Cikoski, p. xix.: “I include few citations from 説文. Its glosses are unhelpful in reading CC texts; is dissection of graphs is arbitrary.”
Bottéro & Harbsmeier, p. 249.: “The Shuowen is not a dictionary of basic meanings of words. Its is a dictionary of graphic etymology, and etymology needs to be carefully distinguished from semantic analysis. It provides only meanings that are relevant to the explanation of the graphs used to write words. Similarly, the Shuowen only refers to pronunciations of graphs insofar these are relevant to the explanation of phonetic constituents in those graphs. When the author, Xǔ Shèn 許慎 ... adds alternative pronunciations, all of these are construed as possibly graphologically relevant, and when he adds alternative meanings with or without sources for them, these are again construed as being in principle similarly graphologically relevant. Such list never attempt to provide any comprehensive overview of the different meanings of words written by the graph under discussion. And moreover, the meanings indicated in his glosses are very often not the basic meanings of words in question, but those meanings that Xu Shen considers serve best to explain the structure of the graph. ”
Sampson & Zhiqun “The reality of compound ideographs” in Journal of Chinese Linguistics, vol. 41, no. 2, June 2013. “...Xu Shen certainly got many things wrong (which is in part explained by his lack of access to data that have come to light in the past hundred years)...” (p. 256).
Qiú, p. 95: “The Shuōwén ... was not completed until the mid Eastern Han. The shapes of some small seal forms as written at that time were already erroneous. In addition, the scholars of the script of that time, including Xǔ Shèn, sometimes could not avoid having misconceptions about the structural composition of the small seal forms. Their misconceptions sometimes led them to distort the shapes of the seal forms.”
Qiú, p. 154: “In certain cases Xǔ Shèn construed graphs which were composed of abstract symbols as being xiàngxíng graphs.”
Qiú, p. 161: “...the Shuōwén takes great delight in giving forced explanations of the phenomenon of semantic extensions...”.
Qiú, 230 ,,,
By the Qin dynasty, the character has lost its original shape. Based on the distorted form, Shuowen mistakenly interprets the character as a pictogram (象形) of a plant growing outwards.
Qiú, p. 238-241 has the section “Erroneous Analysis of Graphic Form” which explains perhaps two dozen errors of this kind in the Shuōwén.
Qiú, p. 287: “Loangraphs are ready-made graphs which themselves have meanings; therefore it is easy for people to mistake the polysemy caused by the fact that a character can have several meanings as being due to semantic extension. The founder of the study of writing in China, Xǔ Shèn, confuses these two notions; he often misinterprets what we call the loangraph phenomenon as semantic extension.”