Ernest Hemingway writing a letter, probably.

Should we worry about Readability?

Last week I was taken in. For once, they saw me coming. I was out and about, making my way into the city, attempting to bypass the news feed to get at some messages. And yet something caught my eye. A blog, which shall remain nameless, had sponsored a post containing the words “surprising”, “reading” and “analysis.” They had me at “this.”

Scanning through the oddly long piece, I saw the author had compiled “readability” levels of a wide variety of authors. He had obtained this data by copying and pasting excerpts from his e reader into an online reading test generator, a method he had apparently stumbled upon. The idea had been generated by a real life conversation.

The data yielded startling results. Ernest Hemingway was indecently readable, leading a host of literary greats, more recent releases, best sellers and ‘political’ texts (ghost written auto biographies). The author had also taken the time to include a couple of writers who had ‘bought their way onto’ best sellers lists in the US. Maybe he didn’t realise his piece would be bought onto my feed.

A variation on the Fleisch Kincaid test, which assigns a US grade for different levels of readability, also happens to be built into WordPress, which we use here at HeadStuff. Both are equations where word count is divided by the number of sentences, and the number of syllables by the number of words. As it turns out the test was originally to test the readability of manuals, to be used by the US armed forcesExciting stuff.


The piece concluded that with the desired lower scores on the test, the more accessible the thesis of the text would be. It turns out Hemingway uses more plain language than his friend and contemporary F Scott Fitzgerald, who was placed further down the list. Perhaps if F Scott had been around today he would have pursued a career in content marketing rather than script writing, with similarly disappointing results. It is after all Ernest who has his own branded reading ease app, from beyond the grave.

Jane Austen fared better, though I began to wonder whether there were extraneous circumstances. For example, in reading Cormac McCarthy, would it not be useful to read the body of literature concerning conquest and exploitation in the American West, in order to get a sense of place? Can you really read Tom Clancy without completing Rainbow Six on PSone?

More pertinently, in Austen’s case, do we need knowledge of British colonialism to appreciate the more sinister undertones in Mansfield Park? When it comes to Hemingway, a readability test cannot measure who is writing. While his prose is in one sense clear, it projects a singular vision of masculinity.

Traditionally, Western notions of the literary canon often lack in representation of women, other nationalities, and sexualities. A convincing school of thought argues the thoughts and lives of the marginalised are continually rendered into the dominant mode of expression, meaning they are not truly expressed no matter how plain the language.

In this context, is it actually possible to just slap a metric on a text, without contextual information? It seems a dangerous precedent to lob all of these rich and complex texts into a vacuum. It was also an idea which seemed familiar. In 2016, one of the biggest fake news stories, turned out to be fake news itself. That said, in combination with social network feeds, fraudulent updates and partisan algorithms form a dangerous cocktail, filled with “filter bubbles” and “echo chambers.”

Ultimately, it wasn’t necessarily the substance of the article which bothered me, but more the principal. Rather than crude, obvious clickbait, this represented a more insidious attempt to keep people online. Though what we read should never be politicised, we should spend our time more wisely.

There have long been concerns that with more and more information at our fingertips, we become paradoxically more isolated, and harder to reach. (Unless you’ve got dolla for promoted posts.) Though there is data and evidence to support both sides of the argument, the overall feeling in my personal bubble was leaning towards the saturation argument.

I had recently moved to New York. A friend told how they’d felt over stimulated and depressed after wandering through Time Square at sundown. I found this oral content highly relatable. On the other hand there were piles of events and opportunities you wouldn’t get at home. Cake by the ocean and all that.

At the Center for Fiction, a panel recently discussed the modern family in literature. Much of the discussion focused around representation, with Tanwi Nandini Islam remarking that people need to read more writers from other parts of the world. It is not enough to take any handful of writers as representative of Bangladeshi literature, for example. When the discussion turned towards education, Min Jin Lee made cited the quantity of research linking screen time to degradation in our frontal lobes. We simply aren’t reading enough content with bad readability scores. We aren’t reading enough full stop.