As a literacy tutor, this is a hill I am prepared to die on.
Walk into any local library these days and ask the librarian how to help your young child learn to read and they will guide you to the collection of “beginner” or “early” readers. The brightly illustrated, “pocket sized” books have become ubiquitous and they come in two flavours: the controlled “high frequency” vocabulary variety and the “decodable” variety. The first belongs to the system of teaching reading known as “whole word” or “whole language” (or a bunch of other names) and the second is supposed to support the teaching of reading via a system known as “phonics.”
And, if you’ve spent even ten good minutes researching beginning reading you’ll recognise that these two flavours represent two very different schools of thought on the subject that are often very antagonistic towards each other. But here’s a bombshell: for a large portion of young learners both types will cause major issues if not used carefully.
Not only that, what if I tell you that for literal centuries there was no such thing as an early reader and people managed to learn to read much more fluently than many do today?
So what is the problem with early readers? In one modern phrase: cognitive load. Modern neuroscience has revealed what educators in the past knew in their gut. The mind is only able to attend to a very small number of things at once. The number varies from one person to another but it’s generally recognised to be in the single digits (around 5 give or take).
This is especially true when learning something new and even more true when learning a new skill. An added complication is that different people have different rates of “uptake” – one person will pick up a skill after only one or two practice sessions while another may take 1000. In most cases, both can be equally intelligent, they just are wired differently with different strengths. Change the skill and you might see the opposite happen.
In the context of reading instruction, it has now been shown by watching the brains of good readers that reading involves a process known as “orthographic mapping”. The brain is a very organised, and highly efficient computer. It doesn’t store individual words, rather it stores individual spelling “chunks” and it records “words” as links between those chunks.
This is a highly efficient way to store and retrieve what is potentially an almost infinite body of knowledge. They are getting somewhat rare, but at your next library visit, look for a printed copy of the complete Macquarie Dictionary. It’s one big book! Yet, the head of a competent reader has the system to decode the entire book stored in a relatively small space.
This system is not innate: unlike speaking, reading is not a natural skill – it’s one that everyone has to be taught (although some kids will virtually teach themselves). And for a very large portion of learners, the system must be built step-by-step.
When we break it down, reading has three components. The first is recognising the individual spelling “chunks”. The second is the ability to combine those chunks into a word which the reader can pronounce correctly. And finally, individual words need to be combined into meaningful phrases, sentences and eventually stories.
Now the problem with any beginner reader series is that they try to do all three at once. Some children will seem to learn the spelling chunks almost without any instruction at all and will develop the orthographic mapping with very little repetition. But those children are the exception rather than the rule in my observation.
The next batch of children will learn the spelling chunks fairly quickly and only need a little repetition to start the process of orthographic mapping. These children also will generally fare “well enough” with a beginner reader series.
However, at this point, we’ve only accounted for about half of the learning population. These are the kids that give many teachers and teacher aides the conviction that “most normal kids will teach themselves to read.” There’s a quietly spoken subtext in that assertion: “if they don’t, they have a learning disability.”
But the other half of the class is not necessarily learning disabled. They are probably not as “attuned” to linking letter groups to their associated sounds as the readers who “teach themselves” but most of them are fully capable of learning the skill if it is broken down into its components.
So the problem is usually not with the learners. It’s actually with the educators who are unaware of cognitive load theory. Now, I don’t blame the individual teachers: they are simply a product of their education and beginner reading instruction (and therefore teacher instruction) went awry over a century ago.
Now, in order to read text you actually need to process many different things at once. First, you need to recognise that the text is divided into words. Then you need to understand that each word is made up of letters. Deciding which letters combine to make a single sound adds another layer of complexity. And to really muddy the waters, we often need to decide which sound of several possibilities is represented by a letter (or group of letters) in a word (cough vs plough for example). All that needs to happen before we’ve read a single word, let alone attempted to extract any meaning from the text.
In competent readers all that happens pretty much subconsciously. This is a necessary precondition to being able to read accurately with good comprehension. In educational terms, we say that these skills need to be learned to automaticity. The key insight is that fast learners develop this automaticity very easily. The rest can still develop it, they just need more targeted teaching and practice to do so.
And this is where the common sense method, developed “on the shop floor” of the one-room school house, shines. You see, for literal millennia, all across Europe, reading was taught using the same basic approach. In fact, they didn’t really teach reading at all. Instead, they actually taught spelling first and reading just “happened” as a by-product. And they taught spelling in the only way that spelling instruction really makes sense in an alphabetic language: by breaking words up into syllables and spelling each syllable independently.
This fits beautifully with cognitive load theory because it breaks a big problem down into smaller chunks. Let’s look at how this works with an example word: “elephant”. The word has nine letters and eight sounds. This is way beyond the working memory capacity of most people. What that means is that by the time your child has sounded the last letter, they’ve probably forgotten the first.
But, if we break the word up into syllables it now looks like this:
el – eph – ant
Now we have three short syllables. A child that has been sufficiently drilled in decoding (reading) syllables will be able to decode each of those syllables “at sight” so then reading the word becomes almost as easy as reading a simple three letter word like “cat”.
Reading programs that are based around decodable readers (even those that follow a “good” systematic, synthetic phonics sequence) have one major weakness. They do not give sufficient practice at reading the basic syllable blocks before progressing to full words and connected sentences.
For a learner to become a competent reader, each step along the way (syllables, then single syllable words, then multi-syllable words, then sentences and finally fully connected stories) needs to be automated before proceeding to the next step. The fast learners can generally do this at the pace of a typical “decodable reader based” program. The rest simply can’t: they need a slower pace with more repetition built in. This is not a fault of the child: it is a fault in the program that does not have sufficient flexibility built in to cater for varying learning paces.
And the real danger of decodable readers for the slow readers is that, rather than giving the learner sufficient practice to automate the important skill of “decoding”, the story line and pictures provide him with a crutch: context clues that encourage “guessing.” Even if you do not encourage a child to guess, if they are being asked to read connected sentences when they have not automated the decoding for all the words in that text, they will resort to guessing based on context clues. It’s almost as sure as night follows day.
The old spelling book method of teaching reading was highly effective but the old spelling books need a teacher with a patient, kind manner who understands the best way to implement them. Unfortunately, like many old crafts: this is an almost lost skill.
What the world needs is a modern version of the spelling book that a parent with little prior knowledge can pick-up and present to their child one page at a time with gentle pacing instructions built into each lesson.
I have been using the spelling book method for years in my tutoring practice with some incredible results. And the same little boy that inspired Milestone Maths has inspired me to write a modern spelling book “first reader” for today’s children. Development is in early days but I hope to be in a position to announce a release date soon. Watch this space.
