There have been very few dedicated studies on the subject of orthographic variability and ancient character forms in early Chinese writing, in English or even in Chinese scholarship. We are glad to see Professor Imre Galambos’ significant volume coming out with new materials, approaches, and findings.

Traditional views prescribe the evolution of early Chinese writing as along a single and simple line from Shang-Zhou oracle-bone and bronze inscriptions to Qin-Han small seal and clerical scripts. The effect of the Qin unification and standardisation of the writing system has been taken at face value. When explaining cases of variations, scholars have generally assumed that every character had a correct or standard form while all other forms were derivatives, and have attributed most of the differences to either temporal or spatial factors.

Galambos tries to correct these views by using new archaeological materials and a set of clearly defined criteria and approaches. He analyses the descriptions of Xu Shen and Ban Gu on the evolution of writing and the Qin unification and concludes that Xu and his later interpreters seriously overstated the effect of the writing reform. The differences between the character forms on the Qin edict plates show that the Qin small seal script in common use was not consistent. The differences between the characters inscribed on the Yuan'an stele dated in 117 ad and the Shuowen character forms also show that the Han small seal script was not uniform. Archaeological data reveals that the standardisation of writing did not happen overnight in 221 bc but instead lasted several centuries, beginning before the establishment of the Qin dynasty and lasting far into the Han or even longer. While some scholars have questioned the notion of the Qin unification and standardisation, Galambos discusses this issue with unprecedented depth and convincing arguments.

To identify a character form, the common approach is that characters are graphically defined by their components, an approach that is based on the concepts of “correct form” and “variant/erroneous form”. Galambos indicates that, because the structure of Warring States character forms is not constant, graphical attributes are not sufficient for defining the identity of a character form. Instead, it is more useful to define them based on the word they represented in context, which serves as an unchanging entity behind all variant character forms. Two characters, which from a modern point of view appear to be phonetic loans for each other, often represent cognate words, even if they do not share a graphically identical component, as seen in the case of shou 獸 (to guard, hold) and shou 守 (to guard, hold) in the Guodian Laozi. Moreover, the structural variability of character forms in excavated texts implies that in Warring States times the concept of a standard form did not exist. A standard script is a result of a prescriptive measure enforced by an authority, but there are no records of such an authority during the Warring States period. Therefore, it is more reasonable to look at character forms without the assumption of a standard, in order to avoid the common mistake of looking at variants in the light of modern character forms. Galambos then further defines two more parameters. The first is ‘local script’ which defines the boundaries of the community that used the same script. Character forms of Warring States manuscripts and inscriptions differed not only between the six states but also within the same state, the same written dialect, and even the same text. In an entirely local script, such as the ones defined by the Guodian or Baoshan corpora, variability of character structure was a part of the script. Analysing an entire local script offers a chance to look at character forms in their totality. The second is the parameter for determining which character forms are variant forms. Two forms are variants of each other when they stand for the same word and yet are structurally dissimilar, differing in at least one component. Some scholars have noted the absence of the concept of a standard form and the differences within the writing of the same region and even a single text in early writing, but Galambos extensively explores these issues with unprecedented nuance and, more importantly, clearly sets up new, useful parameters and terminology for the examination of Warring States character forms. His emphasis on the analysis of the word the character forms represented in context is especially insightful. To use components as the parameter and to exclude strokes for defining variant forms, however, may lead to further discussions, as He Linyi's 何 琳 儀 study demonstrates that stroke differences also often produced obviously variant forms during the Warring States period.Footnote ¹

Applying these new parameters and approaches, Galambos chooses the Houma covenant texts for examining the patterns of structural variability. Unearthed in the 1960s from Shanxi and dated to the end of the Spring and Autumn period around the beginning of the fifth century bc, this group of texts consists of hundreds of nearly identical copies of the same oath. From the corpora, Galambos selects three words, zhi 志 or intention (232 character forms), fu 復 or to restore (207 character forms), and fu 腹 or abdomen (225 character forms), and analyses every written form of these words to detect orthographic patterns. The Jin scribes wrote a word with specific form more often than others, but they also consistently used other forms. This shows that variability was an integral part of Warring States writing. Among the variant forms, there was generally a dominant character form, showing that the scribes of the Warring States period adhered to certain conventions regarding structure, but with a larger degree of tolerance. Despite the great variability of semantic components, there usually was a semantic connection between the varying component and the usage of a character form. Most importantly, the components that varied were usually semantic in nature, and the phonetic remained the same in nearly every case. The flexibility of the graphic form reflected a higher orthographic tolerance on both the scribe and the reader.

From these observations, Galambos further draws some implications and speculations. The methods that use isolated examples of orthographic forms by themselves for dating texts and objects or determining the geographic origin of manuscripts and artifacts should be avoided. The Qin-Han standardisation represented not a return to an original state of orthographic uniformity but to the initial standardisation of earlier writing practices. Chinese characters evolved not along a linear path that stretched from the oracle-bone inscriptions to the modern script but followed a complex process involving distinct cultures and languages.

These conclusions and speculations are incisive and inspiring. In order to convince the reader with so many arguments, however, a more comprehensive framework and sophisticated study, instead of the analysis of a few words, may be further expected. A couple of mistakes in the translation and interpretation of the “Postface” of Guo Zhongshu's 郭 忠 恕 Hanjian 漢 簡 could be avoided through a more careful reading.Footnote ²

Galambos’ work offers useful approaches and terminology to the study of Warring States manuscripts and variant character forms; it draws many insightful conclusions and observations, and sheds fresh light on the development of the Chinese script. It is no doubt a significant contribution to the field of early Chinese writing.