A friend of mine used this site: http://www.typegreek.com/ to write a document containing Greek characters. The site converts Beta code – a way of representing written Greek using only ASCII characters – into Greek characters. I believe it uses the UTF-8 encoding.
For example, typing a into the input would produce an α.
When I received my friend's document over email and opened it in Word, the Greek characters had been converted back into Beta code.
For example, instead of ἑ, I would see e(.
Interestingly, this is not an issue caused by Word reading the file using the wrong encoding – if it were, the Greek letters would be displayed as unknown characters (e.g. �) or similar, rather than Latin characters.
Do you know why this would be happening, how it can be stopped, and (if possible) how documents such as this can be fixed?