See yesterday's discussion for an answer.
You are most likely copying unicode (M$ sence of it) characters from Word. Try to see it using HEX display option of the string control. If it has 2 bytes per symbol, so this is unicode I think.
Next open exported strings file and check font names in it. Then check fonts actually set for the wrong label. Don't use label, use Caption instead, BTW.
Discover which fonts you are using in Word document(unicode or not). Then you can try to export/save text as an ASCII file. I don't remember which is an exact file type for it in MS Word, Don't using it at all. Then open it in ,say notepad
and try again.
Another way to try is to use non unicode font to format your text and then try to copy it again (depends on Windows)
.
As a quick workaround you can see if the first byte of symbol is 0 or not. If so, just remove it and you'll get normal english text. If not, you can use divisor number, which will be different for the each language. This is clear from the unicode table (2 bytes). If you'll reach this place, let me know I'll try to help how to calculate it.
Hope that unicode is a reason of this behaviour.
P.S. You can determine unicode by open fonts folder and double klicking to fonts. If you see boxes instead of some groups of symbols, so this is most likely unicode (I don't insist on correctness of this way. On my WinNT box this works).
Sergey
