4 ms·
That seems on par with https://en.wikipedia.org/wiki/Woordenboek_der_Nederlandsche_Taal https://en.wikipedia.org/wiki/Woordenboek_der_Nederlandsche_..., a simil
by Someone 9y ago
That seems on par with https://en.wikipedia.org/wiki/Woordenboek_der_Nederlandsche_Taal https://en.wikipedia.org/wiki/Woordenboek_der_Nederlandsche_..., a similar work for Dutch, which took from 1849 to 1998 to complete. To get to the finish, they did move the goalposts a bit by adding "as used before 1976" at some point. It also was digitized fairly early by having two groups of people in, IIRC, India type out a copy in duplicate and comparing the results.
(No, you don't have to speak a language to type out a manuscript)
- tgb 9y agoNot knowing the language might even help in accuracy, though probably not in speed! I've heard that copy editors would sometimes read the text backwards as that made spelling errors more obvious. When read as part of a sentence it's easy to skip over a spelling mistake.
- thaumasiotes 9y agoDifferent classes of transcription error tend to come from literate vs illiterate transcribers. I believe haplography and dittography (writing a sequence of letters once when it should be written twice, or twice when it should be written once) are more common from illiterate transcribers.