Guessing words from context is not reading, by Fay Tran

Tags: guessing, whole language, reading, Fay Tran

By Fay Tran

Before the mid 1980’s, reading was taught by a combination of phonics and whole word methods, with structured practice provided by controlled vocabulary readers and work books. This was before the age of computers and interactive white boards, so teachers used flashcards and the blackboard to teach decoding skills and build fluency and comprehension. For reading lessons the class was usually divided into three groups, so that children were taught at their individual level. The teacher also aimed to hear every child read from their graded reader every day.

And then Whole Language arrived, like a new religion that everyone had to follow.  Kenneth Goodman was the god to be worshipped and Frank Smith, Brian Cambourne, Donald Graves and Mem Fox were the disciples to be obeyed. From 1985, all student teachers were taught the Whole Language method exclusively for teaching literacy, and all education lecturers were expected to be committed to the method.

Initially called ‘The Psycholinguistic Approach’, Whole Language is based on the belief that children can learn literacy skills the same way that they learn language, through plenty of exposure to, and interaction with text.  Teachers are expected to work as facilitators rather than instructors and direct instruction of phonics and decoding skills is banned, as are controlled vocabulary readers. Children are encouraged to guess words from context, rather than use phonics skills to decode them, correction of errors is minimal and predictive text readers with supportive illustrations are used to make the guessing easier. 

At the time I was fortunate to be working as a learning support teacher in a school that bucked the trend and continued to teach phonics for both reading and spelling.  Knowing that children really needed direct instruction of phonics to develop reading skills, I thought reason would soon prevail and there would be a general return to the proven method, but here we are, more than 20 years later and the return to sanity has only just begun. For all of that time about half of my work involved the support of children at risk of reading or spelling failure because of learning difficulties.  The other half was rescuing children who came from other schools with already established reading failure.

Without exception, regardless of underlying learning difficulties, these children responded to direct instruction and practice in phonics skills and spelling rules, starting right from the basic level.  Reminding a child to sound out a new word rather than guess from context is a very simple way of making a huge difference to the development of their reading and spelling. While it was very rewarding to put these children back on track and know that their futures would no longer be limited by illiteracy, I became increasingly concerned that what was happening at our school was quite unusual and that thousands of children across the country were being deprived of the opportunity to develop literacy skills.

I know I was not alone in these concerns. I attended almost every SPELD and AREA or LDA conference in Melbourne, Adelaide and Brisbane and every one reinforced my beliefs that the way I was teaching was the most effective and efficient way to ensure every child developed essential literacy skills, but somehow outside these organizations, a few schools and some private tutors, the rest of the world stayed in denial.

The problem is that children who don’t develop efficient literacy skills at school are not just disadvantaged but are seriously traumatized and often damaged for life, and this starts very early on. Let’s not beat about the bush here. Functional literacy is absolutely essential in today’s society and without it, people are at serious risk of a dismal future. If you don’t believe me, ask a social worker. Even the child who falls behind his friends by the end of the first year develops negative attitudes to his or her learning ability and by the third year any child who is struggling to read is likely to have written themselves off academically. These children develop either avoidance strategies where they work so slowly that they never complete tasks, or behave in a disruptive manner so that the teacher focuses on their behaviour rather than their learning problems. Some manage to conceal their problems by copying friends’ work and learning their take-home readers by heart.  

Amazingly many of the children I taught with significant reading and spelling problems came with reports showing at least average performance with no hint of difficulties. But their mothers knew. Other parents told me that their children’s teachers had told them not to worry as the problems were transitory and would eventually go away. Nothing could be further from the truth, as learning difficulties do not just go away, even those that are just the result of inadequate teaching. Children with learning difficulties need good teaching, involving direct and systematic instruction in the basic skills.

A few years ago, a grade two teacher and I had an interview with a parent of a child that had come to the school at the beginning of the year with significant learning and behaviour problems.  We were able to report that the child was on track with his reading and spelling skills and his classroom behaviour had also improved. The mother was in tears with relief and after she left, the teacher and I walked to the staffroom feeling very pleased with ourselves.  We were discussing the reaction of the mother when another staff member asked what would have happened to the child if he had not come to this school. That brought us down to earth because we had to admit that the child would be likely to experience continued failure and struggle for the rest of his school career and beyond, unless of course, he was rescued by someone else along the way.

This situation is unfathomable in our wealthy, educated and technologically advanced country. I simply do not understand why every primary school cannot teach every child to read and spell to a functional level in the seven years of attendance. We knew how to teach literacy skills back when I first trained fifty years ago and now we have the evidence from neuroscience to validate the direct instruction of phonics skills and spelling patterns that worked then and still work today, given the chance.

The initial campaign to promote Whole Language as the only way to teach reading and spelling was so powerful that almost no one dared to challenge it, lest they were ridiculed as old fashioned and closed to new ideas. Now, after 20 years of rescuing children who were failed by the method, I feel it is time to be brave and openly discuss why the method did not work.

The main reason is that guessing words from context is not reading. Furthermore, training children to guess from context after noticing just the first letter of a word is not teaching them to read at all.  In fact it actually prevents children from developing the skills they need to become accurate and fluent readers. Even if they are introduced to phonics incidentally, they are taught to use this knowledge only as a last resort, if guessing from context does not help. This is why so many children that I have had to teach over the years had no underlying learning difficulty and responded quickly to instruction and practice in the basic reading skills.

When a child reads a word by sounding it out, he or she can be pretty sure that the decision is correct. Some words are not easy to determine even when sounded out, but most of these are the high frequency words that are quickly learnt in the first year of school and I have no quarrel with children using context to help when sounding out a words like ‘pour’ or ‘once’. It is fine to use context to facilitate one’s decoding skills. However if the word is guessed from the first letter only, then there is no way of knowing if it is actually correct or not. ‘D-o-g’ can’t be anything else than ‘dog’ regardless of the context, but reading ‘dog’ for ‘dogs’ can change the whole meaning of the sentence without the reader being aware of the error.

Mandy is a year 2 student that I am helping with reading and spelling. The other day she argued with me about the word ‘tent’ in her Fitzroy reader.  She is quite capable of sounding the word out but she said it had to be ‘trunk’ or ‘stump’ because there was no tent in the picture and there was an old tree. Even when I insisted that she sound the word out and she conceded that it did sound like ‘tent’, she wasn’t happy and blamed the book for confusing her. If not checked at every word, Mandy will confidently and fluently read a sentence with more than half the words and the entire meaning incorrect. It is very hard to replace faulty strategies with more effective ones, especially when you only have the children for one hour a week.

Having unknowingly read one word incorrectly, the child is likely to make further mistakes to maintain the meaning of the sentence.

Wally is a boy in year 3 who came to me at the beginning of the year barely reading at all, simply because he had not been taught decoding skills by any of his three previous class teachers. He is now reading at about year 2 level and his skills are developing.  The other day, he was reading a sentence which started with ‘They could’ but read ‘The colour’. This illustrates how one mistake leads to another, which happens all the time with guessers. By just glancing at the first letter or two of the first word, he read ‘The’ for ‘They’ and this lead to him reading ‘could’ as ‘colour’ even though he was able to read both ‘They’ and ‘could’ as soon as I asked him to have another look at each word.

Another problem that I see involves memory and the learning process. The aim of reading practice is to develop the accuracy and fluency needed to enable good comprehension. This involves using decoding skills with increasing speed and efficiency, but also involves the quick recall of known words and parts of words from memory. Every time a word is sounded out or synthesised from its syllables or phonograms, it is actively processed by the brain and thus likely to be stored in memory for future recall. This cannot happen when a word is guessed, because there is nothing to remember except perhaps the first letter and the sentence it occurred in. So no useful learning occurs.

We know that focussed attention is vital for learning to take place, but if the child’s attention is focussed on predicting what the next word might be and not on the structure of each word, his or her decoding skills and memory for words are not advanced by the experience and again, no useful learning occurs.

It is good that the high frequency words, like ‘was’ and ‘where’ which are often difficult to sound out, are being directly taught in most, if not all, beginners classes. All the Prep/Kindergarten children I know bring home lists of these magic or golden words to practise. The danger is that if this whole word reading is the only strategy apart from guessing that is taught, some children rely on it too much. They appear to thrive at first but lose momentum and then fail around year three when the memory load becomes too great. Often it is the bright child with a good visual memory that has this problem.

When masses of little books replaced the use of graded readers I was not overly concerned. I knew that children with learning difficulties needed controlled vocabulary books to provide them with the systematic practice of sight words and phonics skills and continued to use them for my students. I did not realize the harm that the books could do, if they used predictive text, as they mostly do now, which train the children to use the title of the book and the pictures to guess the words.

Lucy is a little girl in year one with attention and memory difficulties. She learnt almost nothing in her first year at school, but is making at least normal progress working with me twice a week. She is now able to read the first two sets of the Fitzroy readers and can read and write just about any phonetically regular one syllable word. Occasionally she asks me to listen to her read her school take-home reader, which she reads fluently without more than a glance at each page. But this is not reading! Lucy thinks she is reading and no doubt her teacher does too, but I know that most of the words in that book could not be read in isolation, whereas any word she reads in the Fitzroy readers can be read anywhere it occurs.

I have no quarrel with short books with uncontrolled or natural language for children to read as long as they have the decoding skills to read them. With the little books now available from Dandelion Readers, Little Learners Love Literacy and SPELD-SA, children can start reading real books even when they only know a few letter sounds.

Graded decodable books provide systematic practice of developing skills and can also provide practice of the high frequency words like ‘said’ and ‘they’ which must be recognised automatically. The problem with the predictive text books now in proliferation is that they encourage the dreaded guessing and fool both the child and his or her teacher that he or she is actually reading, when the child is just guessing from the pictures and context. This may explain why so many children manage to hide their difficulties until the NAPLAN test in year 3.

A further issue I have with the Whole Language method is the emphasis, right from the beginning, on fluency rather than accuracy. In my experience, fluency comes naturally when word recognition and decoding skills are automatic. Children can develop fluency by reading very easy books or by repeated reading of the occasional page of a more challenging book. To encourage fluency, teachers have been taught not to interrupt a child’s reading to correct errors, but to wait until the end of the sentence or paragraph and then discuss one or two, but not more, of the errors made.

In contrast, I find that a very important aspect of effective teaching of reading is to provide instant feedback, which enables a child to quickly correct his errors as they occur.  I do this by pointing above each word, stopping when an error is made. The feedback provides a second chance for the child to decode or recall the word from memory while the word is still in the short term memory stage. Learning takes place because the child has recalled or decoded the word correctly and he or she is then likely to remember this experience for future reading. If the feedback is not provided until later, or not at all, then incorrect information is likely to be stored in memory, which once again, interferes with learning.

Whole Language reading is supposed to put the emphasis on reading for meaning, but it is probable that reading comprehension is compromised by the concentration on guessing from context at the word level. Reading with 100% accuracy, even if it is slow, ensures better comprehension than reading riddled with mistakes. Even one mistake, like reading ‘can’ for ‘can’t’ can change the meaning of a sentence without any clue from the context that a mistake has been made. One or two mistakes in a sentence compounded by more in following sentences can result in the meaning the child gleans from the text being quite different from the meaning intended by the author.  So learning by reading is also compromised.  On the other hand, when the child is reading by automatic phonics and word recognition skills, his or her brain is freed to consider the meaning of the text as a separate task.

I was heartened in 2009 by the Julia Gillard’s announcement, when she was the minister for education, that the teaching of phonics for reading would be included in the national curriculum. Great! I thought. That will help so many children. I have also been pleased to hear of schools introducing phonics programs in the early years for spelling programs. However, when it comes to using phonics skills for reading, it does seem that the necessary revolution is not yet happening. There are a few schools introducing phonics based programs, but most are still teaching children to use guessing from context as a first strategy and this is reinforced by the predictive text readers that all children take home every day.

Maybe now is the time to say ‘enough is enough’. Enough children have suffered from literacy failure; enough teachers have suffered the frustration of not being able to teach in a way that ensures success; enough parents have suffered the trauma of watching their children disintegrate rather than thrive at school.

The Whole Language method of teaching reading does not work because of faulty methodology.  Teaching the strategy of guessing words from context, along with little or no correction of errors and the use of predictive text reading books have combined to make it more difficult for all children to learn to read and extremely difficult  or impossible for those with learning difficulties. It is now time to acknowledge that as fact and move on to the task of re-educating teachers and saving our children from literacy failure.