No. Political candidate ____ doesn’t speak like a 4th-grader. Reading scores don't explain speeches!
We in the US are now less than a month away from 2021’s election day. We’ll soon see some articles popping up telling us that candidate X “speaks at a fourth-grade level!”
Journalists, perhaps out of a love of the printed word, rush to apply the Flesch-Kincaid reading scale to political speeches. Trump, in particular, was a tempting target. He scored low on Flesch-Kincaid and spoke brutishly. For many, a reading test so conveniently solved the puzzle of his rhetorical success. “4th grade? Oh, that explains it.” This hot take on Trump showed up in The New York Times, The Washington Post, Newsweek, Politico, Esquire, The Week, The Daily Beast, and even Wired.
Earlier this year, Bloomberg News pointed out that Biden delivered his Inaugural address at the 8th-grade level. Both Fox News and The Week lamented when Obama delivered his 2012 State of the Union Address at the 8th-grade level.
Unfortunately, Flesch-Kincaid tests don’t explain political speeches.
The Flesch-Kincaid scale was developed back in 1975 to determine the difficulty of technical manuals. The readability formula is based on the number of words in a sentence and the number of syllables in a word. You dump a chunk of text into a Flesch-Kincaid reader, and it spits out the presumed reading level (4th grade, 10th grade, etc.). It’s a straightforward process, which probably explains its prevalence. BUT, the scale doesn’t evaluate meaning, which is kinda important in political talk. Speakers can score low and speak beautifully. Why?
First, syllable length is not the same thing as depth of thought. As humans, we tend to use shorter words when we speak. That’s not dumbing down content; that’s the nature of spoken language. According to Flesch–Kincaid, Martin Luther King’s “I Have a Dream” comes in at 8th grade. Have you met many 13-year-olds with King’s mastery of metaphor? Of course not. We don’t judge King’s oratory by the complexity of his consonants, but by the content of his characterizations.
Second, we often don’t speak in sentences, at least not sentences as they appear on a printed page. We don’t distinguish between a comma, semi-colon, and period when we’re talking; we vary pause length. Any transcript with punctuation is, at best, an approximation of what was said. The one thing that a Flesch–Kincaid test can do, measure sentence length, is completely arbitrary when applied to speech.
Mark Liberman, a Linguist at Penn State, ran Trump’s candidacy speech through a Flesch–Kincaid analysis with varied punctuation and posted the result on Language Log.
It’s coming from more than Mexico. It’s coming from all over South and Latin America. And it’s coming probably — probably — from the Middle East. But we don’t know. Because we have no protection and we have no competence, we don’t know what’s happening. And it’s got to stop and it’s got to stop fast. [Grade level 4.4]
It’s coming from more than Mexico, it’s coming from all over South and Latin America, and it’s coming probably — probably — from the Middle East; but we don’t know, because we have no protection and we have no competence, we don’t know what’s happening. And it’s got to stop and it’s got to stop fast. [Grade level 12.5]
Any measure that can have that much variance based on where the journalist just happens to place punctuation isn’t worth the column space. Also, the implication is that we long for the halcyon days when our political leaders routinely declaimed in polysyllabic run-on sentences.
The linguist John Haiman has written on our “cult of plain speaking,” where sincerity of speech seems to reveal the authenticity of character. Of course, no speech is merely “plain.” Language is chosen. How we evaluate this plain speech is the challenge, but applying a reading test to political talk won’t give us the answers we need.