The science of readability measurement has its roots in early 20th-century educational research, driven by the practical need to match reading materials to students' abilities. Before quantitative formulas existed, teachers relied on subjective judgment to assess text difficulty — an approach that was inconsistent and unscalable.

Rudolf Flesch, an Austrian-American readability expert and writing consultant, transformed this field with his 1948 paper introducing the Flesch Reading Ease formula. Flesch's insight was that two measurable features — average sentence length (words per sentence) and average word length (syllables per word) — correlated strongly with reading difficulty as measured by comprehension tests. His formula, 206.835 - 1.015(ASL) - 84.6(ASW), produces scores from 0 to 100, where higher scores indicate easier text. Flesch's 1946 book "The Art of Readable Writing" influenced generations of writers and editors, and his work directly shaped the Plain English movement that eventually led to U.S. government mandates for readable public documents.

J. Peter Kincaid, working for the U.S. Navy in the 1970s, adapted Flesch's formula into the Flesch-Kincaid Grade Level formula, which converts readability into an American school grade level. This made the metric more intuitive: a score of 8.0 means an eighth-grade student should be able to understand the text. The U.S. Department of Defense adopted the Flesch-Kincaid Grade Level as a standard for military technical manuals, requiring that they score at or below the reading level of their intended audience.

Robert Gunning developed the Fog Index in 1952 specifically for business writing. His formula counts "complex words" (three or more syllables, excluding common suffixes and compound words) as a proxy for vocabulary difficulty. The Fog Index roughly corresponds to years of education needed to understand the text on first reading. Gunning's consulting work with newspapers and corporations helped popularize the idea that clear writing is good business practice.

The SMOG (Simple Measure of Gobbledygook) Index, developed by G. Harry McLaughlin in 1969, uses only the count of polysyllabic words in a sample of 30 sentences, making it simpler to calculate by hand. McLaughlin argued that SMOG was more valid than other formulas because it correlated better with comprehensive comprehension tests (100% comprehension) rather than the 50-75% comprehension threshold used by other formulas.

All these formulas share fundamental limitations that are important to understand. They measure surface features — word and sentence length — not actual comprehension factors like concept density, prior knowledge requirements, text coherence, or logical structure. The sentence "Ox ax ox" scores as extremely easy (short words, short sentence) despite being meaningless, while "The mitochondria generates energy for cellular metabolism" scores as difficult due to polysyllabic scientific terms that any biology student would know. Domain-specific vocabulary, cultural references, implicit assumptions, and logical complexity all affect real readability but are invisible to formula-based metrics. Modern natural language processing research uses machine learning models trained on comprehension test data to produce more nuanced readability estimates, but classical formulas remain widely used due to their simplicity, transparency, and established benchmarks.