Has the "Science of Reading" Gone Overboard on Phonics?

Some educators are spending too much time on phonics instruction in the early grades, experts warn.

---

The “science of reading” movement has yielded some much-needed reforms. For decades, schools of education have been providing prospective teachers with inaccurate information about how kids learn to read, and widely used reading curricula haven’t introduced phonics in the systematic way that many kids need. Although those problems aren’t entirely solved, there have been some dramatic improvements in the last few years.

But, as often seems to happen in the world of education, the attempt to correct one problem has given rise to another. Some educators, and some curricula, appear to be trying to teach kids everything there is to know about phonics—for example, according to cognitive scientist and reading expert Mark Seidenberg, all 44 phonemes in the English language and all 137 phonics rules, plus other rules about spelling, syllable types, and morpheme types.

Seidenberg and some other experts say that’s unrealistic and unnecessary. Children do need some explicit instruction in the correspondences between letters and sounds (in other words, phonics), they say, but they’ll pick up most of the knowledge they need about those correspondences implicitly, through reading.

Once kids have learned a certain amount about the patterns of English spelling, Seidenberg says, they’ll achieve “escape velocity.” That’s a concept borrowed from rocket science, but Seidenberg defines it, in the reading context, as “the ability to learn from reading with less and less supervision.”

Spending more time than necessary on phonics is a problem because children need other things beyond phonics to become proficient readers—especially the knowledge and vocabulary that will enable them to understand the texts they’ll be expected to read in the future.

When kids are still learning to read words, the best way for them to acquire that kind of language is through listening to read-alouds and having discussions in which they use the vocabulary they’ve heard. When teachers read a series of texts and lead discussions on a single topic, over a period of weeks, children are likely to retain the words in long-term memory. If all or most of the two-hour “reading block” is devoted to phonics, there’s no time in the schedule for this equally crucial kind of reading instruction.

This Alarm Was Sounded 40 Years Ago

While the critique about overteaching phonics has been cropping up lately, it isn’t new. The same alarm was sounded almost 40 years ago, in a 1985 report by reading experts called Becoming a Nation of Readers—long before the term “science of reading” came into vogue.

“The goal of phonics is not that children be able to state the ‘rules’ governing letter-sound relationships,” the report warned. “Rather, the purpose is to get across the alphabetic principle, the principle that there are systematic relationships between letters and sounds.” Even at the time, the authors worried that many reading programs were trying to teach too many letter-sound relationships and dragging phonics instruction out too long.

One of the authors of the report, Dr. Elfrieda H. Hiebert, is still trying to relay that message as the science-of-reading movement sweeps the country. She’s dismayed that, for example, one publisher promises that its curriculum will guide teachers in “explicitly teaching the 150 spellings for the 44 sounds of English.”

One possibly tricky question—as reading teacher and author Harriet Janetos has observed—is how teachers can determine when children have reached “escape velocity” and it’s time to back off on explicit instruction. But the 1985 report had an answer: “Phonics instruction should aim to teach only the most important and regular of letter-to-sound relationships, because this is the sort of instruction that will most directly lay bare the alphabetic principle.”

Hiebert says it makes sense to focus on vowel patterns like ee and ea, as in keep and eat, each of which appear in about 250 English words. Other patterns appear much more rarely and don’t need to be explicitly taught. The pattern ei-e, as in seize, for example, shows up in only six words.

While Hiebert doesn’t use Seidenberg’s “escape velocity” analogy, one way of framing her message is that explicit instruction in the most frequently encountered phonics patterns can enable beginning readers to achieve it.

The Problem With Current Decodable Readers

Hiebert also questions the current form of “decodable” readers, books that provide beginning readers with practice in the letter-sound patterns they’ve been taught. If they've learned the short "a" sound, children will get a book with sentences like "Pat sat on the mat."

Kids do need practice reading books with words that align to the patterns they’ve learned about, Hiebert says. But she believes that the decodable readers now in use don’t focus enough on repeating actual words as opposed to patterns. The premise is that if certain letter-sound correspondences have been taught, children will be able to decode any words following those patterns—including unusual ones like "jag," "vat," "nab," and "bam."

But Hiebert contends that expecting students to recognize a letter-sound correspondence just because it was presented in a lesson places too heavy a burden on their cognitive abilities. And if children don’t already have the words in a text in their oral vocabulary, that’s an additional challenge.

"It's hard to decode unfamiliar words," Hiebert says. "And texts have too many of them, even at lower grade levels."

The situation is only likely to get worse at higher grade levels, both in terms of students’ ability to read individual words and to understand text. The relationship between decoding and comprehension can be complex, but it’s clear they’re closely related.

Researchers have found, for example, that students who fall below a certain “decoding threshold” by the end of fifth grade are unlikely to grow in reading comprehension. A recent study done with a large sample of upper elementary and middle school students found that more than 20 percent fell below the threshold.

That doesn’t mean, Hiebert says, that millions of children “can’t read.” Rather, the evidence indicates that a significant majority of struggling readers at higher grade levels can read, in the sense that they can decode simple words. The problem is that they can’t decode easily enough to read with comprehension, especially when texts use longer and less familiar words.

While schools may have been overdoing decoding instruction in the early grades, there usually isn’t enough support for word-reading skills at higher grade levels, where intervention is likely to focus only on comprehension. When older students do get decoding-focused help, it’s generally aimed at simple, one-syllable words rather than the complex, multisyllablic words that are causing students trouble.

Focusing on Frequently Encountered Words

A better approach, Hiebert says, is to concentrate on words that are frequently encountered in text. Using digital tools, Hiebert has identified the 20,000 words that are most commonly used in school texts. Almost all of them can be sorted into about 5,500 families of words that share the same root—for example, the words improve, improvement, and unimproved.

That still sounds like a lot of words to teach, but Hiebert says teachers can leave out words that children acquire more or less naturally. That leaves about 2,500 word families to focus on, for both vocabulary and reading instruction. Hiebert’s “critical word zone” approach is based on that group of families, only about half of which appear in primary-grade texts. Several studies have shown that such an approach, even when used for only a brief time, can increase students’ scores on standardized comprehension tests.

Like Seidenberg, Hiebert notes that students learn to read as a result of their actual experiences with words rather than by memorizing phonics rules. If students read a lot, encountering lots of phonics patterns as they do, they become increasingly automatic in recognizing words with those patterns, along with their meanings. That gives them the cognitive bandwidth they need to learn to recognize irregular and rare words—and for comprehension.

For kids who struggle with reading, the problem is circular. They need to read a lot in order to become proficient readers, but their difficulties with reading prevent them from reading a lot. Hiebert’s solution is to increase their volume of reading by using books that repeatedly use the same round of words, from the most frequently encountered word families.

To do that, it’s best to use groups of nonfiction books organized by topic, Hiebert says. Nonfiction texts tend to use the same words repeatedly, whereas authors of fiction vary their word choices. Reading sets of books on the same topic can develop automaticity in word recognition and simultaneously build the kind of knowledge that fuels reading comprehension.

Teachers don’t need to create those groups of books themselves. Hiebert has founded an organization called TextProject that offers free text sets for struggling readers from the elementary grades through high school. A “Level A” set of books is organized around the topic of mud, for example. It includes four texts: What Is Mud?, Animals and Mud, Mud Baths, and Clay Art. A magazine aimed at teenagers includes an issue focusing on pets, with articles on “fur babies” and therapy animals.

Fluency, Silent Reading, and Text Sets

Many older students who struggle with reading lack “fluency”—the ability to read at an appropriate rate with appropriate expression. The usual response is to have students read a text aloud repeatedly.

There’s a correlation between oral reading fluency and silent reading comprehension, but Hiebert says there’s no evidence that improving oral fluency will produce better comprehension. Some evidence indicates that oral practice helps students read more quickly without boosting their comprehension.

One possible explanation is that when students read aloud, an adult or a device will prompt them to continue if they stop. But when they’re reading silently, they need to monitor their reading behavior themselves. In addition, oral reading tests generally use much shorter passages than tests that measure silent reading comprehension. Struggling readers may not have the stamina needed for those longer passages.

Rather than having students repeatedly read the same text aloud, Hiebert urges having them read silently from several texts that feature words from the most commonly used families. Studies of the approach, which has been used with young adults learning English as a second language, have found that it boosts reading comprehension as well as fluency, vocabulary, and knowledge.

None of this is to suggest that kids don’t need explicit phonics instruction in the early grades. Most do—just not as much as many schools are now providing. At the same time, many older students need more support in deciphering words—but probably not the same kind they would get in kindergarten or first grade. If we want students to achieve the escape velocity they need to become proficient readers, we need to add more nuance to the concept of the “science of reading.”