What Do We Know About Literacy Rates in the Ancient World?

Mar 2015
Those estimates of 0,05...0,1% in Han Guangdong or Korea are suspect for me.
Going into the details of Egyptian sources: the best represented group is small town middle class.
But we see how the definition of "literacy" needs qualifications.
When they were requested to write 1...3 lines, 50 % of the Egyptian subscribers were able to do so. But when the length of the text to be written was 4...10 lines, just 27 % could. So, roughly half of the people who could write a couple of lines were "slow writers" who were forced to give up altogether when needing to write more.
Of the remaining quarter of men who did write longer texts at need, from quarter to half did so with heavy spelling errors (possibly poor handwriting), etc. So only about 10...15 % wrote flawlessly and comfortably.

People who worked as clerks in government employment would have needed to write reasonably comfortably and smoothly. Thus the people in society who could write included people whose jobs required habitual writing, plus people who did not hold such jobs but who were qualified candidates... plus substantial numbers of slow writers who could not write long texts, or could do so with heavy errors.