本文是以殷周金文中所見的親屬稱謂「姑」進行分析，與傳世文獻記載的「姑」的定義互相比對以釐清在每一銘文中「姑」的意義。帶有親稱「姑」的銘文大體上可以區分為早晚兩期，早期（商晚期至西周早期）銘文中之人物都是以十日命名者（商人貴族），其「姑」的意義是丈夫的母親(HM)。此組金文中所反應的基本關係是「婦姑對稱」。晚期（西周晚期至春秋）銘文中的人物則多是以伯、仲、叔、季命名者，此組銘文中之「姑」的意義變得多元，既可以是丈夫的母親，也可為女姓稱謂人的姑媽，而且同時也出現了「姪」與之相對，也可以是男性稱謂人的姑媽。 根據兩組的時間順序，以及親稱衍生的邏輯，我們推測此種變化，可能包含以下幾個步驟：首先「姑」這個親稱是十日命名者的原生親稱。在周人統治（屬於伯仲叔季的命名者）之後，由於史官、作冊、書吏仍由文化水準較高的商人擔任，因此將「姑」這個稱謂借用到周人的書寫系統中。然而傳世文獻顯示姬姓的周人與姜姓集團或其他古姓間，原有「母方交表婚」或「雙方交表婚」的關係，因此有「夫之母」與「父之姊妹」合一或同屬一類的現象，故形成了「姑」既有婆婆的意義，也有父之姊妹的新的意義。同時金文中也出現了前所未見的「姪」這個親稱，此一稱謂可能是周人既有口語的稱謂被行諸文字，其原始的關係是姑媽與姪女間的關係，亦即「姑姪對稱」。進一步發展而有因為兄弟隨姊妹稱姑，產生的「姑侄對稱」，以及丈夫隨妻子稱姪，而發展出「叔姪（侄）對稱」等次生之關係。 從「姑」的用法不同，我們已經可以感覺到，商人的親屬體系與周人有別。進一步分析我們發現，以十日命名群體的女性稱謂人的親屬稱謂體系，與男性稱謂人的親稱體系，在婚前是相同的，但是婚後有兩點差異，一是婦人稱丈夫為「辟」，一是稱婆婆為「姑」。其餘各類親屬的稱謂則婚前與婚後基本上不變。根據女性婚前婚後自稱以及稱謂體系的變化推論，我們得到一個相當重要的結論：以十日命名群體是多世系群（三個或更多）的系統。此一現象顯示張光直與持井康孝、松丸道雄提出的「十世系說」值得我們進一步探討。

　　　　　This paper begins with an analysis of every occurrence of the term ku 姑 in the Shang and the Chou bronze inscriptions. They are compared with definitions of ku in the classical texts in order to clarify the meaning of each occurrence. The bronze inscriptions with the term ku can be classified into two groups. In the first, dated from the Shang to early Chou, all the dedicated persons in the inscriptions are represented by one of the ten kan 十干 signs (Heavenly Stems). This means that they may have been the Shang royal class. The meaning of ku in this group is the husband’s mother. The basic relationship is a reciprocal one between a married woman fu 婦 and her hubsand’s mother ku. In the second group, dated from the late Western Chou to the Spring and Autumn periods, the persons in the inscriptions often carry a name with po 伯, chung 仲, shu 叔 or chi 季 (such as Bo-Mao-Fu 伯懋父), making the meanings of ku in this group much more complicated. It can be the hubsand’s mother in some cases and father’s sisters in others female egos. At the same time, a new term chih 姪, meaning the brother’s daughter, occurred as a new reciprocal kinship term for ku. In this period, ku was also used as the father’s sister for male egos. Based on the chronological order of the two groups and the logical evolution of the meanings for kinship terms, we infer that there were several possible steps for the development of meanings of the kinship term ku. We argue that ku was a generic kinship term for the group using ten kan signs in their “temple names (miao-hao 廟號).” In the first hundred years of the Chou Dynasty, most of the posts of historians, document makers, and clerks remained occupied by the Shang nobility perhaps because their literacy rate was much higher than that of the previously barbaric Chou people. They introduced the term ku to the writings of the Chou Dynasty. However, matrilateral cross-cousin marriage or bilateral cross-cousin marriage were popular long before the establishment of the Chou Dynasty among both the Chi 姬 clan of the Chou royal house and the Chiang 姜 clan, a powerful alliance of the Chou, as well as other ancient clans. For them, the husband’s mother is either the same person or of the same category as the father’s sister. Therefore ku took a new meaning of the “father’s sister” in the second period. In this period, the term chih also occurred in the bronze inscription for the first time. It is possible that chih was a generic kinship term of the Chou people and was introduced to the writing system to distinguish the brother’s daughter for a female ego. The meaning of ku became broader during the second period, as brothers designated their father’s sisters as ku by imitating their sisters. Thus a new reciprocal kinship dual of ku and chih 侄 (male) occurred. Subsequently, husbands designated their wives’ brothers’ daughters or sons as chih by mimicking their wives. Then, a secondary reciprocal kinship dual of shu 叔 and chih 姪 or chih 侄 developed. In this period, all the relationships associated with the kinship term ku recorded in the Er-ya Shi-ch’in 《爾雅•釋親》 occurred in bronze inscriptions. From the difference in the meanings of ku for the Shang and the Chou, we begin to sense the differences between the two groups of population. Our further analysis reveals that there was no difference in kinship terminology for male or female egos before marriage for the Shang people. However, a married Shang woman designated her husband as pi 辟 and husband’s mother as ku. These were the sole differences in kinship system for a female ego, and the rest of the kinship terms remained the same. A kinship system with such characteristics practised neither matrilateral-cross-cousin marriage nor sister-exchange marriage. Instead, there should be three or more parallel lineages associated with such a marriage alliance. This conclusion endorses the argument by K. C. Chang, Mochii Yasutaka and Matsumaru Michio that there were ten lineages in the Shang royal “clan,” each with a designation of a kan sign. This is an important theory deserving further attention.