Before I get into my Nerdy Writing Project, I have to mention that the James River Writers' Writing Show last night was great. The theme was "Where the Wild Things Are: The Irrational World of Children's Literature," and the panel featured three children's book authors and a librarian who sits on the 2009 Caldecott committee. Happily for me, at least two of those authors write in middle-grade and young-adult, one of them in fantasy. Also, I bought a book: The Eternal Hourglass, which is Book One of a new series by Erica Kirov, one of the speakers. When I was checking the panelists out online, I saw the book (sadly, I can't now find the site where I originally saw it, which made it look even more awesome). I thought it looked really neat. Then, at the Writing Show, I saw the actual book and was kind of stunned by how, um, gorgeous it is. I feel sheepish about being swayed by the literal book's cover, but what the heck. I'd already thought it was cool, and the JRW events are served by a small local bookseller, so I feel good about buying there and spending more than I would on, say, Amazon.
Right! On to my Nerdy and Threatening-to-be-Endless Project: Scene Analysis. I got the idea from a Writer's Digest article that suggested going through books you like and noting how scenes of tension are often interspersed with catch-your-breath scenes that move the plot in ways that are perhaps less intense. While that was the inspiration for my project, I admit I've gone quite a bit further. Objective observations have always appealed to me, and I was curious as to whether splitting books into different kinds of scenes would yield results (say, for the sake of argument, in pie-chart form) that supported my subjective thoughts on various books, particularly in terms of pacing.
So I made an Excel spreadsheet. Well, actually, I've made five so far. I decided that the first fifty pages was enough of a book to analyze - a conclusion supported by the similarity, in the one book I analyzed all the way through, between the overall results and the first-fifty-pages results. Each book has its own chart, and each chart records the following for every scene:
Section Type. I'm dividing each book into four types of sections. I call them Scene, Scene* ("scene-star"), Summary, and Summary*. "Scene" means a section that takes place entirely in the moment - dialogue, a blow-by-blow description of action, even a character's thoughts - basically, anything that isn't summarized. "Summary," conversely, is stuff that doesn't tell you exactly what's happening right now. I found the two most common uses to be description and to note time passing. "Scene*" is any section that, overall, is definitely in the moment, but includes a non-negligible element of summary - say, several sentences of description, or a introduction that describes time having passed. "Summary*" is any section mostly not grounded in the moment, but with a smattering of lines that are, often a brief exchange of dialogue. Generally speaking, "Scene" and "Summary" correspond to the "Show" and "Tell" of writing. Not every book has all four section types. Howl's Moving Castle, for example, had no Scene* at all.
Length. Length in pages, in increments of 0.5. Sometimes something that should have been a section would be only a paragraph or so; if it seemed to really merit sectionhood, I rounded up, and otherwise just incorporated it into one of the neighboring sections (accounting, in some situations, for those sections going from Scene to Scene* or Summary to Summary*). It's not a perfect system, but it's difficult to measure pages to increments of less than one half, and I think the Scene* and Summary* labels help maintain the overall sense of what kinds of sections we're looking at.
Story Time. Length of story time represented by the section. Generally, these were Minutes, Hours, Days, Weeks, Months, Years, or Explanation (the last used for sections of Summary that are all description, and take no time in the story).
Intro. Here I recorded the introductory sentence, or part of it, or occasionally more than one sentence; as much as seemed relevant. This took by far the most space on the charts, but I found it worthwhile for the interesting discoveries. I've had trouble integrating the passing of time into my stories, and many Summary sections do that, often in the first sentence or so. Check out a few of these from Summary bits of Harry Potter and the Philosopher's Stone:
"Nearly ten years had passed since . . ."
"On Saturday morning, things began to get out of hand."
"Perhaps it was because he was now so busy, what with Quidditch practice three evenings a week on top of all his homework, but Harry could hardly believe it when he realized that he'd already been at Hogwarts two months."
"Quirrell, however, must have been braver than they'd thought. In the weeks that followed, he seemed to be getting paler and thinner . . ."
"In years to come, Harry would never quite remember how he got through his exams when he half expected Voldemort to come bursting through the door at any moment. Yet the days crept by . . ."
(All copyrights, etc., of course, belonging to J. K. Rowling.)
Synopsis. A very brief description of what happens in the section.
Having done that, I made a small graph at the bottom of each worksheet totalling the number of pages written in each section type, then used that graph to make a pie chart displaying percentages.
The pie charts show some fascinating trends. "Scene" is the big winner in terms of volume, in part because dialogue sections are so long in pages, and I have no practical way of counting words or otherwise being more precise about length. Other than that, though, there is surprising variation. The five books I've analyzed so far are Diana Wynne Jones' Howl's Moving Castle, J. K. Rowling's Harry Potter and the Philosopher's Stone, Susanna Clarke's Jonathan Strange and Mr. Norrell, Stephanie Meyer's Twilight, and Terry Pratchett's Mort. Keeping in mind that this process, while as objective as I could make it, is hardly perfect science, a few statistics are as follows:
Most Scene: Howl's Moving Castle, 84%
Least Scene: Jonathan Strange and Mr. Norrell, 52% (Twilight a relatively close second at 59%)
Most Summary: Jonathan Strange and Mr. Norrell, 24%, more than double that of any other book
Least Summary: Howl's Moving Castle, 2%
Most combined Scene and Scene*: Mort, 94%
Most combined Summary and Summary*: Twilight, 25%
Parts of this startled me. Terry Pratchett's books seem to include endless detached description, valuable because of its humor, but long nonetheless. I was startled at how little Summary I found in Mort. On the other hand, it had more Scene* (24%) than any other book, indicating the presence of many short but non-negligible Summary pieces. Another interesting discovery: amount of Scene versus Summary - similar though it is to Showing versus Telling - has less bearing than I had theorized on how fast the pacing seems to be. Of course, none of the books I've analyzed so far threatens whiplash, pace-wise. I plan to soon add Artemis Fowl and Mister Monday, both much faster-paced, to the list and see what that does to my conclusions.
Also: Like many writers, I've heard time and again that one should "show" and not "tell." Having looked at both in some detail now, I would say that there are definite benefits to both. It would be very difficult for J. K. Rowling to make a whole school year (not to mention ten years of growing) pass for Harry in one book without sentences like those above. If you're not doing it often - say, if it were only the one ten-year jump - it's fine to just leap ahead and make it apparent that the time has passed, but that can work less well with smaller time blocks and can be disorienting for readers.
I recommend this exercise to anyone who is interested in writing and has the time to spend on it. I wouldn't do this to a book I was reading for the first time, but it's less distracting than you might think, so I can enjoy a rereading while taking these notes. Alternately, if done without really reading the text (but on a book I know fairly well or have recently read), it's very quick to analyze the first fifty pages. A worthwhile project to help one really grasp the differences between, and implications of, showing versus telling.