The latest installment in an ongoing series by Matthew Davis, Director of the Core Knowledge Reading Program. See Matt’s other articles from past issues of “Common Knowledge” here.

Version 1.0 of the Core Knowledge Reading program for kindergarten is now being piloted in over forty classrooms around the U.S. The program combines systematic phonics instruction with frequent reading aloud to build vocabulary, background knowledge, and cultural literacy. We believe this is going to be a powerful one-two punch that leads to increased reading achievement across the board while also narrowing achievement gaps between demographic groups.
You can read a brief description of the kindergarten materials and the pilot initiative here.

If you would like to see some sample pages from the actual components that are being used in the field, click here. The sample pages include lessons from the Skills Strand and read-alouds from the Listening and Learning Strand, as well as general introductions to both strands. These introductions will tell you a great deal about the sequence of instruction, as well as the instructional and philosophical ideas behind the program.

When Two Vowels Go Walking

By Matthew Davis

One of the most famous of all phonics rules is “When two vowels go walking, the first one does the talking.” This rule has been taught in many phonics programs and in thousands of elementary classrooms over the years. There is a song version of this rule, which the clever people at PBS and Between the Lions have made into a cute animated video. The song presents several examples of the rule: “In boat,” the singer explains, “you hear the ‘o’ but not the ‘a’. In meat you hear the ‘e’; the ‘a’ sits quietly. The second vowel you see but you don’t say.” You can watch the whole video here. There are also several musical interpretations of the song on YouTube. Here’s a bouncy version of the tune by Korean students. For an acoustic guitar version by a young strummer, click here. To hear the walking-talking rule mixed in with other traditional phonics lore and set to rap music, click here.

These videos showcase musical talent, creativity, cleverness, and youthful energy. But do they represent good reading instruction? Before you get out your video camera and try to produce a better video treatment, let us take a moment to consider how reliable the walking-talking rule really is.

As part of our work on the Core Knowledge Reading Program, we have developed a database that includes 25,000 frequently-used English words, a breakdown of the phonemes in each word, and tags for each of the spellings, or graphemes, used to represent those phonemes. Among many other things, this database allows us to find and tally all the words for which the walking-talking rule guides a student to choose the right pronunciation. The database tells us that the rule points the reader to the correct pronunciation for 2,634 words. (Click here to see a list.)

That seems pretty powerful—until you count up the cases in which the rule leads the reader astray. It turns out that there are 2,592 words in the database that a reader would pronounce incorrectly by following this rule. (Click here to see a list.)

Alas, it turns out that this much-taught rule works only about 50.4% of the time. In other words, it is not much more reliable than a coin flip.*

The database can tell us more. For example, it tells us that, if a student followed this rule slavishly while attempting to read natural text, on average, he or she would make about 59 reading mistakes in a run of 1,000 words — that is to say, 59 mistakes that are directly attributable to bad advice provided by this rule, above and beyond whatever other mistakes he or she might make. These mistakes by themselves are enough to put the poor youngster right on the edge of the “frustration level” as it is commonly defined.

The unreliability just noted is reason enough for teachers to drop this rule from their teaching repertoires, but there are other reasons as well. The rule also exemplifies and helps to perpetuate a widespread misunderstanding about what vowels are. The walking-talking rule applies the word vowel to the letters. It says, if two vowel letters go “walking,” then the first one “talks” while the second one keeps mum. But vowels are not letters — they are sounds. Specifically, vowels are sounds made with an open mouth and more or less unobstructed airflow. Vowel sounds are distinguished from consonant sounds, which involve a closing of the mouth or some blockage of air flow, whether momentary or sustained. It is only by extension and a kind of transfer of meaning that one can go from speaking about vowel sounds to speaking of letters like ‘a’, ‘e’, ‘i’, ‘o’, and ‘u’ as “vowels.” Really, we should speak of these as the letters that stand for the vowel sounds. That is a bit clumsy, but at least it is accurate.

Unfortunately, however, that is not the way things have traditionally been taught in this country. Many of us were taught that the vowels in English are ‘a’, ‘e’, ‘i’, ‘o’, and ‘u’ —and sometimes ‘y’ and ‘w’.” That is the formula you will find in the phonics rap linked above. It is also the formula I was taught (without the rapping) when I was young. I knew it by heart, and yet I remember being very confused by the part about ‘w’. I never understood how ‘w’ could be a vowel. I could not think of a single word that had the letter ‘w’ standing for a vowel. I was not able to think of one in elementary school or middle school. I remained stumped throughout high school, four years of college, and eight years of graduate school. It was only a few years ago, when I was in my thirties, that I figured out what was meant. The letter ‘w’ is sometimes used in tandem with other letters to stand for a vowel sound, as in the word snow (an example where the walking-talking rule works) or now (an example where the rule does not work).

The letter ‘y’ is classified as a “vowel” in such schemes chiefly for the same reason, because it is part of various letter teams that stand for vowel sounds in words like key and boy, although ‘y’ can also stand for a vowel sound unaided, as it does in myth or funny. So that is what is meant by “sometimes ‘y’ and ‘w’.” But you could argue that this addition does not go far enough. Probably we should also add ‘g’ and ‘h’, which are members of vowel-team spellings in words like sight and night, bought and sought, weigh and neigh. We should probably add ‘r’, too, since most linguists view /er/ as in her as an indivisible vowel sound, in which the /r/ is thoroughly amalgamated with the vowel sound. So maybe the line up of English “vowels” should be “a, e, i, o, u, and sometimes, y, w, g, h, and r.”

But there is a better solution at hand, if only we are willing to stop thinking of vowels as letters and go back to the original definition of the word vowel as a particular kind of sound. Once we make this shift and begin thinking about vowels as sounds rather than letters, we can see that the walking-talking rule really makes no sense. When two vowels go walking, they BOTH do the talking, because two vowels means two sounds. In fact, English vowels rarely go walking without a consonant in between. One does find two vowel sounds adjacent in some words, like Maria (/ee/ walking with /u/ or schwa), poet (/oe/ walking with /e/) and create (/ee/ walking with /ae/). But words like these, with back-to-back vowel sounds, make up only a tiny percentage of English words.

If we return to the example words boat and meat with our sound-based definition of a vowel, we can see that neither of these words actually has two vowels. Both words are monosyllabic, containing only a single vowel sound. In each case, the single vowel sound is written with two letters.

As for the letters, the truth is that none of them talk. If letters could talk, reading would be much easier. All we would have to do is listen to the page.

Thus, the case against the walking-talking rule is not only that it is a very unreliable guide to pronunciation; the rule also sows confusion by obscuring the true meaning of the word vowel and perpetuating a misleading identification of vowels with letters rather than sounds. To these two charges I will add a third: the rule ignores, and encourages students to ignore, a fundamental fact about English spelling. Written English is loaded up with digraphs, spelling units in which two letters stand for a single sound. There are dozens of two-letter spellings that stand for single sounds. Among the vowel spellings we have ‘ai’, ‘ay’, ‘aw’, ‘au’, ‘ey’, ‘ei’, ‘ee’, ‘ea’, ‘ie’, ‘oe’, ‘oa’, ‘ou’, ‘ow’, ‘oi’, ‘oy’, ‘ue’, and ‘uy’. Among the consonants we have ‘ch’, ‘sh’, ‘th’, ‘wh’, ‘ng’, and ‘ph’. Learning to recognize these digraphs and connect them with single sounds is one of the challenges involved in learning to read English.

It is a challenge that must be squarely faced if the student is to succeed, and yet the walking-talking rule does not encourage students to face the challenge. On the contrary, it promotes what can only be called an escapist fantasy. The fantasy is that one can succeed in decoding by focusing on the first letter of a digraph and ignoring the second letter. The reality is that the grapheme consists of two letters that work together to specify the sound. Both letters matter, and ignoring the second one is a recipe for making about 59 mistakes in every 1,000 words read. The walking-talking rule encourages students to treat as twos things that must be recognized as single units if the reader is to succeed. Thus, the rule may be said to promote a maladaptive strategy which has the potential to do harm even beyond the cases in which it is supposed to be applied. Suppose, for example, that a student tried to read the following words by letting the first letter in the underlined digraph “do the talking” while ignoring the second: thanks, ship, chop, ping, phone. Such a student would misread every single word. The walking-talking rule seems to express a longing for a writing system in which there are no digraphs, a writing system in which all sounds are represented by single letters. Unfortunately, that is not the writing system we have.

I have been hard on this phonics rule, but I hope I will not be labeled a phonics basher. I consider myself a strong phonics advocate. I believe that systematic instruction in the grapheme-phoneme correspondences is the best way to teach reading, and that beginning readers should not be asked to read stories loaded up with graphemes they have not been taught.

In the heyday of whole language, hecklers denounced Marilyn Jager Adams and other scholars as “phonicators.” Well, I consider myself a phonicator. Indeed, I say Let phonication thrive! But let us not phonicate promiscuously. We phonics advocates need to really understand our writing system and its spelling patterns and not just repeat jingles passed down from generation to generation. We need to be careful not to teach rules that are unreliable and misleading, like the walking-talking rule. We need to be clear about what a vowel really is — and is not — and we need to anchor our teaching in the phonemes of the language. The rules issue is particularly important since “the rules don’t work” has long been one of the most frequently-used arguments in the rhetoric of whole language. As long as phonics instruction features misleading, ineffective rules like the walking-talking rule, it will be open to criticism from advocates of whole language and whole-word reading. Fortunately, there is no reason why it has to be.