This paper attempts to show that the graph □ expressed the word jiǔ 灸 'moxibustion, moxa'. In contrast to many a graph with variants, ranging in number from one to more than twenty, the graph in question has no apparent variant. There is, however, a graph □ seen on the same piece as the graph □. These two graphs occur in a series of ten related inscriptions that form a contextual hierarchy (CH). Since they share the same grapheme □-suggested in this paper as the protoform(初文) of jiǔ 久 'long time; moxibustion'-it has served as a point of departure for further examination. The paper analyzes the graphemic constituents of □, namely, □, □, □, and □, while those of □ are just two: □ and □. They were configured in ways that must have been meaningful to the scribe who originally "designed" □ and □. The paper argues that □ was an "imaginative variant of □", a major claim of the paper. It rejects the prevalent transcription of □ into "㐱" read zhěn/*tənʔ and glossed "tufty hair(鬒)". Because this makes no sense in the context in which it is used, a recent study has proposed that 㐱 stood for the word lì/*rəts 沴 'water disadvantageous' (SW says: 水不利也 ; 从水㐱聲) . The problem is that it is wrong to transcribe □ as 㐱 , further interpreting 㐱 as 沴 . The paper explains why. Based on various pieces of evidence and analyses presented in the paper, we submit that the contemporary Shāng scribes had no trouble identifying both □ and □ wrote the word jiǔ/*k^wəʔ/h 灸.
本文試圖證明甲骨文字□表示「灸」這個詞。與許多具有異體(數量從1到20多不等)的字相比,我們的初步印象是此字併沒有任何明顯的異體字。然而,有一個寫作“□”的字與□字曾出現在同一塊甲骨片上(《屯南》2219),構成所謂「同版關係」。而且這兩個字也曾出現在十条相關命辭中,形成「上下文層次結構」(contextual hierarchy)。由於□和□這兩個字具有相同的字位(或字素)即—我們認為該字位本身是本義為「長時間;艾灸」之「久」的初文—,它成為進一步研究的出發點。本文分析了□的字位(或字形成分),即:□、□、□、□,以及□的字位,即:□和□。對於原先設計□和□的造字者來說,這些配置方式一定是有意義的。本文試圖對此進行解釋,提出□是□的一個「富有想像力之變體」—這是本文的一重點,我們將在本文中進行解釋—,由此否定了普遍的將□隸定為「㐱」的說法。因為㐱字表示的「鬒」義於上下文不通,最近也有研究提出「㐱」(常讀zhěn/*tənʔ)表示lì/*rəts「沴」這個詞(說文云:「水不利也;从水㐱聲」),但是,其根本問題在於契文□本不應隸定為㐱,因此,將㐱進一步釋為沴就是完全沒有根據的。在本文將說明其理由。根據本文提出的各種證據和分析,我們認為商代當時的刻寫者清楚地知道□和□這兩詞同為jiǔ/*k^wəʔ/h「灸」這個詞的書寫形式。