Nabokov's Favorite Word Is Mauve

by Ben Blatt

  The black bar represents the reading level of the median book in each decade. The shaded region represents the middle 50 % of all books. In the 1960s the median book had a grade level of 8.0, with the middle half of all books coming in between 7.2 to 9.3. While 7.2 could be considered low fifty years ago, in 2014 there were 37 bestsellers, and 36 had a grade level of 7.2 or below. The floor for simplicity has become the ceiling. The number one bestseller with the highest reading level in 2014 was Daniel Silva’s The Heist, which had a score of 8.0. Out of all 37 books it was the one book that had a score that half a century ago would have been typical.

  On the upper end, James Michener’s 1988 novel Alaska had a grade-level score of 11.1, making it the number one bestseller since 1960 with the highest reading level. Twenty-five books since 1960 have had a grade level of 9 or higher. But just two of these were written after 2000.

  On the lower end, eight books tie for the lowest score of 4.4. All eight of these books were written after 2000, all by one of three high-volume writers: James Patterson, Janet Evanovich, and Nora Roberts.

  These ultrapopular bestsellers with low reading levels are a recent phenomenon. Twenty-eight of the number one bestsellers since 1960 that I collected had a grade-level score below 5. Just two of these were written before 2000.

  To see the trend before your eyes, below is a graph showing the percentage of books with a grade level greater than eight, the median score in the 1960s.

  And on the next page is a graph showing the percentage of books with a grade level less than six. This is the median today.

  The New York Times bestseller list holds a rarefied place in the book world. To have written a New York Times bestseller is to have made it. And for the general public, the list often serves as the public face of fiction, a guide to what’s worth reading. Yet in the last fifty years, there is no way around it: The books that we’re reading have become simpler and simpler.

  There are two reasons this could be happening. The first would be that all popular books today are filled with simpler sentences and more monosyllabic words. The alternative is that the New York Times bestseller list is getting “dumber”—as the Guardian would put it—because more books of a “dumb” genre are reaching the top. I’ll call this the “guilty pleasure” theory. If quick reads like thrillers and romance novels now make the list more often than they did thirty years ago, the median reading level would move down even if each genre’s grade level stayed the same.

  I’ve checked both theories, and the answer, it turns out, is: both.

  There have always been “guilty pleasure” books on the list. In the 1960s it was Valley of the Dolls, in the 1970s, The Exorcist, in the 1980s, the books of the Bourne Trilogy, and in the 1990s The Lost World of the Jurassic Park series.

  But without a doubt, there are more guilty pleasures on the list today than there used to be. In the 1960s a book would hold its top position on the list for many months at a time. Today, books jump up and down the chart much more rapidly. James Michener’s Hawaii and Allen Drury’s Advise and Consent were the only two books to reach number one bestseller status in 1960. In 2014 there were 37 that did, and the longest any one book claimed the top spot was four weeks (Grisham’s Gray Mountain). Prize-winning literary novels like The Corrections and The Goldfinch make the number one spot on occasion, but today it’s much more often dominated by commercial novels. This makes the contribution of the literary books less important to the median.

  Looking at prizewinners rather than bestseller lists, we find that literary books haven’t declined in reading level nearly as much. That being said, they are not as complicated in terms of sentence length and word length as you might think. Complicated themes don’t always translate to complicated writing. The average for Pulitzer Prize winners in the 1960s was a 7.6 grade-level score, and in the 2000s a 7.1. In the years in between, the average was 7.4. There are many more outliers among the Pulitzer winners (Chabon’s The Amazing Adventures of Kavalier & Clay scored a 10.0 while Alice Walker’s The Color Purple, from the perspective of fourteen-year-old Celie, scores a 4.4), but there has not been a systematic shift over the years.

  The growing presence of guilty pleasures is not the sole reason for the decline in bestseller reading levels, however. If we break down bestsellers by genre, we find that there has been a long-term shift within those guilty pleasures. Thrillers have become “dumber.” Romance has become “dumber.” There has been an across-the-board “dumbification” of popular fiction.

  In the following graph I have plotted the 25 authors with the most number one bestsellers since 1960. All of these writers have had at least seven number one hits in their career, and just about all of them are writing for a broad audience: suspense, mystery, romance, action, etc. They are shown by the average reading level of their books and the year in which their first number one bestseller was written.V

  Robert Ludlum is known for thrillers like the Bourne Trilogy, which debuted in 1980, but he still wrote at a Flesch-Kincaid reading level of 7.2, not common in popular fiction today. Tom Clancy and Dean Koontz, who both got their starts in the 1980s, write at a level higher than any of the rising popular writers of the last twenty years. Your average John le Carré novel had a reading level higher than 36 of the 37 number one bestsellers from 2014. Danielle Steel ranks as a low outlier for her time period, but she still writes at a higher grade level than many of her even more modern counterparts.

  There aren’t just more guilty pleasures representing popular books. The pleasures have gotten guiltier.

  Though it is the most prevalent, the Flesch-Kincaid test is just one of many tests of reading level. Most use sentence length as a large component. Today’s bestsellers have much shorter sentences than the bestsellers of the past, a drop of 17 words per sentence in the 1960s to 12 in the 2000s. This means any of these similar tests will show similar declines.

  One interesting alternative is the Dale-Chall readability formula. While it too uses sentence length, it has a separate component that factors in the number of “complex” words that appear in a text. In 1948 Edgar Dale and Jeanne Chall compiled a list of 763 words they did not consider complex. From this list it’s possible to count the number of “complex” and the number of “not complex” words in a text.VI The thought is that it’s not just sentence length that can make a book hard to follow for young readers, but the number of words that are unfamiliar.

  Since their original list, Dale and Chall have expanded the list to almost 3,000 words. Over 99% of the words in Dr. Seuss’s The Cat in the Hat are considered “not complex.” Seuss’s only two exceptions are thump(s) and plop.

  But 1 % complex is unheard-of when it comes to novels. The closest any number one bestseller has come to this is Danielle Steel’s 1993 Star, which uses a record-low 7% “complex” words. The book’s first sentence is below with its one “complex” word bolded.

  The birds were already calling to each other in the early morning stillness of the Alexander Valley as the sun rose slowly over the hills, stretching golden fingers into a sky that within moments was almost purple.

  On the other side we have Robert Ludlum’s 1984 thriller, The Aquitaine Progression. Twenty-two  percent of its words are considered “complex” by Dale and Chall, more than any other number one bestseller. Below are the first three sentences with words considered “complex” in bold.

  Geneva. City of sunlight and bright reflections. Of billowing white sails on the lake—sturdy, irregular buildings above, their rippling images on the water below.

  Just as we see Flesh-Kincaid scores decline in recent decades, so do the “complex word” counts measured by Dale-Chall. Though the results are much less pronounced than the change in Flesch-Kincaid reading levels, there is a clear downward trend since 1960.

  A bestseller once considered typical in complex word usage would now be on the high end of the spectrum. A 2% decrease is small in absolute terms but when you consider the small range a book may fall in, somew
here between 7 and 22%, a 2% fall is significant.

  *  *  *

  Where will the bestseller list be in ten or twenty years?

  The New York Times bestseller list is looked up to in the book world. It’s prestigious for authors and a guide for readers. By the changes the New York Times has made over the years, it’s clear that they think about its composition. And while their exact methods are undisclosed, they’ve acknowledged that unlike other lists they weigh the sales of certain independent stores more heavily than bigger retailers, tending to give a fighting chance to smaller, more “literary” books versus the commercial, grocery-aisle thrillers. But if the New York Times is conscious of the trend in its bestseller list, they face a question: Should they ever step in to exclude certain authors or genres from the bestseller list for fiction?

  It might sound outrageous at first for the Times to try to “shape” the list, but they have done it before. In 2000 a major change was made that excluded the Harry Potter books from the list. In the previous year, the number one spot was filled by a Harry Potter book on twenty separate weeks. The result was a new “Children’s Book” list, which has since splintered even more into distinct “Young Adult,” “Middle Grade,” “Picture,” and “Series” lists.

  One obvious fix to the dominance of guilty pleasures would be to split up the fiction list, the marquee list of the New York Times, into one focused on literature and one focused on genre fiction. If the New York Times wanted to promote a diverse range of books, they could make the former list their most touted. A safe haven could at least please those serious readers who want to know what’s popular in the world of books beyond pulp fiction. (Admittedly, the line between genre fiction and literary fiction would be difficult to draw, especially if publishers have a financial interest in gaining a certain categorization.)

  A similar effort was already attempted by the New York Times editors. Though it did not alter their main fiction list, in 2007 a “trade” paperback section was unveiled, which was supposed to shine a light on a certain brand of fiction. Here was the description used by the New York Times’s own editors upon its release: “This issue also introduces a new bestseller list, devoted to trade paperback fiction. It gives more emphasis to literary novels and short-story collections. . . .”

  The counterpart to the “trade” paperback was the “mass-market” paperback. A book qualifies for the “mass-market” bestseller list based not on its genre or potential audience, but if it’s printed within certain parameters (smaller pages, cheaper paper; often they are those pocket-sized editions you tend to see in grocery stores). And it just so happened that genre paperbacks tended to be printed as mass-market paperbacks. However, the market for these inexpensive books has since shrunk with the rise of ebooks, so more and more genre or commercial books are instead being published in the trade paperback format—that is, as higher quality paperbacks meant to be more lasting. As a result, the trade paperback list has not lived up to its initial selling point. As I wrote this chapter, the number one book on the trade paperback list throughout was the infamous erotica novel Fifty Shades of Grey. It was followed by Fifty Shades Darker and Fifty Shades Freed. All three have been on the list for more than 100 weeks. The rest of the list had some more literary fiction, but also works by authors Gillian Flynn, Nicholas Sparks, and James Patterson, who seem to defy the list’s goal of giving “more emphasis to literary novels.”

  If the New York Times wants to accomplish that, they’re going to again need to adjust their categories. Perhaps it’s time to bite the bullet and try to define that elusive “literary” genre rather than base their decisions on the book’s physical format. If they wish to hold on to their cultural perch, they’ll likely need to change again as they have done before.

  The broader question that I keep thinking back to, however, is: Should we worry at all about the overall reading level of the fiction bestseller list?

  For this question, I say no. I have devoted an entire section to showing how bestsellers have become . . . dumber. It would be easy for me to lump the New York Times reading level decline in with the rise of knee-jerk arguments that the country’s intellect is at an all-time low.

  But I don’t think this is fair. Remember, the reading level is supposed to be a rough cutoff for who is excluded from a text. You don’t have to be in sixth grade to read a book written at a sixth-grade level. Books with simpler texts can appeal to a wider audience.

  Simple can be great. It includes more people. Writing doesn’t need to be complicated to be considered either powerful or literary. The winner of the 2014 Pulitzer Prize for fiction, The Goldfinch, was also a number one bestseller and has a reasonable reading level of 7.2. While many classics do have high scores (The Age of Innocence at 10.4, Oliver Twist at 10.1, The Satanic Verses at 10.1), just as many have surprisingly low scores. To Kill a Mockingbird has a reading level of 5.9, The Sun Also Rises at 4.2, and The Grapes of Wrath measures down all the way at 4.1. All three of these books are revered by the literary community, but are accessible enough to be taught in high school classrooms across the country.

  This inclusion is needed to reach broad audiences. It’s logical that our most popular books are not complex, and I would not expect the future of popular reading to revert back to the lengthy sentences of George Washington’s first State of the Union Address. Kerouac’s most popular book, On the Road, scores at a reading level of 6.6 on the Flesch-Kincaid scale. And while I don’t think Kerouac was referring to sentence structure when he said it, I still think that the following line is worth considering in this discussion: “One day I will find the right words, and they will be simple.”

  * * *

  I. About 15 % of books from the earlier decades are missing from the sample because they are not available in digital form.

  II. Up until the creation of radio and television the State of the Union was often a written document sent to Congress.

  III. The 2010s only cover 2010–2014.

  IV. About 15 % of the books from the first few decades were not available in electronic form and were not included in this analysis. However, even if they all were written at an extremely low level it would not be enough to move the median of these decades below modern levels.

  V. The cutoff to make the chart above was seven bestsellers by the end of 2014. There is a possible skew in the data due to the fact that in order for a writer starting to write in the 2000s, he or she had to write much more quickly (and therefore possibly more simply) to achieve seven bestsellers by the end of 2014.

  VI. Conjugates of verbs are accounted for. Proper nouns are discounted.

  England and America are two countries divided by a common language.

  —GEORGE BERNARD SHAW

  A Tale of Wizards, Blokes, and Knickers

  For many young American readers Harry Potter introduced them to a new world. I’m not just referring to the magic that takes place at Hogwarts, but the wonderful world of British English. For every fictitious word like Muggle for American children to learn, there was a British one like bloke. For every magic spell like wingardium leviosa, there was a British curse like blimey. Left and right, characters were snogging by common-room fireplaces. It was not just the magic that drew American readers in, but the brilliant British language.

  And when we look at the data, we can actually see this: British words became so entangled in American readers’ memories of Harry Potter that it changed the way they saw the characters, compared to British readers.

  Let’s consider those three British B-words: bloke, blimey, and brilliant. It may be a simplification to say these are British words, but I would not be the first to do so. Bloke and blimey are both listed in the book A to Zed, A to Zee, which is in essence a British English to American English translation dictionary. Brilliant is listed in the book Divided by a Common Language as a word “Not to Say in the US” because it “can mean something quite different in the US.”

  You could take exce
ption to any of these words: Blimey is a cockney word that most in Britain would not use. Bloke is also used in other English-speaking countries such as Australia. Brilliant has a universal meaning, even if the exclamatory “Brilliant!” is not common elsewhere.

  But these are words that at the very least are not quintessentially American. If you heard someone describe another person as a “Poor bloke. Brilliant mind,” you’d guess those words were spoken by someone from the U.K., not the U.S. And those exact words happen to be spoken by Hagrid in the first Harry Potter book to describe Professor Quirrell.

  Are these words actually favored by British writers if we look at the numbers? Do all authors from the same side of the Atlantic have a shared style? Without looking at the data, it’s hard to tell how many of the phrases and words we associate with either American or British writing are real and which are exaggerated stereotypes.

  On a statistical level, the stereotype holds up when we turn to the data to look into these B-words. If you look at the British National Corpus (1980–1993) and the Corpus of Contemporary American English (1990–2015) you can see the disparity. These are both curated collections of hundreds of millions of words, meant to serve as benchmarks of the language as it’s put into practice on either side of the Atlantic. In these samples bloke is used 27 times more often in British writing than American, blimey 30 times more often, and brilliant 45 times more often.

  But I wanted to go a step further, looking at how these differences work their way into readers’ minds. Let’s start with bloke. Though used more in Britain, it’s still not a big part of common language. In the British National Corpus bloke showed up just 1.2 times for every 100,000 words. However, Americans used it 0.045 per 100,000, perhaps making it attention-grabbing on the rare occasion it is read or heard. In the seven Harry Potter books, J. K. Rowling uses this word even more often than the average Brit, at a rate of just under 3 per 100,000 words—and it stuck out to American readers.