Some thoughts on the reliability of 說文解字

Disclaimer: This article will be very technical and very probably very uninteresting if you are not

familiar with Chinese character etymology. My apologies in advance. 

I got into a debate with someone online under one of my videos recently. The video was about the 辡 character phonetic series. In the beginning of the video I argued that 辡 was a character formed by two 辛 characters. According to my sources, 辡  means 'litigation' and one of the older meanings of 辛 was 'criminal' and that 辡 'litigation' is a semantic compound character with one 辛 'criminal' and another 辛 'criminal' pointing to its meaning (two criminals litigating in front of a court).

Since 辛  doesn't mean 'criminal' today, someone correctly asked in the comments, what my sources were.

I wrote:

(I am) Inferring (that 辛 had the meaning of criminal) from the following:

《說文》《辡部》辡：辠人相與訟也。从二辛。凡辡之屬皆从辡。

And the existence and ancient interpretation of characters like 宰 辠 and 辜

《說文》《宀部》宰：辠人在屋下執事者。从宀从辛。辛，辠也。
《說文》《辛部》辠：犯法也。从辛从自，言辠人蹙鼻苦辛之憂。秦以辠似皇字，改為罪。
《說文》《辛部》辜：辠也。从辛古聲。

To which the person argued, that the 說文解字 dictionary is not a reliable source and that it regularly misinterprets characters, among other things also because it uses an extremely limited data set and that not a single entry in the entire work makes use of 甲骨文 data (due to it being unavailable).

A debate ensued which went on ad infinitum and produced enough material to be published as a small article:


Regular is a relative term, but I would argue that 說文解字 does indeed interpret characters correctly on a regular basis and to say that that it uses an extremely limited is debatable and makes no reference to 甲骨文 is debatable.

The authors of 說文解字 refer to 古文 characters which are (maybe not all of them) what we call Oracle bone script characters today (or Oracle bone script type characters to be more precise).

Example:

䇂部: 䇂：辠也。从干、二。二，古文上字。

二，古文上字 clearly says that 二 was 上 in what they call 古文.

Now 二 meant 上 in the Oracle bone script: It can be found in the 商 dynasty, the late 商 dynasty, the 西周 dynasty and in the 春秋 period. (说文新证上 p. 34). Are the authors of 說文 referring to what we call the Oracle bone script in this case? I think yes. Does this and many similar cases mean 說文 makes referrals to the Oracle bone script type characters in general? I think yes. Does this mean the authors of 說文 had access to what we call Oracle bone script type characters? Again, I think yes.

From what we know, Oracle bones date to about 1000 BCE, but that doesn't mean that Oracle bone script type characters were not used anymore after that date too. We either have no archeological evidence after 1000 BCE or texts found after this date were in general structurally a bit different and weren't considered to be Oracle bone script texts anymore. The transformation to later more modern scripts at that time however was gradual and many Oracle bone script style characters were still used. A great many of 甲骨文 characters can be found on bronze inscriptions, they can even be found in the much later Seal script and characters like 司， 工，己 and others are written almost the same way as 甲骨文 characters even today. With that said, the referrals to 古文 found in 說文解字 to me are referrals to what we call 甲骨文 in one way or another.

Maybe the authors had no access to the exact same pieces of bones and shells with Oracle bone script on them excavated in the 20th century, but that doesn't mean they didn't have access to what we now call Oracle bone script characters at all. The characters trickled down from 1000 BCE to later periods of time and people who wrote 說文 were referring to 甲骨文 style in one way or another.

Did they have a more limited access to Oracle bone script type characters than we have today? I don't know but I wouldn't be so sure. I would argue they had access to a lot of Oracle bone script type characters, how many I don't know, but I do know, that to say that 'not a single entry in the entire 說文解字 makes use of 甲骨文 data' is incorrect.

Further arguments to support that 說文 is regularly interpreting characters correctly, maybe more correctly than we can interpret them today:

The authors of 說文 had access to a ton of books written in the Seal script, many of which are lost to us now. They also used a language and had access to dialects and variants of a language which were all much closer to the time when many of these characters were created, which made research much easier for them in many ways in my opinion.

If something, the problem of 說文 is that its authors did not use methods scientifically rigorous enough and not the fact that the dataset they were working with was 'extremely' limited.

Furthermore, most characters in 說文解字 are phono-semantic compounds anyway which for the authors of 說文 must have been relatively easy to interpret, given that their language was so close in time to when these phono-semantic compound characters were created. Characters had much less time to change in pronunciation and in meaning.

Example:
In 的 the character 白 is the semantic and 勺 is the phonetic. None of it matches today, but in the past 的 used to mean 'bright white' and the pronunciation of 的 and 勺 matched almost perfectly according to later data:

的：《傳》發，發矢。的，質也。又叶之若切，音灼。
勺：《唐韻》之灼切

For the authors of 說文解字 this was much less of an issue, since the characters preserved their meanings and pronunciations much better at the time they were writing 說文

And..I think (disregarding 轉注字) the most difficult characters to interpret are 象形字 and 會意字 anyway (which were most prone to misinterpretation by the authors of 說文)  and they objectively form a much much smaller portion of 說文 so to argue that 說文 doesn't regularly interpret characters correctly is wrong.

Returning to the original topic of the video - did the 辛 character mean 'criminal' at one point in history?

Disregarding all online and book sources that confirm this:

http://www.zdic.net/z/25/xs/8F9B.htm https://baike.baidu.com/item/%E8%BE%9B/34249 http://vividict.com/WordInfo.aspx?id=3815

古文字詁林 Book 10 p. 1020

Looking at data only from 說文解字 we have the following:

辡：辠人相與訟也。从二辛。
宰：辠人在屋下執事者。从宀从辛。辛，辠也。
辠：犯法也。。。改為罪。
辜：辠也。从辛古聲。

Translation:

辡 : criminals engaging in mutual litigation. Formed by two 辛
宰 : criminals carrying out their activities under a roof. Formed by 'roof' and 辛. 辛 has the meaning of crime.
辠: means 'to violate law'... (later) character changed to 罪 (crime).
辜: means crime. Formed by 辛 and 古 phonetic.

The above to me is quite strong evidence that 辛 meant crime earlier in history. There are ways to argue how the above 4 entries could all be wrong, but it seems improbable. The last entry especially is very important:

辜 gū - crime
古 - gǔ (phonetic)
辛 - crime (semantic)

Again, there are ways to argue, that this could be a mistake of 說文解字, but it really is very improbable.