Etymology of 四
四 is a special self-explanatory character. Its oracle glyph represents it is the twice of 二. The original idea of character construction: twice of 二. Its bronze glyph is derived from the oracle glyph. If we write the strokes of the bronze script vertically, it becomes ; and then combine and (two, meaning 4 is twice of 2), it becomes . Some bronze script is written as adding a two to , emphasizing the multiple relationship between "four" and "two" . The seal script omits in the bronze script .
My Annotation: note that there's NO or in oracle scripts. Dictionary of Pictograph here discusses three types of scripts -- oracle, bronze, and seal. For more interesting glyph of 四, see the Chinese etymology site.
Explanation in 甲骨文字典 (Dictionary of Oracle Scripts)
[Character Forming] The method of counting strokes is used for representing the number in oracle 一, 二, 三, and 四. Their glyphs originate from counting rods that are put horizontally. Early bronze scripts were also carved as (on 孟鼎 Meng cooking vessel), (on 牆盤 Wall plate) -- the same as oracle glyph. During the Warring States period / Spring and Autumn Period, bronze 四 became (on 郘鐘 Shao bell), (on 徐王子鐘 Prince Xu bell), (on 大梁鼎 Big Liang cooking vessel) -- they were the origin of the seal glyph in Shuowen. The bronze 四 was a loan character. In 數名古誼 (The Ancient Meaning of the Chinese Number Names) by 丁山 (Ding Shan), the origin of 四's bronze glyph is explained as "from 呬(breathing) in 呬息 (breathing)"; In 中國文字之源流與研究方法之新傾向 (New Trend of Research Methods of the Etymology of Chinese Characters) by 馬敘倫 (Ma Xulun), it is explained as "from 泗 (snot) in 涕泗 (tears and snot)".
My Annotation: loan character / phonetic loan character / rebus character, as the name suggesting, are characters that are "borrowed" to write another homophonous or near-homophonous morpheme. So for the question "where did bronze glyph 四 come from", 甲骨文字典 (Dictionary of Oracle scripts) adopts two viewpoints: it came from 呬 or 泗.
孟鼎 Meng cooking vessel. Kept in National Museum of China.
牆盤 Wall plate. Kept in Zhouyuan Institute of Administration of Cultural Relics, Fufeng Town, Baoji City, Shaanxi Province.
郘鐘 Lü bell. Kept in Shanghai Museum.
徐王子鐘 Prince Xu bell.
Let's play a game: can you find
四 in these rubbings? Open the links in a new window/tab to view full-resolution image
郘鐘拓片 (Rubbing of Lü bell).
徐王子鐘拓片 (Rubbing of Prince Xu bell).
大梁鼎拓片 (Rubbing of Big Liang cooking vessel).
The major points of statements for 四 in Dictionary of Pictograph are
亖 meant twice of two.
was the result of combining and , emphasizing four was twice of two.
(Though Dictionary of Pictograph doesn't clearly state that, it implicitly suggests) the bronze four came from .
For statement 1, it is OK to explain it like that. Either
counting strokes gives the correct answer. In 說文解字注 (Annotations on Shuowen) by 段玉裁 (Duan Yucai), it was
This is a calculation "two and two make four".
"二二如四 (two and two make four)" would support the "multiple relationship" viewpoint, but not necessarily be the truth. Because there is already a "multiple relationship" rule (we will discuss it in the Etymology of 十) for carving a number:
>= 5 times of ten, combine the multiplication factor below the character ten. For example,
= 50, = 60
It was not necessary to make a specific rule for constructing
亖. So I would prefer the "counting strokes" explanation.
For statement 2 & 3, if there's any oracle scripts carving like , then no doubt it will be the correct answer. However, not only the oracle has never been found, but also, classical annotations by ancient scholars didn't support such a statement. The entry 四 in Shuowen says
Like dividing a shape into four pieces. Characters that related to 四 have the radical 四. 𦉭 is the ancient form of 四. 亖 is the oracle glyph of 四.
Shuowen's viewpoint about the transformation from 亖 to 四 was then further confirmed in Annotations on Shuowen by Duan Yucai in Qing Dynasty:
The ancient form (some bronze script) was just like that. Seal script 四 modified it a little.
This is a calculation "two and two make four". Because the two strokes in 二 have equal length, the four strokes in 亖 should also have equal length. 覲禮 said "四享", and 鄭玄 commented on that: "The four should be three. Both writing three and four are all counting strokes, 三 and 亖 look alike, so it is misread." Similarly, for "朝貢禮純四只" in 聘禮注, 鄭玄 said "four should be three"; "天子巡守禮制幣丈八尺純四𦐖" in 周禮內宰職注, the same, "four should be three"; "是四國者專足畏也" in 左傳, 劉炫 said "four should be three". Thus, these mistakes were all because of the similarity of 三 and 亖: following the convention of Shuowen, when copy-writing a book, one should write oracle first and then ancient bronze glyph, but the mistake would happen when messing up the order.
In short, Duan Yucai's opinion suggests: "because 三 and 亖 can be easily mistaken, seal script 四 modified a bronze glyph 𦉭 a little to avoid that."
So, now we can see, both Annotations on Shuowen and Shuowen didn't support the statement " was the result of combining and ". What's worse, the statement of Dictionary of Pictograph contradicts the bronze glyphs 𦉭 and (on Lü bell)!
But how about the statements for "oracle 亖 => bronze 四" in Dictionary of Oracle Scripts? Are they more reliable? I would say both 丁山 (Ding Shan) and 馬敘倫 (Ma Xulun) had done much more research work on it, though whether they were correct or not has not been confirmed yet, they did provide brilliant opinions on how the transformation happened. And there's no contradiction between their statement and the bronze scripts we have found. Thus, the statement of Dictionary of Oracle Scripts for 四 is more reliable.
Part of paper 數名古誼 (The Ancient Meaning of the Chinese Number Names) by 丁山 (Ding Shan)
Part of paper 中國文字之源流與研究方法之新傾向 (New Trend of Research Methods of the Etymology of Chinese Characters) by 馬敘倫 (Ma Xulun)