Joel of Far Outliers has a thought-provoking post discussing the question “Why are there so many writing systems in India and and so few in China?” He describes a 1994 article by Victor H. Mair from the Journal of Asian Studies:
The premise of Victor H. Mair’s wide-ranging article is that written Chinese emerged not as transcribed speech, but rather as a special, radically shortened cipher with its own grammatical and expressive conventions. He calls this written form Literary Sinitic (LS) and finds the disparity between it and any form of spoken Chinese, which he refers to under the general heading of Vernacular Sinitic (VS), is of a wholly different nature than the contrast between written Latin and any modern written or spoken Romance language. Indeed, he argues, Literary Sinitic remained incapable of serving as a means of recording spoken Chinese or any other language. Thus, for Mair, the question becomes: How did vernacular written forms emerge in a milieu in which Literary Sinitic dominated intellectual life? He finds the earliest instances of written Vernacular Sinitic occur typically in Buddhist texts. He believes the Buddhist emphasis on teaching through the local dialect (desa bhasa) was a major impetus for the development of written vernacular, but concludes it is difficult to determine exactly which aspects of Buddhism had the greatest influence on the slow maturation of written Vernacular Sinitic.
This provides a convincing answer to an interesting question. Mair goes on to say that the Manchus’ proclamation of kuo-yü [= guoyu, Mandarin] as the official written language of the nation “marked the formal end of the multimillennial separation between book language (shu-mien-yü [= shumianyu]) and spoken language (k’ou-yü [= kouyu]) in China.”
Addendum. Be sure and read xiaolongnu’s extremely interesting comment about the chain of spoken translators sometimes required in order to render sutras into Chinese, resulting in translations that “read more like the transcripts of lectures — complete with reports of the audience response — than like the translations of written texts.”