Introduction to the Premises of Reading Level Advice
Writers use reader demographics to guide their choice of language and syntax. A writer composing a dissertation on nuclear physics does not have the same reader demographics as an author writing for children. Word choice, sentence structure, inclusion of pictures or equations are all guided by the intended audience. Authors of novels, specifically genre novels, have a greater flexibility in prose decisions than do academics publishing scientific research. Google results for the query “what reading level should you write novels,” suggest authors write at a reading level no higher than seventh grade which, in the United States, equates to the proficiency of a twelve- or thirteen-year-old. A slew of results advocate writing to the third-grade level, or that of a typical eight-year-old. The premises supporting these assertions fall into three general categories:
1) Hemingway wrote at a third-grade level
2) The general (American) public reads at the seventh-grade level and authors should write to that level, or lower
3) Authors who write books using simple language make more money
While text editors identify spelling or grammar errors, lexical analysis tools evaluate prose complexity and assign a reading level. The intended use of lexical analysis is to provide authors guidance in crafting prose suitable for a specific audience. The secondary use is to help writers tune the relative difficulty of their prose to ensure they are writing at a given grade level. In and of themselves, these tools are not bad or good. Likewise, a writer’s choice to use lexical analysis tools carries no moral qualifier. It is the arguments for minimizing text complexity that are bad. This paper examines the logical fallacies of the reading level advice, explores the usage of a lexical analysis tool, and argues against the advice to write as if no one can read.
An Examination of the Logical Fallacies
To assert the premises of reading level advice are deficient is not a condemnation of lexical analysis tools. There is no good or bad technology. There are bad arguments, including those supporting the trend to write at a specific reading level. Lexical analysis tools amplify the effects of this faulty logic, making the ability to “level down” language accessible and normalizing the writer’s adherence to illogic. The following section identifies the logical errors of each premise of the argument to write at a reading level no greater than that of a thirteen-year-old.
Premise 1: Hemingway wrote at a third-grade level
Premise 1 is a logical fallacy of composition and division.
An example of the composition fallacy is demonstrated in this statement: Because the words in a book are inexpensive to print, the book must be very cheap. This composes (or equates) the price of the words to the final price of the book itself; not only is this statement false, but it is also incomplete because it does not account for the cost of manufacturing a physical book, nor does it account for the payments to the publisher, book distributor and author. By collapsing Hemingway’s skill as a writer onto the mechanical skill of a reader to recognize his words, Premise 1 results in a false composition of excellence of craft and reading level.
In addition, Premise 1 commits the fallacy of division. An example of the division fallacy is the inverse of the composition example: Because a book costs a lot, each word must cost a lot to print. This statement is false for the same reasons the composition example is false.
Premise 1 divides the parts of prose (words, sentences) from the sum of those parts (storytelling). This results in the conclusion that the sum of comprehension is the mechanical ability to recognize syntactical elements (words, sentences). A corollary false conclusion is that good storytelling is solely determined by the writing level of the language used to tell the story. Neither of these conclusions is valid.
Premise 2: The general (American) public reads at the seventh-grade level and authors should write to that level, or lower
The second premise combines three logical fallacies: middle ground; tu quoque and special pleading.
The middle ground fallacy occurs when nonsensical answers are produced by arguments based on artificially constraining data to a mathematical mean. A simple example is: “I say the sky is blue, you say the sky is red, so we can compromise and say it is purple.” The “middle ground” of claiming the sky is purple is silly. Restating Premise 2 to match the blue/red/purple analogy results in a similarly silly compromise, to wit: “If an elementary student says the prose is too hard, and a college student says the prose is too simple, the author should write prose at the golden mean of the seventh-grade level.”
Statements that return criticism with criticism, such as “the public reads at a seventh-grade level, therefore authors should write at a seventh-grade level” fall into the tu quoque fallacy, better known as the “two wrongs don’t make a right” fallacy. Premise 2 makes the following critical statements: first, the general population’s reading proficiency is low and second, authors should write novels that do not exceed the average reading level. The shockingly low reading proficiency in America is a condemnation of the public education system, which has done wrong by the students. To conclude that authors should limit the complexity of their prose implies readers are neither capable nor desirous of improving their reading proficiency. A semiliterate populace and books written to enforce this condition are two wrongs that, when combined, erode development of critical thinking and empathy born of struggling to understand a complex novel.
These errors are compounded with the special pleading fallacy. Special pleading is typified by ignoring facts that make an argument weaker. Consider a person stopped for speeding. The driver says he should not get a ticket because he was rushing to the hospital. The driver is making a special plea by sidestepping the fact that speed limits don’t come with exceptions based on personal circumstance. The only situation in which this is not special pleading is if the vehicle is an ambulance transporting a critically ill patient. The premise that authors should write at a seventh-grade level is the equivalent of insisting an ambulance driver convince the police officer treat him as a minivan driver. Authors, like ambulance drivers, can bend the rules. If readers are the police, they’re going to wonder why so many ambulances are moving out of the way of speeding minivans. The second premise is a special plea that authors accommodate readers with skills within the mathematical mean. It is also a special plea that readers accept a library of books suited to the needs of the average reader. Worse, Premise 2 is an arrogant assumption that the mean cohort cannot or does not want “level up” and read at more than a seventh-grade level.
Premise 3: Authors who write books using simple language make more money
The third premise encompasses at least three logical fallacies: ignoratio elechi; correlation without causation; and retrospective determinism.
The fallacy of ignoratio elechi occurs when a relevant conclusion is reached but misses the point. For example, if someone says pandas must not be dangerous because they look cute, they come to a relevant conclusion (pandas are cute) but miss the point that cuteness of an animal does not equate to the danger posed by an animal. To state that authors who write at a third- to seventh- grade reading level make more money than authors who write at levels above that range is ignoratio elechi. While it may be true that any single book written at a fifth-grade reading level outsells any single book written at the post-graduate level, generalizing the conclusion to all books misses the point. The complexity of the language does not equate to a specific income. Given an identical royalty rate per book sold, the sale of a one copy of a post-graduate book will make an author more than the single sale of a middle-grade book; anyone who has purchased college textbooks can vouch for that. On the flip side, readers pay less for middle-grade books but the market for those books is much larger. It is worth noting that any sale, regardless of reading level, makes more money than no sales. If an author chooses not to publish a book because it is “too difficult to read,” that author is undercutting any financial gains while simultaneously paying obeisance to the tyranny that a story is worth only the amount the author reports on her taxes.
The statement equating reading level of a novel with the income of the author is a correlation without (necessarily) any causation: a simple novel written by a popular personality will make more sales than a complex novel by an unknown person, but this does not mean that if the popular personality wrote a complex novel it would not sell equally well, nor does it ensure that a simple novel written by an unknown person will sell at all.
The third fallacy of Premise 2 is that of retrospective determinism. The very nature of book sales accounting is retrospective (the total is not known until a sale is made). Likewise, attributing a determining factor to a sale is done retrospectively. The sale could be due to the simplicity of the language; it could be due to a price drop; it could be due to a passing fad; it could be due to the popularity of the person who authored it. It could be due to anything or a combination of everything – which is why saying the reason for the sales is driven by the reading level of the book is retrospective determinism, and a narrow determinism at that.
A Case for The Logical Use of Lexical Analysis Tools
Technical manuals, laws, and medical treatments should be written in the simplest language possible because these are not fiction, they are instruction. These documents need to be more like Hemingway’s news reports and less like his short stories. There is no illogic to using a lexical analysis tool to suit the purpose of technical writing. Moreover, a reader desiring a deeper understanding can advance to more complex texts on the subject: for example, a reader can advance from the reduction of a law written in layperson terms to reading the law itself. The corollary, and it is a dangerous one – is that there is no avenue for this progression in fiction or, rather, there is: CliffsNotes. In a future shaped by an adherence to writing to the average reading level, most fiction will become the equivalent of CliffsNotes that do not correspond to a more “complex” text.
Brief Overview of One Lexical Analysis Tool
The lexical analysis tool, Hemingway Editor, claims to “make your writing concise and correct.” Hemingway Editor outputs a “reading level score,” along with color-coded highlights indicating sentences or words the AI deems too complex. Paid subscribers can configure the “accessibility” level of the tool, finetuning it to the reading level of the intended audience. The default on the free web version, sadly, is not set at “most accessible” (the reading level of a third grader). Users must pay for the Pro edition to be advised on how to write for seven-year-olds. It is not stated on the UI, but it appears the free edition is intended to help writers produce prose for ninth graders (13/14 years old). This is the Hemingway Editor output for the previous paragraph:
Refutation of Premise 3: Hemingway Editor Readability Score
POOR Grade Post-graduate (22 -? years old)
8 of 12 sentences are very hard to read.
Please note the use of the name “Hemingway” without the descriptor “Editor” refers to the actual Hemingway, not an AI with a clever marketer who uses Ernest Hemingway’s likeness in a way that would enrage him more than the bulls in Spain.
The Inequivalence of Reading Levels and Storytelling
Hemingway’s language is simple, with sentences pared down to the minimum complexity and a vocabulary equally minimalistic. Hemingway’s background is as a news reporter in a time when newspapers were made of paper. The physical format limited news stories to a physical space measured in inches. To transmit relevant facts in two or three inches of newsprint requires cutting language to the bone. The minimalism of Hemingway’s literary voice is informed by his background, resulting in a low lexical analysis score. According to calculations, The Master of the English Short Story writes at a third grade reading level.
Paragraph 1: Hemingway Editor Readability Score
OK Grade 11 (16/17 years old)
1 of 6 sentences are very hard to read.
Lexical analysis tools conflate the analysis of parts (sentence structure and words) for the sum of the parts (evoked emotion and the revelation of a deeper understanding of human nature). Yes, a seven-year-old can read the words in a technical sense – or so I hope. They can understand the parts. Comprehension is the sum of the parts. Comprehension level is what creates the magic of ‘seeing’ a book differently when read at various stages in an individual’s cognitive and emotional development. A discussion about Hemingway’s short story, Hills Like White Elephants, with an elementary schooler, a teenager, and a forty-year-old will be qualitatively different due to the reader’s comprehension level. While each person can recognize the words and parse the sentences, the meaning and impact of those apparently “easy” words result in anything but “easy” existential considerations. To say Hemingway writes at a third-grade level demeans the power of his prose. Fiction isn’t transmission of fact: it is the transmission of emotion and ethical perspective. Good fiction submerges a reader. Great fiction drowns the reader, forcing them to confront their conceptual framework of what it is to be human.
Paragraph 2: Hemingway Editor Readability Score
OK Grade 11 (16/17 years old)
4 of 11 sentences are very hard to read.
On average, Americans read at a seventh-grade level, or the equivalent of an eleven- or twelve-year-old. A twelve-year-old recognizes a greater variety of words than a third grader. Not to harp (but actually, to do just that), words and sentence structure are not the story in and of itself. Hills Like Elephants is far more than the literal words comprising Hills Like White Elephants. Lexical analysis software cannot differentiate between writing like Hemingway and storytelling like Hemingway, whose shortest story is: “For sale. Baby shoes. Never worn.” A reader with rudimentary skill can parse the words (and there are no sentences). An experienced reader comprehends and feels the emotional impact of the words. No mathematical analysis can spit out a number representing the gut punch of those sentence fragments. Quantitative measurements of literacy exclude the qualitative purpose of fiction.
Paragraph 3: Hemingway Editor Readability Score
OK Grade 10 (14/15 years old)
1 of 11 sentences very hard to read.
The Misuse of The Reading Level Rubric
The age-equals-reading-level rubric is a tool developed for teachers to evaluate a student’s mechanical proficiency in the reading skill. It is not a rubric to “grade” the conceptual, emotional, or ethical comprehension of adult fiction readers. Lexical analysis software perverts the original intention of the reading level concept: instead of measuring the student’s mechanical proficiency, these tools measure the author’s proficiency at diminishing the richness of their vocabulary and dethroning the majesty of a well-placed semicolon. Readers, and authors, would be better served by exchanging the reading level rubric for a methodology more like the film industry’s qualitative movie rating system. Movies are categorized by the maturity of the themes and imagery, not by the complexity of dialog or closed captions. If we must categorize fiction, let’s at least do so in a way that embraces the soul of fiction instead of the presumed near illiteracy of the populace.
Paragraph 4: Hemingway Editor Readability Score
POOR Grade 16 (22/23 years old)
3 of 6 sentences very hard to read.
Remarkable as Hemingway is, there is more than one way to craft gut-punch fiction that existentially drowns the reader. If the majority of authors use reading levels to guide their prose, the majority of books will read like Hemingway’s “Baby Shoes,” but without the substance of “Baby Shoes.” The identity principle in mathematics states that X=X. In the same way, only Hemingway is Hemingway. Math would be trivial to the point of uselessness if there were only one variable, not to mention one equation (X=X). Fiction is trivialized when the writer is handcuffed to the “easiest” word (one variable) and simple sentences (one equation). Current authorial “guidelines” encourage homogenization of new fiction, replacing the multiplicity of voice with a cesspool of faux-Hemingway that utterly fails to be Hemingway.
Paragraph 5: Hemingway Editor Readability Score
OK Grade 13 (18/19 years old)
2 of 6 sentences are very hard to read.
More damning than the impending boredom of homogeneity is the insult to the reader. The implication of writing to a third-grade level is that the author believes the reader has not and can never attain the intellectual curiosity to encounter “difficult” fiction. It is condescending. It is also a self-fulfilling prophecy that trains readers to avoid dictionaries, despise the beauty of compound sentences, and avoid the classics. I, for one, reject this dismal subjection of the reader to the continually devolving beauty of the English language in an effort for authors to make more money – especially given that the arguments for doing are erroneous.
Paragraph 6: Hemingway Editor Readability Score
OK Grade 13 (17/18 years old)
3 of 5 sentences are very hard to read.
Do “easy” books sell better than those that are “difficult?” The answer depends on the author’s skill, the population size of the reader niche, and a variety of factors that are largely unknown or unquantifiable. It is not surprising that the advice to write “easy” books comes during a historical era where the majority of the American public is stressed, living paycheck to paycheck, scarred by a pandemic, and consumed with international news foretelling the end of civilization. Escapism is essential to psychological survival. It is a public good that there are books offering a reprieve. The dark side is that if a book is truly “easy” (and not Hemingway “easy”), escapism lasts only as long as it takes to read the book. The so-called “difficult” books (and the one true Hemingway) have an impact that goes far beyond the last page. Books that challenge a reader provide psychological escapism, yes, but they also provide grist for the intellect, increased empathy for others, and the satisfaction of achievement.
Paragraph 7: Hemingway Editor Readability Score
OK Grade 12 (17/18 years old)
3 of 8 sentences are very hard to read.
Conclusion
Lexical analysis tools are not inherently “bad,” whereas the reasons to apply these tools in writing fiction are illogical. Conclusions built upon faulty premises are themselves faulty. Given the proliferation of logical fallacies in the three categories of popular advice supporting the use of lexical analysis tools, authors who do not consider the consequences are constraining their creativity based on literal nonsense. The act of writing to an average reading level disowns the powerful complexity inherent to the English language while simultaneously demeaning the reader’s ability to rise above the average American’s proficiency in comprehension. As an author, I reject the tyranny of these faulty arguments. I refuse to treat readers as if they are incapable of putting in the work to stretch their literary limits. I intend to drown my readers, and trust that they are stronger swimmers than Google thinks they are.
Paragraph 8: Hemingway Editor Readability Score
POOR Grade 14 (19/20 years old)
3 of 7 sentences are very hard to read.
Note: A great overview of logical fallacies is available at https://www.logicalfallacies.org/ .
Erzsebet Carmean 
Poppy read and enjoyed your blog. He said to tell you great job.
Get Outlook for iOShttps://aka.ms/o0ukef
LikeLike
Thank you, Poppy!
LikeLike