The Text and Academic Authors newsletter, The Academic Author, has abounded with reader interest in better writing. In an opinion column, Mark Hochhauser advocated more simplicity in writing, not "dumbing down" but writing more readably for students. Hochhauser said readability formulas indicate some textbooks are written above their intended student audiences. He advised text authors to use shorter sentences and smaller words.

Norma Roch, a free-lance text copy editor, responded that her editor colleagues don't discuss readability and asked, "Is it really as simple as using good grammar and short sentences?"

The fact is that readability formulas are a good starting point for considering better writing. However, the formulas go beyond shorter sentences, smaller words and good grammar.

The formulas were developed for newspaper writers and editors in the 1940s, when. World War II reduced newsprint availability and forced thinner papers with more tightly written articles (See Note 1). Robert P. Gunning developed the Fog Index that counted the number of words and the frequency of multisyllabic words. He told his newspaper clients: "Write as you talk ... Why should a police reporter say an accident victim suffered 'contusions and abrasions' when he really means 'cuts and bruises'?" (See Note 2)

Also in the 1940s, Rudolph Flesch developed a formula based on word length, sentence length and the human interest content of an article (See Note 3). One journalism professor said of Flesch's readability formula: "He and later scholars were good medicine for the media!s language ills" (See Note 4).

But the formulas are not without detractors. Researchers have tested the formulas for reliability and found that they may underestimate the reading difficulty level for math texts, poetry and other "terse and condensed prose," the kind often found in college textbooks, and "materials that contain subtle variations in meaning" (See Note 5).

Taking a closer look at the rationale on which readability formulas are based, without going into the multipliers and constants, consider these three:

1. Flesch Readability Formula: Longer sentences and longer words lead to reading difficulty. Take a sample of the writing and count the syllables, then average the number of words in a sentence. Combine the two measures, then convert the score to a range of grade levels.
   
   
2. Gunning Fog Index: Same approach, but less focus on syllables (See Note 6). Count the noncapitalized words of three or more syllables other than compound words or third syllable suffixes such as "ing" or "ed." Add the number of these "hard words" to the average number of words per sentence, multiply by a constant and convert to grade level.
   
   
3. Dale-Chall Formula: More highly regarded by education and reading researchers (See Note 7), this formula omits long words in favor of "unfamiliar" words. Unfamiliar are those not on a list of 3,000 common words. Get a count of the unfamiliar words and the average sentence length; sum the two scores, add a constant and look up the raw scores on a reading grade-level chart (See Note 8).

In a newspaper readability test using the three formulas, scholar Ron Smith found that Dale-Chall consistently scored prose at a significantly lower reading level than the Flesch and Gunning formulas. He noted that latter two are generally equivalent because both are based on length and syllable assumptions, while Dale-Chall uses unfamiliar words (See Note 9).

Another readability formula is called Cloze Procedure. Developed in 1953 by Wilson L. Taylor (See Note 10), this system randomly omits words from a passage and asks readers to fill in the blank with the correct word: Nannies watched the __________ at play. "Children" successfully reproduces the original text, based on our shared experience with the language.

Taylor noted that long words such as "respectability" are more easily understood than some short words such as "erg," and argued that short sentences aren't necessarily easier to read; ease depends on context.

The four formulas noted are not the only ones developed. Others are based on content of certain parts of speech, active voice, "concrete" terms, and the use of personal nouns and pronouns. And while readability formula proponents quibble about minute aspects of their choice formula, they seek validation for it by comparing it against others. The usual outcome is that each of the procedures seems to do a pretty good job of assessing readability. However, a recent study by Katherine McAdams (See Note 11) used the Fog Index to discriminate between two sets of newspaper articles: one at about the eighth grade level and another at about the 15th grade level. This study had a startling outcome: level of difficulty had no effect whatsoever on reader satisfaction and readers cared only about the article's topic interest level.

What might textbook authors take from this little jaunt through the maze of readability formulas?

First, Hochhauser's counsel to write more concisely definitely has merit. Short sentences and small words do form the basis of most of the formulas. A textbook writer would be well advised to heed his admonition to write with brevity. In fact, the surest way to lose any reader is to write a complex, 50-word sentence replete with multisyllabic words.

The concepts that textbook authors try to explain are difficult enough even if we strive to make our sentences short and concentrate on using simple words. The material we're presenting is probably the first time a student has been exposed to these ideas. Successful text authors recognize the difficulty and work to keep their presentation clear. Second is the "contusion-abrasion" advice. Textbooks offer enough complexity in the principles they attempt to present. Many technical terms are required. Still, the readability formulas remind us to use common words whenever possible in structuring sentences. We can define "hypothesis" as "a formal statement stipulating predicted relationships among variables designed to investigate causality." We can also define it as "a formal statement to test if one variable affects another."

Next, the McAdams study cited should be a wake-up call for textbook authors. McAdams said readability takes a back seat to interest: Readers prefer an interesting topic with very difficult prose to an easy but boring topic. Dare we consider that many in our textbook audience aren't there by choice? If we concede that our subjects may not always be scintillating reading, we should stretch to counteract the interest factor with the best writing we can muster.

Finally, the remaining readability formulas remind us about the tried-and-true tactics of writing for our readers: active voice, "concrete" rather than vague terms, personal nouns and pronouns. These are among the elements of good writing in any genre, and they should certainly be employed by the teaching writer.

How much of better writing is explained by the readability formula precepts? Hochhauser was right: They constitute the basic principles, perhaps 60 percent of writing quality. But Norma Roch was right too. These are only the basic guidelines. The other 40 percent of better writing can't be reduced to formulas.