The Jark on the Spiv’s Deed

I love getting reviews that make me think.

I got a review of a book chapter this morning that did that. This reviewer told me I used obscure words that would be unfamiliar to readers. Well, I can’t argue with the merits of that advice in general. I love Hemingway in no small part because of his spare style. When an author uses a word that I have to look up, it almost always annoys me. It’s not that I don’t like learning new words–to the contrary. The problem is that looking up a word takes me out of the story. It breaks the fictional dream, which is at the heart of modern fiction.

My first reaction was to run the two tests of readability that come with most word processors: the Flesch grade level and the Flesch-Kincaid reading ease score. My chapter had a Flesch grade level of 4.1 and a reading ease score of 82.5. According to Wikipedia, that grade level score means that a US fourth-grader should be able to read and understand it. The same Wikipedia article says a reading ease score of 90 is easily understood by an eleven-year-old, while one of 70 is easily understood by a thirteen-year-old.

It’s interesting to see how various authors stack up on the readability scales. This blog actually gives some comparative scores. The Old Man and the Sea, for example, has a grade level of 4 and a readability score of over 90. This means that Hemingway’s masterwork is exceptionally readable: lower grade levels and higher readability scores are good things when it comes to lucidity. Authors like Stephen King, Dan Brown, and John Grisham have grade levels of 6 to 8 and reading scores from 75 to 80. These are all readable authors by any measure. Depending on the passage you choose, other famous authors don’t fare so well. One passage in particular from Proust has a readability score of -515.

This blog isn’t doing so well, either. It has a readability score of 67.3 and a grade level of 7.7. If this were fiction, I’d be looking for ways to make it more readable.

The point is that if you are writing fiction, you probably don’t want your use of language to obscure your meaning. You want your fiction to be readable. You’re not trying to impress your reader with your proclivity for polysyllabic elucidation. Oops. I meant to say, “your habit of using big words.”

Okay, then, back to my thought-provoking reviewer. My scores are roughly the same as King and Grisham, so that must mean my chapter is peachy-keen, right? By the way, “keen”–as in making a high-pitched sound–was one of the words my reviewer objected to. If you think the scores absolve me, think again.

Consider this sentence

The jark on the spiv’s deed wricked the truth.

This unintelligible sentence has a readability score of 100 and a grade level of zero, so by the above measures it’s more lucid than, say, the children’s book Goodnight Moon, which has a grade level slightly less than 3.

Say, what?

Suppose I had written instead:

The forged seal on the criminal’s deed twisted the truth.

It turns out, this says the same thing as the earlier sentence, except it uses plain English rather than obscure words. It has a grade level of 3.6 and a readability score of 86. It’s also a sentence anyone reading this can likely understand.

So what’s going on? Like Sheldon on Big Bang Theory, I like to do the math. If you look at the formulae for the Flesch scales, you’ll see that they are based on counting characters per word and words per sentence. In fact, all the scales that mechanically calculate readability do the same kind of thing. What they don’t do is decide whether or not your words are obscure. So my first sentence, the one with the jark on the spiv’s deed, is short and uses one-syllable words, hence the high readability and low grade level scores. Readability in this case is a deliberate artifact–I constructed a sentence to “falsify” the score.

Now, another way to test readability is have real people read the sentence and see how many know what it says. In fact, you could have fifth graders read it and find out if the “average” –whatever that means–fifth grader understood it. Of course, that kind of real-world test can’t be automated and built into your word processor. I think we don’t need a test, though, to conclude the first sentence above is unreadable, despite its score to the contrary.

Does this mean that the Flesch scales and all their relatives are useless? No, of course not. In fact, if you compare the results of the Flesch scales with the above real-world test using real people, the results generally have a 90% correlation. That’s really quite good, especially for social science results where there are often multiple unmeasured variables. One of those unmeasured variables in this case is how common the words are in the passage being tested. As the above example shows, it’s an imperfect measure. Still, for most documents, readability fails more due to long words and complex sentences rather than obscurantism. Thus, the scales do in fact measure something useful. They’re just not 100% accurate.

This takes me back to my thought-provoking review. It’s a matter of taste and style whether or not use a particular word. Indeed, Dean Koontz has almost made a trademark of using at least one obscure word per novel–kind of like Alfred Hitchcock’s appearance in all of his films. He does this with a certain insouciance–one of his words–and this is no doubt part of his appeal to his legion fans.

Now, the fact that Dean Koontz does this doesn’t make it “okay” for a beginning author. If I used the wonderful word “insouciance” in an opening paragraph, I’m sure the acquisitions editor would stop reading and toss my manuscript in the reject pile. Koontz, on the other hand, could open with the page 178 of the Tulsa phone book and he’d still get a contract and readers would still buy his book. He’s not only more talented, he’s a best-selling author. The “rules” don’t apply to him: nothing sells like success. Whether it’s an agent, editor, or an ordinary Joe picking a paperback out at the supermarket, Koontz’s solid reputation means that he doesn’t have to prove himself with those first paragraphs. I do.

So how about that verb “keened” that I used? Well, it’s onomatopoetic (there I go again!), so I like it. I’m also pretty convinced it’s not so obscure that people will have to look it up, although I concede it’s not a word most people would use in conversation. Bottom line: I think I’ll keep it, but I’m glad I thought about it.