I found my first real case of minim confusion, which I previously said was theoretically possible in Manchu because medial “a”, pre-consonantal “n”, and one form of “k” are all made up of identical strokes.
By my “first real case”, I mean two words that are attested in dictionaries, having the same written form but different pronunciation, i.e. they are homographs.
First of all, the theory behind it. Initial “a” looks like . Initial “e” looks like . Medial “n” when followed by a consonant looks like , so when you have a word that starts with “en” followed by a consonant, the “en” looks like , the same as initial “a” .
Sima gave me a copy of Gertraude Roth Li’s Manchu: A Textbook for Reading Documents last week, and I’ve been sticking my nose in it whenever I have a minute. I had just gone through romanizing the first reading selection, and noticed that one of the words could be romanized in two different ways: acu or encu (“c” sounds like English /ch/), which both look like:
Acu is the sound Manchu people make when they sneeze (or when they touch something hot). Encu means “separate”, “foreign”, or “different”.
Another kind can occur with medial “k”, which looks like. So “nka” would be the same as “kan”, either at the end of a word, or before a consonant. Haven’t found one of those yet though. But it does present the possibility that some words could have an undetermined pronunciation. In most cases one should be able to verify a word by checking how the Xibe pronounce it, but there is a miniscule chance that there are some words with this pattern that are not in any dictionary because they’re in a document that’s rotting away in some underfunded archive somewhere. It’s an extremely small chance, but a chance nevertheless.
Another thing reducing the size of the chance to the nano scale is vowel harmony. A word with “a” in it normally wouldn’t have “e”. I’ll be surprised if I find another one of these rare gems.