文例研究的方法，主要是掌握成組辭例為基礎的詞組研究法。我們建議，了解文字必須以分析文句為基礎，先弄懂整個句子內容的大方向，透過成串已知符號的詞位、詞序和上下文義，從而鎖定待考的未知字形在詞性和用義上的可能性，最後才由形音義的角度判斷文字。 本文先就趙誠先生《甲骨文字學綱要》一書中若干由於缺乏文例的通盤考量而產生據形誤釋的例子，強調文例在考釋甲骨文字的重要性。本文復申論文例的功能，有： （一）由文例對字用的認識，幫助對字形的理解。如甲文丰之應讀為圍，井可以讀為毋。（二）由文例對詞組的認識，幫助對上下文的理解。如“人牛＂應釋為“牝牛＂，“司工＂應理解為“祠示＂。（三）由文例互較破讀屬省文的字形。如天為茲省，夭為寅省。（四）由文例可幫助通讀卜辭的行款。如〈集22147〉舊釋的“癸巳卜妣妥＂“用＂“庚……＂三段應合讀為“癸巳卜，用妥妣庚？＂〈集22186〉舊釋的“己丑卜啟丁今。＂應讀為“己丑卜：今日啟？＂。（五）由文例分析可解決卜辭的標點。如〈丙30〉的“沚乃其來＂應讀為“沚、乃其來？＂〈集6057〉的“長友角告曰＂應理解為“長、友、角告曰＂。為“疾止，唯有入？＂之省。〈集6559〉的“王于來春伐刁”應為“王于來春伐刁，災？＂之省。

Since late Chin Dynasty, scholars of ancient etymology have comsciously induced the methodology of etmological studies. They have contributed quite much rational empircism in deciphering ancient characters. This paper emphasizes on the essentiality of syntax in analyzing oracle-bone inscriptions. We suggest that sentence analyzing should be done before trying to understand the individual characters. Namely, by way of already known word positions, word orders, and contexts, we first lock on the possibility of the meaning and grammatical form of unknown characters, and then, with the help of form, phoneme and meaning, decipher the words.