According to legend, the Chinese writing system was produced four and a half millennia ago by a four-eyed man called Cangjie, the court historian under the Yellow Emperor, the mythical father of Chinese civilization. Seeking to represent everything by a distinguishing characteristic, he devised thousands of symbols for each word of his language.
In actuality, the oldest specimen of Chinese writing is a little over 3000 years old, and features characters so different from modern Chinese that until about 100 years ago, no one had realized it was writing. Of course, that was before paper or even bamboo strips had come into use as a writing surface. Rather, writing was etched into turtle shells and ox scapulae (shoulder blades) for use in divination.
You might think writing out the sounds you make in your language is the most obvious – perhaps the only obvious – way to write down language. But the ancients did not have the benefit of hindsight. Many civilizations developed methods for recording ideas, especially accounts and transactions, thousands of years before proper writing emerged, involving pictographs and ideograms which expressed messages, but not the speaker’s language per se. Systems which transcribed language emerged from these earlier systems after thousands of years of use. These new, proper writing systems, like their antecedents, were composed of pictographs and ideograms. Written language was developed independently in at least two places, Mesopotamia, where Cuneiform was born, and Mesoamerica, where the Olmec began to write 2500-3000 years ago, and both languages follow these patterns. Egyptian hieroglyphs and Chinese characters may also have emerged independently, and both used images to depict the meaning of the words they represent as well. Generally, however, these systems were not oblivious to the sounds of the words recorded, and suggesting the sound of the word intended was often necessary for effective communication. In Egypt, this led to hieroglyphs representing consonants (but not vowels), and in China, symbols were used as syllabic homophones for other words.
The first Chinese dictionary to attempt to explain the logic behind every Chinese character was the Shuowen Jiezi, which placed all Chinese characters into one of six categories.
First is pictograms. These are pictures of what the words refer to, often used for animals. 犬 is a dog. 象 is an elephant. 虎 is a tiger. These were often somewhat clearer when they were first written thousands of years ago, but they were crude, like cave paintings. Some still look vaguely like what they depict, even in their modern form. 人 is a person. 山 is a mountain. Some are clear, but their meaning has changed. 木 is a tree, but its meaning is wood.
Second is ideograms. These are pictures too, but they do not refer to what they depict, rather an idea behind what they depict. These are common for more abstract things. 一，二，and 三 are all ideograms, essentially just Roman numerals turned 90 degrees. 上 and 下 are arrows pointing up and down.
Third is compound ideograms. These are characters which put one or more characters together to suggest a meaning. The Chinese word for forest is 森林, the first character being three trees together, and the second showing two. Each suggests a place where there are many trees. Another example is 休, meaning rest. On the left is a person (a reduced form of 人), resting by a tree.
Fourth is rebuses. These are for words which are especially hard to express, leading writers to simply use homophones in their stead. For example, 四 depicts nostrils, but its meaning is four. 北 depicts two people standing back to back, suggesting a meaning of back, but its meaning is north.
Fifth is phono-semantic compounds. The vast majority of modern Chinese characters are of this kind. These are words which have one part suggesting the meaning of the word, and one part suggesting the sound. 背 is an example. Since 北 originally meant back, but was co-opted to mean north, when the meaning of back was intended, 背 was invented to “reclarify” the character. The original is on top, suggesting the sound, and 月 has been added underneath. 月 is the moon (a pictogram), but here, as in many characters, it serves as a reduced form of 肉 (depicting sliced meat), meaning flesh, common in words for body parts. As with English, suggestions about pronunciation are centuries if not millennia out of date, and as such, often misleading. While the Shuowen Jiezi suggests 信, meaning trust (among many other things), is a compound ideogram, depicting a person and words (言 depicts words coming out of a mouth, 口) later scholars consider it a phono-semantic compound, with 人 suggesting the sound and 言 suggesting the meaning. The modern reading of 信 is “xìn”, and the modern reading of 人 is “rén”. However, in Old Chinese, 信 is reconstructed as “snjins” and 人 is reconstructed as “njin”. For similar issues in English, see “ghoti”.
And last is the derivative cognates, or transfer characters. No one really understands these. Shuowen Jiezi cites 考 and 老, which look similar, and rhyme. Derivative cognates are extremely rare, and honestly, I have nothing to say about them.
For my game’s language, I’ve heard a lot of people throw around all kinds of words for the writing in my game. Strictly speaking, I would call it a syllabary, with each symbol corresponding to one unique syllable. There are ideographic elements, such as in using parallel lines to indicate “yes” (suggesting agreement) and using perpendicular lines to indicate “no” (suggesting disagreement), but ultimately the symbols come from math out of a desire to draw on something abstract and absolute, rather than something physical and concrete as the ancients often did. So far, I’ve chosen the symbols I have on two criteria: their significance to mathematics, and aesthetics. There are two “pages” of 20 characters which are fully implemented at this point. The first draws on Euclid’s definitions from the beginning of the Elements, and the second draws on proofs that follow. At this stage, my process is designing symbols, writing sentences, and then trying to match them up.
I hesitate to call it a syllabary on the other hand, because I never intended for pronunciations to come into gameplay. I think it might be fun to have one or two challenges involving rebuses, but for the most part I think players should be able to beat the game without having any idea how to actually say anything. Ultimately, the association between symbol and word is arbitrary – at least in the current stage of the game. It’s something I’ll continue to refine as I implement more of the game’s vocabulary.