This is an embarrassingly basic question. In the first "Hello World" example in the book Java Persistence with Hibernate they have a very simple class:

Message
- id : Long
- text : String
- nextMessage : Message

What I'm having a difficult time with is the fact that they specify a many-to-one relationship in the Message.hbm.xml mapping file for nextMessage. I simply don't understand why it's many-to-one instead of one-to-one.

My understanding of "one to many" is that one row in some table may refer to multiple rows in another table. Whereas with "one to one", one row in one table refers to exactly one row in another. Since a Message only has "one" nextMessage, why wouldn't this relationship be one to one?

Thanks.. sorry for the basic question.

Your question is very valid.

Think of it as if you were mapping all objects of the child class altogether, not one individual row of it.

"Many" (rows, objects) of the class Messages will have "one" parent.

The reverse association of a "one-to-many association", is a "many-to-one" association, even if each individual message only relates its one parent.