Saturday 26 August 2017

Who’s in your reading instruction family tree?

Many people spend vast tracts of time trawling through family photos, birth, death, and marriage certificates and online repositories of church records and the like to compile an understanding of where they came from, culturally and geographically.  From time-to-time family-tree research even turns up some bragging rights – paradoxically in Australia, that often comes in the form of finding a convict in the higher branches of the family tree.

However, I wonder how many in education (whether as classroom teachers or as university academics) give thought to the ancestry of their beliefs and practices concerning early reading instruction? I often hear (or read, e.g. on Twitter) “I am not a Whole Language advocate”, or "I don’t use Whole Language approaches”, yet the practices these same people go on to advocate do in fact, have their origins in Whole Language.

So let’s have a look at some common philosophical positions, beliefs, and practices, and shake the family tree to see where they come from. In some cases, we can trace lineage to a particular epistemological camp, and in others, we find there is mixed lineage – as in our own family trees.

I may have missed some ideas, and am happy to add them in. You may also think I have muddled the lineage of some ideas or approaches, in which case I am happy to hear from you and be directed to sources that re-calibrate my understanding.

I will contrast “Whole Language”- derived approaches here with approaches derived from cognitive science, as no phonics advocates argue for a single focus on one aspect of the language system over and above the others. 

Lineage and comment
Learning to read is natural – just like learning to speak and understand oral language.
Whole Language
This is such a lovely, but incorrect idea. Reading and writing are derived from spoken language, but they are not a simple representation of spoken language in a different modality. Written language tends to be more formal, has conversational dysfluencies and pauses edited out, and historically, has not occurred in “real time” between two parties. (That has changed in recent times, with the advent of email, texting, etc). 
Spoken language is a faculty humans have developed over millions of years of evolution, such that the human brain devotes significant amounts of its real estate to producing and understanding language.
Reading and writing, on the other hand, have only existed for 5-6000 years – a mere blink in evolutionary terms. This means that the human brain has had to “re-purpose” language pathways for reading, and it requires skilled instruction for optimal development. Interested readers are referred to the work of Professor Stanislaus Dehaene on this subject. 
It should also be remembered that much in all as speaking and understanding may be “biologically natural” children do receive an enormous amount of specific 1:1 input from adults to develop oral language skills in the pre-school years. Oral language doesn’t magically pop up out of nowhere; it develops within the interpersonal space between children and their carers. 
Unfortunately, however, children have quite variable experiences of oral language in the pre-school years, such that some arrive at school with richly developed phonological/phonemic awareness, vocabularies, narrative language skills, and so on, and others are more impoverished. This means that early teaching needs to accelerate the progress of those who start from behind. It's not enough for these children to be making progress at the same rate as their linguistically more able peers. 
It is these same children who start from behind, who often remain behind, and then make up the “long tail of under-achievement” in reading outcomes.   
Oral language skills are fundamentally important to the acquisition of reading.
Here we find some mixed lineage in our family tree, and a good thing that is too, but it creates that awkward moment at family gatherings of look-alike cousins who may not be as similar as they initially appear.
It would be odd if Whole Language advocates such as Goodman, Smith, Cambourne et al. did not emphasise the importance of oral language for learning to read, as it is axiomatic to their views on where reading skills are derived from.
Strangely, however, advocates of cognitive science in early reading instruction are sometimes falsely accused of promoting “phonics only” approaches (I have never actually heard such calls but it is claimed by some Whole Language advocates, without any evidence).

However, a proper look at the arguments from cognitive science shows that they are embedded in a wider emphasis on the so-called “Big Five” of phonemic awareness, phonics, vocabulary development, comprehension, and fluency. This is nowhere better represented than in the Five from Five Project which provides abundant resources for teachers and parents. Readers should also check out this paper by Australian education academic, A/Prof Deslea Konza, in which she advocates for a widening of this notion to The Big Six, explicitly including oracy in the framework. 
As I have outlined previously, what separates phonics out from these other language elements is that it is contentious and in many cases, poorly understood.

The bottom line is that what occurs within words is part of the language system – phoneme-grapheme links and morphological affixes are part of the meaning system that we call language. Whole Language cannot be whole unless it takes account of what occurs within words, not just what happens between them. 
Every child learns differently
Whole Language
This is one of those simplistic truisms that got into the education water and is now difficult to remove. If there’s 7 billion people on this planet, there are not 7 billion different learning needs. Teaching would be impossible if that was the case. Yes, all children are individuals and need to be cherished and respected as such, but human brains, like hearts, livers, kidneys, gall bladders, skeletons etc have more in common with each other between individuals, than they do differences.

Of course, some differences do occur,   and of course these differences do sometimes matter – especially in the context of developmental disabilities. But in the main, we should assume similarities in needs not differences. For teachers, this translates into pattern-recognition, without which teaching would not become easier over time and experience would count for little at all.
Literacy develops from whole to part
Whole Language
This view reflects the “top down” approach to early reading instruction espoused by Whole Language advocates such as Kenneth Goodman, who also asserted (2014, p. 85) that “There is no hierarchy of sub-skills, and no necessary universal sequence”.  
This is the kind of thinking that sees a Blunderbuss approach taken to early reading instruction – children are “sprayed” with instructional bullets such as sight-words lists to rote-learn, predictable texts, and encouragement to guess. They are left to infer the alphabetic principle (if they are lucky), and are not afforded the developmental scaffolding of simplified sentence structure, vocabulary, or word structures.
Beginning readers should be encouraged to rote learn lists of sight-words
Whole Language
Because English has only a semi-transparent orthography (i.e. links between sounds and letters are sometimes clear and sometimes not), Whole Language advocates deal with the issue of high-frequency “irregular” words by presenting them to children early in the learning-to-read process as a massive visual memory task. Such words might be written on flash-cards and children simply have to memorise them as wholes. Never mind that most, if not all have some regular features, and never mind that some children have had very little text-exposure in the pre-school years, so just the abstract notion that these black squiggles on the page represent spoken words can be a mystery in itself.
Over time of course, we want nearly every word to become a sight word – a word that readers cannot not read, even if they choose not to – because the level of automaticity for the brain is so high, that reading is not a matter of conscious choice.
Children should be encouraged to invent spellings, and to experiment with punctuation
Whole Language
Delightful in all as some of the products of this process may be, it is an inefficient means of learning, as it provides too many opportunities to learn error patterns. Teaching children correct spelling and punctuation, and giving them opportunities to practice these results in improved automaticity and frees up cognitive capacity for the next level of complexity. Practising errors is confusing to young children and is an inefficient way to learn.
It’s better for children to discover/intuit the alphabetic principle for themselves, through learning to read and write. Phonics (if it must be taught) is best dealt with in the context of authentic, meaningful texts
Whole Language
These are central tenets of Whole Language-based reading instruction and if they dominate your beliefs and practices then you can see where the ancestry of your ideas lies. Check out Kenneth Goodman’s ideas here.
Three language systems interact in reading: the grapho-phonic, the syntactic, and the semantic, so children should use the Three Cueing System to deal with unfamiliar words during reading.
Whole Language
This is a popular and widely promoted approach  (sometimes referred to as “Searchlights” in the UK), though it is atheoretical and had a mysterious entrance into the educational arena, as outlined by Alison Clarke. 

Alison observes that:

“Strong readers and spellers internalise and automatise the links between words’ sounds and their spellings, and eventually can convert speech to print and print to speech at lightning speed without conscious effort. It’s only weak readers who have to guess from pictures, context, syntax or anything else. Context, syntax etc. come into play after a word is identified, in comprehending the text.”

Multi-cueing (or Three Cueing) is also strongly linked to the Whole-Language based idea that beginning readers should be encouraged to look at pictures to “derive meaning from text”.

The logical inconsistency of course, is that if children are looking at pictures in order to work out what the words are, then they are looking at and naming pictures, which is a different skill from reading.
Comprehension of meaning is always the goal of readers.
Whole Language
This is a bit of a motherhood statement that is intended to sweep away more nuanced consideration of the task at hand for the beginning reader. As summarised by Hoover and Gough’s (1990) Simple View of Reading, the beginning reader’s success is a product (not a sum) of their decoding ability and their language comprehension skills. 
So yes, the end-game is comprehension but we need to get to comprehension via decoding. This becomes only too apparent in the middle primary years, when picture cues and predictability diminish, and some previously apparently “good” readers are devoid of decoding skills. 
An early emphasis on systematic synthetic phonics equates to
·        Drill and Kill
·        Barking at Print
·        A neoliberal conspiracy to destroy children’s love of reading and literature. 
Whole Language (obviously!)
It is better to systematically teach children how to de-code at the outset, while carefully introducing sight words, constantly developing vocabulary skills, and strengthening comprehension skills.
Cognitive science (obviously!)
A “Balanced Approach” to early literacy is what works: making sure that phonics is “in the mix”, but starting with predictable texts, sight words, and encouraging children to guess from context.

Whole Language
I’ve blogged about the problems associated with this “instructional bricolage” before. Balanced Literacy, as the diagram below illustrates, is a “bit of this, and a bit of that”. It is atheoretical and open to an infinite number of ways of being interpreted by different teachers. It does not position explicit phonics teaching as the starting point in early years classrooms, and is basically a re-badging of Whole Language instruction.

Letter of the Week
Someone will need to help me out here as I am not 100% sure of the lineage of LOTW. It does not intrinsically look like it belongs in either camp. Maybe it’s reading instruction’s cuckoo’s egg?
Either way, there is a good critique of this approach here
Identification of word families – onset and rime
Cognitive science
According to Hempenstall (2015, p. 16) “The onset of a syllable is its initial consonant(s), and the rime is its vowel and any subsequent consonants in the syllable". 
Hence in the words “tap” and “trap”, the onsets are “t” and “tr” respectively, and they share the rime “ap”. The aim of this approach is to strengthen syllable and phonemic awareness and to teach decoding by analogy (e.g., knowing that “mug” and hug” belong in a “word family” should help a child to decode “jug” by analogy. This approach seems to have had some popularity in recent years, however its usefulness over an emphasis on phoneme-grapheme relationships for beginning readers is questioned by eminent reading researchers, such as Professors Maggie Snowling, Charles Hulme, and Kate Nation of Oxford University (see Hempenstall, 2015).
Children should read authentic texts from the outset
Whole Language
This is a central tenet of Whole Language instruction and is one of those nice ideas that has inherent face appeal, while lacking empirical substance. It goes something like this:  

There are so many beautiful children’s story books out there, so surely if we use those to teach children to read, we will instil a love of reading and hey presto, will produce a generation of passionate readers.

If only it were so.
Given the influence of Whole Language instruction in our schools over the last three-four decades, I think we can be fairly certain that this logic does not apply. I have blogged before about the authentic illusion in early literacy instruction.
We should not conflate the books that adults read to children (which should be rich and varied in their content: vocabulary, syntactic structures, and narrative complexity) and books that we provide to beginning readers to get them off the blocks. These should be simple and decodable, to allow graduated consolidation of sound-letter correspondences, while also introducing high-frequency sight words that are not readily decodable (e.g. “could”). There are many options for decodable texts and some are listed here
All that’s really needed to improve literacy skills is to better engage parents in the process of reading to their children and instilling a love of reading.
Whole Language
This is a view promulgated by some children’s authors (e.g. Mem Fox). Reading to children is important at so many levels – as a soothing, engaging, entertaining, horizon-expanding activity that parents and children can enjoy together. It is not enough, in itself, however, to get children across the bridge to literacy.
Teachers should provide a rich language and literacy environment for students and should emphasise speaking, listening, reading, and writing.
Mixed lineage.
This is a point of furious agreement between the Whole Language and cognitive science camps, as reflected in the Five from Five Project mentioned above.

It's good to know that when the extended family gets together from time to time, there are some safe topics on which we can all agree. 

(C) Pamela Snow (2017)