Yesterday, Anne Gallagher, Michelle, Scott and I tried an experiment that involved having readers try to match the writer to the writing for 4 short excerpts. (If you still want to guess, feel free to comment on the post before reading further.) The idea to do this stemmed from a discussion over at Scott's place about voice. We had several ideas running. The first was that the narrator's voice may vary with the story, but does writing also have an underlying "authorial voice" or style that a writer can't necessarily hide? Does this authorial voice always peek through? Is it limited by, say, the writer's own vocabulary, intellect, and experiences?
I was surprised to see the wide range of guesses we got. Only one person, Alex, got the right answer (woohoo!), although Scott and Michelle would have also guessed correctly. (Anne, would you have gotten it right as well?)
The correct answer is:
Honestly, I was surprised by what a hard time everyone had on it, including myself. My own guess was wrong, as I switched Scott's and Anne's samples. I realize, then, that the elements I look at in someone's writing are not necessarily what other people, including the writers themselves, look at.
My reasoning behind my guesses will probably seem odd, as your reasoning will probably sound odd to me...although we would love it if you would describe that reasoning. For me, I easily identified Michelle as the writer of 3 because of the description "long, thin body lying limp" and the word "fluttered". I didn't need to read much further than that. Naturally, I knew which one I wrote. I had a harder time trying to decide between Anne and Scott for the first two samples, and I finally decided Scott didn't write the first one because of the use of the words "foolishly" and "foolish", a form of repetition that I personally use a lot, but one I wouldn't have guessed Scott would like. I credited it to Anne.
As I was trying to decide, I also considered things like sentence structure, as well as logic. I am more familiar with Scott's writing than Anne's, and I have seen a certain extended logical thought process in Scott's writing exemplified by lines such as "Algiers is a dangerous city, especially for men who take employment with the French oppressors". But, I detected the same kind of logic in the extended metaphor in the second paragraph of sample 2.
What this makes me realize in myself is that what I consider to be that "authorial voice" has to do with vocabulary, sentence structure, explanations of logic, and rhythms. I didn't pay much attention to subject matter, or to character. But, of course, the criteria are apparently different for everyone, given the different answers. And, so, I'm still left wondering, What makes up an authorial voice, and is there really such a thing?
Granted, this experiment had some flaws to it. The writing samples are short, and if the authorial voice has more to do with the limits of the writers ability--which I tend to think it does--then the short sample probably doesn't yet reach the boundaries of those limits. What would happen if you read multiple books by any of us? I'd bet those limits would become much more obvious.
And, how exactly did we choose our samples? I bet if any of us had wanted to, we could have picked writing samples that were more misleading and/or random. I can't speak for the others, but I'll say that for me, there was no value in that. I decided to choose two writing samples that I felt reflected polished work in its final state. Often times my early drafts are much more tainted by other writers' voices, and I hoped that these samples of mine were rid of such things...although I still see some McCarthy in my second sample. What was important to me was to show what I considered to be my best work in the hopes that people would reveal commonalities in it that I didn't necessarily see before or that people would reaffirm that the elements I consider to be mine were actually apparent. Neither one seemed to be the case.
The other complication is that voice comes on many levels. As we discussed at Scott's blog, there's always the narrative voice or the narrator's voice. This probably varies with the story. If there's also an authorial voice, the extent to which that authorial voice makes itself visible will also vary. In my own work, I personally try to hide that, aiming more for the invisible author. I would have said that in the books I've read by Scott, his authorial voice is more visible, more poetic. Michelle is more visible and poetic to me as well. I would have said Anne was less visible, based on these samples.
Do we care? Well, you tell me. If you were the one to put two samples up, would you prefer that readers could pick your work out right away or that they couldn't? I'm very much on the fence with this question. I do like the idea that I can write broadly. One of my favorite compliments is hearing someone say that they thought one of my stories was written by a woman or by someone other than who I am. At the same time, I was honestly a little disappointed that only three people assigned the 4th sample correctly to me. Those stats, in case you are interested are below:
For sample 1:
6 people correctly assigned it to Scott.
4 people assigned it to Domey.
3 people assigned it to Anne.
0 people assigned it to Michelle.
For sample 2:
5 people assigned it to Michelle.
4 people assigned it correctly to Anne.
4 people assigned it to Scott.
0 people assigned it to Domey.
For sample 3:
6 people assigned it to Domey.
4 people assigned it correctly to Michelle.
2 people assigned it to Anne.
1 person assigned it to Scott.
For sample 4:
4 people assigned it to Anne.
4 people assigned it to Michelle.
3 people assigned it correctly to Domey.
2 people assigned it to Scott.
One strong point in terms of identifying the components that make up voice might be to look at the writers who were ruled out of certain samples. No one thought Michelle wrote 1. No one thought Domey wrote 2. Obviously, there are elements there that people don't associate with these writers.
What about spread? Is there significance to there being slightly less spread in sample 1 versus sample 4? Does 1 maybe have a more prominent voice compared to 4? Is the author of 1 more visible than the author of 4?
I also wonder if factors like credibility came into play. Did people assign their favorite passage to the writer they respect the most? Their least favorite to the writer they respect the least? Did our personalities have anything to do with it? Did our past blog posts have anything to do with it? Can I possibly ask any more questions?
So, tell us your thoughts!