I recently received two packets from England containing copies of the same document each with an appended note to the effect: ‘What do you think of this?’ The document was a ‘working paper’ on the NLS, intended to introduce a dialogue on phonics teaching, and to set out the current position of its authors regarding phonics and the Literacy Hour. As my work was cited in this document to support certain curriculum choices and other arguments, I am using this forum to respond to the invitation for dialogue and to answer the question my friends posed.
I am concerned about two issues discussed in this document. The first has to do with the overlap of real books activities with phonics instruction during the Literacy Hour in the Foundation Stage and Key Stage 1. The second has to do with the sequence of phonics instruction presented on page 8 of the document.
To clarify my position on these two issues, I want to provide some facts.
1. Reading , spelling, and writing involve two basic skills: knowledge of the written code (reading mechanics) and being able to comprehend language (listening comprehension). You can’t comprehend what you read if you can’t decode and/or if you don’t comprehend what people say. Across the school years, reading comprehension is almost perfectly predicted by a combination of two tests: a test of decoding skills and a test of listening comprehension (combined correlation is around .90). This body of research is so consistent that the statistical values from one study to the next are almost identical. This means there is nothing ‘special’ about reading comprehension. That is, it isn’t necessary to teach comprehension by having children read. You can teach it just as well by having a conversation.
2. Children learn early on that books convey meaning through printed words. If they never learn that print is ‘a code’ for sounds in our speech, or if they receive contradictory or conflicting instruction, most children prefer to adopt a ‘sight word’ (whole word) strategy. This seems ‘natural,’ it is easy to do initially, and has some immediate success, that is, until visual memory starts to overload. A ‘sight word’ strategy takes one of two forms: visual memory alone, or decoding the first letter phonetically, plus guessing the word based on its length and patterns of ascenders (b, d, f, h, k, l, t) and descenders (g, h, p, q, y). These strategies have been observed and documented many times in the clinic.
In my research on reading strategies in the schools (McGuinness, 1997), I discovered that nearly all children used a whole-word guessing strategy at the start of Year 1 (age six). By the end of the school year, one-third of the children remained primarily ‘whole word guessers.’ When the children were followed up two years later, the ‘whole word guessers’ had not changed their spots. They still made many of the same errors on the same words on the same reading test they had made two years earlier, and they were far and away the worst readers in the class. Surprisingly, many of these children scored in the superior range on a vocabulary test. This tells us that becoming a whole-word (sight-word) reader is not due to low verbal skills, but is a high risk factor in the general population, and something teachers should curtail at all costs.
3. There are three large-scale, observational studies in the literature on how children spend their time in the classroom. Using large numbers of classrooms, each child was individually monitored for extensive periods of time. When ‘time-on-task’ (what each child actually did and how long he did it) was correlated to various measures of reading, only two activities contributed to reading skill across the early years: a) Phonics activities, specifically learning sound-to-letter correspondences, and segmenting and blending practice. b) Writing of all types: tracing, copying, writing from memory, words, sentences, stories.
Several ‘reading activities’ were unrelated to reading (correlations at zero), such as memorizing letter names, learning about concepts of print, segmenting syllables, etc. More importantly, some activities were negatively correlated to reading skill. This was true for time spent memorizing ‘sight words,’ and most particularly for teacher led ‘language arts’ activities, such as lessons on vocabulary and grammar, and story time. The ‘language arts’ activities were negatively correlated to reading comprehension at very high values (r = -.70 to -.80), meaning there was an almost perfect linear relationship between them. In other words, the more time children spend in teacher-led ‘language arts’ activities, the worse reading comprehension scores become.
There are two possible causes for this result. One is that it reflects a simple trade-off: the more time you spend on language activities, the less time there is available to learn the code. The second cause is more alarming. It is possible that practices in these teacher-led activities, such as ‘reading along’ or doing ‘vocabulary’ work in the context of looking at ‘real books,’ actively undermines phonics instruction. Whichever one is true, this is bad news for the Literacy Hour, certainly during the period when children are trying to master the code.
The evidence from the highly successful phonics programmes (Jolly Phonics, Phono-Graphix, Lindamood, Lippincott) is clear and consistent: Teach the alphabet code and get on with it. And when you’ve done this (a process taking about 11-20 weeks), move on to other reading skills, such as fluency practice, adding spelling alternatives, and reading comprehension.
Taken together, the evidence does not support the position taken by the NLS group:
“Shared, guided and independent reading and writing are also key features of the methodology covered by the literacy hour. -- Children should be taught to apply and coordinate a range of reading strategies in a connected way such that each reinforces the other and all contribute to the development of fast early fluency and comprehension.” (pg. 15)
The Reading Reform Foundation made a similar argument against this view, stating that introducing multiple strategies at an early stage of reading instruction will be “mutually contradictory and will confuse rather than assist young readers.” The NLS group does not agree, but they miss an important point. No one who supports a phonics first approach to mastering the code is claiming that ‘phonics instruction’ can magically suffice for all remaining forms of literacy instruction. The point is simply that if you are teaching the code, don’t open the door to encouraging inappropriate decoding strategies.
The NLS proponents, in a further rebuttal to the RRF commentary, miss a second and equally important point:
“This exclusive approach [meaning phonics] precludes teaching any kind of hypothesizing, problem-solving, predicting or inferring, which are pejoratively dismissed as ‘guessing’ on the grounds that they interfere with the proper business of reading.” (pg. 15)
As we have seen, reading comprehension, which is the real goal of learning to read, is completely linked to decoding and to listening comprehension. There is no reason whatsoever, why ‘hypothesizing, problem-solving, predicting, or inferring’ can’t take place during natural communication – i.e. speaking and listening. The teacher can read a story and ask the children for information concerning what they think the story is about (hypothesis); how they would get out of trouble if they were the characters in the story (problem-solving), what they think the outcome might be if the characters persist in their plan of action (predicting), and speculate on how they got into this mess in the first place (inferring). There is nothing unique about reading comprehension.
The concerns of the NLS group do make sense IF they are based on erroneous assumptions, such as: it takes three years to learn the code,’or ‘language arts goes out the window if phonics lessons dominate.’ But these are false fears, and they melt into insignificance in light of what is really true. They are also dangerous fears, because they are holding us back.
The second issue, the alphabet code itself, is of particular concern, especially as my suggestions for teaching a Basic Code (40+ phonemes and their most common spelling) are outlined on page 7 (middle paragraph). There are other problems that come in here, such as their rationale for teaching the first 6 phonemes, but I will focus on the main problem. It appears that the NLS group does not really understand synthetic phonics or the nature of a Basic Code. They provide brief summaries, and list programmes based on these principles, stating that they “work on similar lines” to what they have designed, yet this is far from the case. On page 8 there is a chart of a sequence of ‘steps’ with lists of phonemes on the left, and lists of graphemes (letters, digraphs, phonograms) to be taught on the right. There is no connection whatsoever between these two lists.
The NLS paper states that good phonics teaching involves teaching “the most common spelling” for each phoneme, yet there are only 11 phonemes on their list. The remaining 30+ phonemes are nowhere to be seen. On the right side of the page, there are seven lists containing a total of 62 graphemes. Anyone who understands how the code works, will immediately see a shift in logic. The code is now almost entirely letter-driven (visual), having nothing to do with the stated goal: “teach how graphemes map on to 45 phonemes.” Not only this, but the order of these graphemes across steps 3 – 7 has no connection to the structure of the spelling code.
At step two, the child hears six phonemes and sees the six common spellings for those phonemes. But at step three, everything changes. Suddenly, the young reader is confronted with a batch of unrelated letters and digraphs which are not linked to phonemes in any way. (These are ss, ck, l, n, d, k, sh, ch.) Notice that four higher levels of complexity of the spelling code are introduced as well. Here are: 1) Letter(s) standing for phonemes that have never been taught (l, n, d, sh, ch). 2) Consonant digraphs (ss, ck, sh, ch). 3) Alternative spellings for the sound /k/ (ck, and c taught at step 2). 4) A spelling alternative controlled by phoneme position in a word (ck). There is only one way for a child to handle this, and that is to fall back on visual memory, using the logic that ‘letters make sounds.’ The alphabet principle goes out the window at this point, and there’s no way to get back to it.
There is a reason why programmes like Jolly Phonics and Phono-Graphix are fast and effective, and this is because they stick with the logic that the NLS document initially describes, but then abandons.