Writers of narrative nonfiction tend to write sentences like this one (which I wrote a few weeks ago): "She realized she would not be the next Sarah Bernhardt." But how do I know what went through her head?
The logical answer is: She said so. For example, Jonathan Harr provided this explanation in the author's note to A Civil Action: "The thoughts and feelings of the characters are based on what they told me they thought and felt, or what they described as their thoughts and feelings in transcribed court proceedings." Seemed fairly obvious to me.
Some say you need attribution ("she said later"). I see no point in clogging up the text that way. (In an unmarked footnote, sure.) They've argued that a lack of attribution frees the writer up to invent things, but the counterargument is that if a writer's of a mind to invent things, he's probably to invent the "she said later" as well.
Most journalism requires two independent sources to verify that something happened before printing it as fact. So, the traditionalists argue, how could anyone ever verify what's going through a person's head?
However, traditionalists would have no problem with my writing: "Critics had savaged her performance. Would she ever become the next Sarah Bernhardt?" These are verifiable facts: the critiques and the existence of Miss Bernhardt. But in putting the two together like that, the journalist is assuming what she's thinking and leading readers to think the same thing, just with the ability to weasel out by saying "Well I didn't actually write that."
To me, if you're going to put readers in people's heads, don't play games. Just do it -- and be ready to back it up in the notes.
Join the discussion at