The oracle bone inscriptions were discovered at what is now called the Yin Ruins near Anyang city in 1899. Sumerian cuneiform is currently regarded as being the oldest known writing system having originated about 3200 B.C. In a 2003 archeological dig at Jiahu in Henan province in western China, various Neolithic signs were found inscribed on tortoise shells which date back as early as the 7th millennium BC, and may represent possible precursors of the Chinese script, although there has been no link established so far.
Four percent of Chinese characters are derived directly from individual pictograms (Chinese: 象形字; pinyin: xiàngxíngzì), and in most of those cases the relationship is not necessarily clear to the modern reader. Of the remaining 96%, some are logical aggregates (Simplified Chinese: 会 意字; Traditional Chinese: 會意字; pinyin: huìyìzì), which are characters combined from multiple parts indicative of meaning, but most are pictophonetics (Simplified Chinese: 形声字; Traditional Chinese: 形 聲字; pinyin: xíng-shēngzì), characters containing two parts where one indicating a general category of meaning and the other the sound, though the sound is often only approximate to the modern pronunciation because of changes over time and differences between source languages. The number of Chinese characters contained in the Kangxi dictionary is approximately 47,035, although a large number of these are rarely-used variants accumulated throughout history. Studies carried out in China have shown that full literacy requires a knowledge of between three and four thousand characters.
In Chinese tradition, each character corresponds to a single syllable. Most words in all modern varieties of Chinese are polysyllabic and thus require two or more characters to write. Cognates in the various Chinese languages/dialects which have the same or similar meaning but different pronunciations can be written with the same character. In addition, many characters were adopted according to their meaning by the Japanese and Korean languages to represent native words, disregarding pronunciation altogether. The loose relationship between phonetics and characters has thus made it possible for them to be used to write very different and probably unrelated languages.
Just as Roman letters have a characteristic shape (lower-case letters occupying a roundish area, with ascenders or descenders on some letters), Chinese characters occupy a more or less square area. Characters made up of multiple parts squash these parts together in order to maintain a uniform size and shape — this is the case especially with characters written in the Sòngtǐ style. Because of this, beginners often practice on squared graph paper, and the Chinese sometimes use the term “Square-Block Characters” (Simplified Chinese: 方块字; Traditional Chinese: 方塊字; pinyin: fāngkuàizì).
The actual shape of many Chinese characters varies in different cultures. Mainland China adopted simplified characters in 1956, but Traditional Chinese characters are still used in Taiwan and Hong Kong. Singapore has also adopted simplified Chinese characters. Postwar Japan has used its own less drastically simplified characters since 1946, while South Korea has limited its use of Chinese characters, and Vietnam and North Korea have completely abolished their use in favour of romanized Vietnamese and Hangul, respectively.
Chinese characters are also known as sinographs, and the Chinese writing system as sinography. Non-Chinese languages which have adopted sinography — and, with the orthography, a large number of loanwords from the Chinese language — are known as Sinoxenic languages, whether or not they still use the characters. The term does not imply any genetic affiliation with Chinese. The major Sinoxenic languages are generally considered to be Japanese, Korean and Vietnamese.