Guessing: why it is not an effective ‘reading strategy’

Guessing

The problem with implementing good, systematic phonics teaching in schools isn’t that is difficult to get right, it really isn’t. The problem is that it’s also easy to get wrong, especially if training has been inadequate. One particular problem area I’d like to consider here is the deliberate or accidental encouragement of guessing from context which, once established as a habit, can cause significant problems and be very difficult to shake.

How does it start?

Often it is deliberately encouraged, doing so was considered ‘best practice’ in most UK schools until very recently and is still advocated by many. At other times it is inadvertently encouraged due to poor choice of reading practice material, misguided feedback to the child, and unhelpful messages/cues to speed up.

Decodable versus ‘Real’ or Levelled/Banded Books

As well as sharing stories and books to develop their language skills; to maximise success and build on good classroom teaching, developing readers need decodable books so they can practice applying their new-found knowledge. A decodable book is one which limits its vocabulary to words that can be decoded by the child given what they have been taught so far (i.e. they increase in complexity as the child progresses). Unfortunately, many of the books still in use by primary schools to support their developing readers are not decodables and were actually specifically designed to promote not decoding, but guessing.

Consider the following, which is based on a real book (in fact these are often referred to by their fans as ‘real books’) and is a fair representation of many books I have seen*. The first page has a cat on a mat. The next page has a recognisably similar picture, but this time a dog is on the mat. Each page follows this pattern and has a picture that clearly and directly represents the text.

Note also the repetitive text. In the early stages there will usually be a phrase repeated throughout the book with a single word changed on each page to match the picture.

Now, especially if they’ve been taught some basic phonics, it is possible that a child being given such a superficially ‘simple’ book could decode cat, dog or even rabbit but it is more far more likely, as intended by the book’s design, that they will just guess.  Giraffe is a much more complex word and the child reader is unlikely to have yet learned that ‘g’ when followed by a ‘i’, ‘e’ or ‘y’ often represents /j/.  But, assuming giraffe is in their vocabulary and they know what one looks like, any child reading this book will be able to appear to ‘read’ it, having been carefully primed by the preceding pages. In fact, they might not decode or even guess any of the text at all as in many ‘balanced literacy’ classrooms the teacher/TA is likely to have read it to the child first so that they are familiar with the structure used and the child has merely to recite the sentences using the picture cues.

Path of Least Resistance

So, is it natural to guess? Well to say ‘natural’ potentially implies that learning to read is a natural process (it isn’t) and that context guessing is similarly a ‘natural’ part of this. it is more accurate to say it is understandable i.e. it is understandable that a child might short-cut the reading/decoding process and jump to making the seemingly obvious guess, after all they are beginning, not-yet-fluent readers so guessing the identity of the word may well be much quicker and easier that decoding cat, dog, rabbit and certainly giraffe. And why not if the result is the same? Where’s the problem? They have extracted the meaning haven’t they? Well, yes, but just not from the text, it was from the pictures or the priming by the teacher, and as they progress through their education such a strategy will become less and less successful.

But don’t good readers/adults use context-informed guessing?

Yes, competent adult readers may well guess the meaning of unfamiliar words from the context but there is a world of difference between occasionally guessing the meaning of a word from context (when the text supports it) and guessing the ‘identity’ of a word in your vocabulary.

Note the mitigation above, “if the text supports it”. Often a text will not give enough information to reliably guess meaning (let alone identity). For example, I encourage my Year 5 pupils to ‘collect vocabulary’ for which they write definitions, draw pictures and write example sentences. I explain that their sentences need to illustrate the meaning of the word (without guidance they often don’t). I give them this example:

I loathe going shopping.

and ask them what they think loathe means. They invariably say that loathe is a synonym for like or love. The penny only drops when I add

…because I hate crowded places.

My point is that, unless the text is explicitly designed to facilitate guessing it is often very difficult to do so accurately, so as a strategy for reading it is very unreliable particularly as even small errors can dramatically change the meaning without the child even realising it.

The following example may further illustrate the difficulties; I asked a new 10-year-old tutee to read a passage aloud about some children looking at the shapes which the clouds were making (it had an accompanying illustration). She has high comprehension when she reads accurately with no speech/language difficulties:

Instead of “see the long nose” she read “see the long noise” (a guess at a visually similar word).  Realising this didn’t make sense she went back, but instead of re-reading each word more carefully, she mistakenly assumed she had misread something else in the sentence and then ‘read’ “hear the loud noise” instead which she was then happy with. Hopefully you can imagine what might be going on when struggling readers are reading silently to themselves and how one incorrect guess can often be compounded as the child tries to mould what they see to match what they think they have read.

Quite apart from challenges posed by guessing, the interference it causes to other cognitive processes is considerable. Not only does the disfluent decoding that encourages guessing cause a significant cognitive overload, but the guessing itself also interferes with decoding. Over 20 years of observing struggling readers one-to-one has shown me how encouraging (even if only by failing to discourage) a child to guess is equivalent to training the brain to ‘shout out’ an answer with similar negative consequences as children habitually shouting out in class:

Imagine inheriting a class where the previous teacher has allowed, or even encouraged pupils to shout out their answers. Speed becomes the aim, not thinking, not accuracy. Even if you manage to re-train the children to at least put their hands up rather than shout, you can still see them squirming in their seats, arms bolt upright, faces straining, nearly bursting with the effort it takes to contain the ‘shout’ within. And as experienced teachers will tell you, as soon as that hand shoots up they’re no longer thinking nor listening, their brain is full of their ‘shout’.  Exactly the same is going on in the brains of guessers. Their brain is running interference, trying to ‘shout’ as they are trying to decode. This is an enormous hindrance to developing decoding fluency, to reading new vocabulary or even reading words that they know but infrequently encounter in print. Of course, this is only noticed if the child is reading aloud and mistakes are picked up on by a listener.

Conclusion

Ultimately, to be able to read and write is vital in our society. But while the effects of reading and writing may be truly magical, how reading/writing works is not. It can and must be explicitly taught if all our children are to be given free access to the pleasure, freedom and power that literacy can bring. Moreover, quite apart from my anecdotes and observations, the clear evidence over more than 40 years of direct research is that disfluent decoding, necessitating a reliance on context guessing, actually interferes with comprehension (i.e. reaching the meaning) and is a marker of a poor reader. As opponents of the systematic teaching of phonics say, and supporters of systematic phonics heartily agree, reading is about meaning, so it is therefore crucial that we not limit children by encouraging them to guess at the identity of some words based on what they think is the identity or meaning of some other words. That way madness, and for many, illiteracy, lies.

 

* you can see some examples of ‘real books’ here https://www.raz-kids.com/main/BookDetail/id/1437/from/quizroom/languageId/1