I define it as how the narrator speaks to the reader. Like Magpie said, 10 people would tell the same story differently--different words, different sentence lengths--different gestures, if you could see them. If you picture the narrator telling the story in person, or over the phone, it's a lot easier to grasp his/her 'voice'.
That's probably why some instructors will tell you to set a place for your narrator at the dinner table and have a chat with him/her, get him/her to tell you the story.