mysql: possible loss of UTF data from char to text field type conversion?

I have a database that I use for a foreign language / vocabulary web application and was interfacing with it through phpMyAdmin (and of course php). I had about a thousand rows for Arabic text, which had been put into a char field. I wanted to expand the size of my entries, so I read that a text field could get me past the 255 character limit. I changed the field type to text- and all of the Arabic entries now display as question mark nonsense etc. I've dealt with plenty of that type of display throughout the site's design, making sure that all of the character sets etc are right, and it has been displaying fine for months until this change.

I returned all of the table's structure to its origi开发者_StackOverflow社区nal state to the best of my knowledge, and the problem persists. This is a pretty catastrophic loss, and I'm basically wondering if these rows have just been scrambled irreversibly, or if I'm simply just not resetting things properly. Any ideas on how to get this data back?

Thanks so much in advance.

---

Maybe you should specify the character set in the alter table statement, because it may default to some nonsense (eg. latin1).

If the data is lost then it is permanent. You need backup to solve this. :-(