Comments 26

Code-Teaching or Code-Breaking?

Code-breaking is what you do when you don’t know the code. 


There are two main approaches to teaching reading, which can be summarised as code-based and meaning-based. Code-based teaching works on the premise that there is a known, culturally shared, symbolic code. This written code represents another, spoken code. Teaching the relationship between these two codes is the focus of systematic synthetic phonics. Because it is not a natural process to interpret written symbols, this aspect of language must be taught systematically and explicitly to ensure that all readers not only acquire, but also master the skill. Once the code is mastered, the meaning of the text is available to the student and the cognitive benefits of reading accumulate.

Enigma Machine

The Enigma Machine

The meaning-based approach to reading, exemplified in whole language and its descendant Reading Recovery, takes a code-breaking rather than a code-teaching approach. The text is approached as a puzzle to be solved, analogous to deciphering an intercepted wartime message. The code-breaker might consider the participants: from whom? To whom? Then there is the question of purpose: why was this written? The cracker looks for clues in the medium, or in other related messages that might have been discovered. Repeated words or phrases are checked. Reasoned guesses are made, checked, confirmed or disconfirmed. The hope is that eventually a pattern will emerge. The more that is known, the fewer possibilities there are for what the remaining text might mean. Assuming that the code is broken, the message emerges and its importance can be judged. Then it is on to the next message. Hopefully the lessons learned from the previous code-breaking exercise will help with this one.


The three fundamental differences between these approaches are accuracy, speed of return and efficiency. In the case of the code-teaching approach, mastery of the code, built with practice, will inevitably lead to greater accuracy than a less systematic approach. In terms of speed of return, the code-teaching approach may delay the deciphering of some messages – but only in the short term. The problem is overcome by efficiency: once the code is mastered, a great many texts can be deciphered quickly, and the intelligence gathered can be put to use.

On the other hand, while the code-breaking approach may yield some meaning early, the context, guess, confirm sequence is inefficient and often inaccurate. At a certain point, (usually by the end of Year 4) readers need reliable information at their fingertips so that they can work with it. This stage is called “reading to learn” instead of “learning to read”. And it is at this stage that the inefficiency and inaccuracy of the code-breaking approach becomes apparent. Students taught this way have frequently not been taught to fluency, and the strategy of guessing and predicting has left them never being really sure what was in the text. While their teachers may feel pleasure that the students are “constructing their own meaning from texts” the students tend to feel like failures – because they are failing.

Enigma Machine2

The code-breaking approach is unsupportable as an educational practice because:

  • It wastes children’s time, teaching them to use code-breaking strategies when we already know the code.
  • As educators we have a responsibility to pass on our knowledge – not to require children to “discover” it.
  • After a few years, the poor guessing strategies of code-breaking prevent children from accessing the knowledge that schools (and society) expect them to be able to find in written texts.
  • The limitations of poor reading hinder the development of language skills, thinking skills, vocabulary and curriculum knowledge.

Regrettably, arguments about the merits of the two approaches will no doubt continue. Ideals about human nature, society and learning have deep roots in political and philosophical streams that are not easily severed – not even by the sharp blows of logic and empirical evidence. Teacher education institutions seem particularly prone to idealizing code-breaking by positing that the code is not teachable; see John Walker’s Literacy Blog for a cogent response to a recent example of misconceptions about the code.

Book Magic

For our part, the code-breaking approach of ‘whole language’ has produced an unending stream of children reaching secondary school in need of help. We look forward to the day when logic prevails, and good teaching at primary school level makes Thinking Reading unnecessary.

Visit our website

You may also be interested in:

Pulling the Strands Together

The Case for Design in Curriculum

Teaching Reading is Rocket Science

What Does Mastery Really Look Like?

A Question of Progress


  1. nemocracy says

    You are confusing two aspects of reading here.

    The decoding aspect (pronouncing the words) is reasonably well served by imagining English spelling as a code and learning the code. It’s not foolproof because the English spelling system is rife with idiosyncrasies and irregularities. It is not a code in the sense of having one to one symbol/sound correspondence.

    Understanding the text is another, different aspect of reading. A beginner reader might be able to decode the word ‘tram’ without understanding it. Understanding the meaning requires learning vocabulary, which sometimes happens through inferring from context. This aspect of reading is not about cracking a code, or learning a code. It is about expecting texts to make sense and using that assumption to find out what sense they make. Using this reading skill to decode won’t work well. It is nevertheless important to decode with the knowledge that texts make sense.

    Both the aspects you identify are important. Nowadays whole language is out of fashion and SP has got a grip on early education. But neither approach is enough by itself.


    • Thank you for taking the time to write a reply. There is no disagreement with (or confusion about) the distinction between decoding and comprehension – indeed, the entire post is predicated on this distinction. The point is that the primary basis for teaching reading in the early years needs to be mastery of the code. If students have language-rich backgrounds, their vocabulary and comprehension will continue to develop. If they do not have a large vocabulary, and have less extensive experience of language, mastery of the code will enable them to access more language more quickly. In other words, students from language-rich backgrounds and those from less language-rich backgrounds will both benefit from mastering the code as early as possible.
      There is, however, a fundamental problem with your assessment of the language code as “rife with idiosyncrasies and irregularities”. If the code is learnable (as clearly it is, since, for example, we are using it to communicate) then it is teachable. The code is in fact far more systematic than many people suppose. The error arises principally because of a supposition that there should be a one-to-one correspondence between sounds and letters. In a 26-character alphabetic system representing 44 – 46 phonemes, this is clearly impossible. If the code is conceived of as symbols or patterns of symbols representing sounds, then it immediately becomes more transparent, more systematic, and most importantly, more teachable. The link in the post to John Walker’s Literacy Blog is well worth following up, as are many of his other posts. See also Susan Godsland’s site,, for example this page:
      The assumption that the code is not reliable enough to teach is frequently made by people who have themselves mastered it, and this is particularly distressing: despite the fact that they have their own evidence of success, they assume that it cannot be systematically taught to others. That assumption results in 20% of children leaving school unable to read and write properly.
      For further reading, have a look at our site on the Professional Reading pages. You will see that we have a page of links on issues around phonics vs whole-language approaches (The Reading DEbate). There are also resources on teaching comprehension, fluency, grammar, vocabulary, and motivation.


  2. nemocracy says

    Thank you for your response. I’m going to stick with my assessment that you are confusing two aspects of reading. The problem I find with your description is that you seem to regard Whole Language teaching as aiming to teach decoding (being able to pronounce). I’m not a fan of Whole Language, but it is my understanding that it approaches decoding as a ‘being able to understand’ process. So to say that it is inefficient in teaching decoding (being able to pronounce) is to confuse its purpose with that of phonic decoding. Clearly if you regard being able to pronounce words correctly as paramount phonic decoding is the way to go. However, I can’t agree that being able to pronounce words correctly is paramount in being able to read. Comprehension is the goal. The fact that you can’t understand text unless you can decode it is obviously really important, but I think Whole Language is right to identify a large picture in decoding (context, vocabulary, morphology) which contribute to the small detail picture contributed by phonology.

    I can see where you’re coming from when you want me to acknowledge more regularity in the ‘code’. However, I again have to demur. SP programmes have done a great job in pinning down the code, but they cannot take that last step for children, the step where the child decides between alternative grapheme-phoneme correspondences in identifying a word that is ambiguously spelt. Over the years various ways of identifying regularities in the language, by creating rules, or more recently by listing all possible alternatives, but they have failed in this regard. Rules are cumbersome,difficult to apply and have exceptions. Lists and charts describe the problem but do not solve it. So something else must be learnt, and that is recall of which whole word uses which alternative spelling. I don’t see how phonics training can do this. Enthusing children to do lots of reading and working it all out through experience stands a better chance – paired up with a careful grounding in phonics. Children who have good vocabularies and good experiences with books will be more likely to enjoy reading than children who don’t have this head start. Schools need to prioritise the provision of speaking, listening and reading experiences as well as teach phonic approaches to give equal opportunities.

    Thank you for suggesting further reading.


    • A few quick points:
      You said: “you seem to regard Whole Language teaching as aiming to teach decoding (being able to pronounce)”
      1. The purpose of phonics is not to “pronounce” but to identify words accurately. The words must be known/learned in spoken language in order to extract meaning. Therefore meaning is important in phonics teaching.
      2. What we have said about Whole Language is that is “meaning-based” and therefore not an efficient strategy for teaching a known code.
      3. Context, morphology, vocabulary and so on are acknowledged as essential skills for acquiring meaning. I don’t think we are in any disagreement over this. However, these are not the exclusive preserve of Whole Language. As you say, they can and should be combined with good phonics teaching.


  3. I am going to respond again, as this blog post has come up on Twitter recently:

    To deal with point 1: Phonics is all about pronunciation, there is nothing else applying synthetic phonics will do for the reader except enable pronunciation. Once pronunciation is achieved the reader may recognise the word from their vocabulary and, if they know the meaning, this will enable them to read. It is a combination of phonics knowledge and language knowledge that enables the child to read. And it’s very useful if the child has a wide vocabulary and a wide experience of books.
    Synthetic phonics, applied without a supplementary search for meaning, is not enough – decoding something using phonics alone won’t give a child the meaning of a word.

    It is also significant that applying SP alone will not necessarily give the child the correct pronunciation of a word. Words sometimes contain graphemes which can represent a variety of phonemes. We hope the child will plump for the right phonemes, or will patiently try out different pronunciations, eventually stumbling on, and choosing, the right one. Phonics is a blunt instrument until combined with vocabulary knowledge and a sense of how syntax works. Children with good vocabularies, who have had experience of books, are more likely to have come across the word before, and will have an understanding of how words work together to express meanings. So they are more likely to choose correct pronunciations from the possibles presented through their phonic knowledge. Indeed, some of these children may bypass sounding out carefully from left to right. They may know what word fits and recognise from sampling letters in the word, that that is the word written.

    SP is a blunt instrument. Children need to encounter a range of books and build their language skills alongside phonics lessons. And it needs to be recognised that children may use strategies which are not strictly SP as they develop reading skill.

    Hopefully I have now addressed your point 1. Your points 2 and 3 are built on the assumption that Whole Language and SP are mutually exclusive: that you have to follow one or the other. It’s a false dichotomy. Teachers need to assess their pupils and teach to their strengths, not blindly follow SP or WL. I think I have shown above that in order to read for meaning pupils apply their knowledge of whole words and syntactical conventions alongside phonics to arrive at accurate decoding, supporting the search for meaning. I’m surprised you don’t see this and yet acknowledge the role of morphology. No, it is not the preserve of Whole Language. Let’s get rid of these false categories and apply intelligence in choosing the teaching methods which support reading instead of disappearing into muttering clusters, turning to glare at the opposition. In blindly following the governmental push for SP we are abdicating our responsibility as teachers, part of which is to be critical thinkers.

    I am sorry this comment is so long! Many thanks for your blog post. It is important to have freely available opportunities for discussion at length. Twitter can only offer a starting point.


  4. The post does not anywhere suggest that synthetic phonics is all that is needed for reading. This post focuses on decoding. See also this post – Pulling the Strands Together – in which I do focus on combining the different elements of reading, a view propounded by many who also propound SSP, e.g. Lyon, Moats, Kame’enui. Your insistence on challenging a view that the post does not advocate is not something I can do anything about.
    I agree that syntactic, morphological and contextual clues are necessary for accurately discriminating some homophones. This does not negate the necessity for accurate and fluent knowledge of the code – indeed, without this, there is no discrimination to be made.
    As I said previously, I advocate the teaching of vocabulary, syntax and morphology. However I do not advocate the use of Whole Language to do so. It is not only unnecessary, but actively undermines systematic code knowledge.
    Given the repeated misrepresentation of the original post’s content, I don’t think there is any point in continuing this line of discussion.


  5. I’m sorry you feel like that.

    Just to make one point which I think encapsulates my different view: you link to a blog by John Walker in which he says that children, looking at a circle, can be quite happy to regard it as a pizza, a moon etc – any circular object. He uses this observation to support the idea that children are able to regard a grapheme as possibly representing a range of phonemes without confusion.

    Indeed a circle could represent a range of objects, and a child might identify these if asked. However, how does the child know which of the objects it represents if there is no context to help? In picture context is provided: a sun may be represented by a circle with rays emerging; a moon might be represented by a circle in a starry sky; a pizza might be represented by a circle adorned with splodges of red tomato etc. Only when the context, providing extra information, is added to the circle will the child know which of many objects the circle is meant to represent. The sun is a good example, as there is an embedded convention of drawing those rays. Similarly a grapheme may represent a range of phonemes, and children may be aware of this possibility. It is through context that the phoneme it represents can be divined. Sometimes the context is like those sun’s rays. The conventions of the language, like the sun’s rays, are a supporting context ie the grammar of the language. The child who speaks grammatically will expect the word to conform to this grammar: language skills, in addition to phonic skills, will be contributing to decoding.

    I agree that your post concentrates on decoding, but you seem to be saying that decoding through phonics is the single means of arriving at the word. Unfortunately, as I said above, phonics is a blunt instrument for this purpose if used on its own, not only for decoding homophones but for decoding any word in which there are graphemes that can represent a range of phonemes. SP helps the pupil to arrive at a possible pronunciation, but to arrive at a word the pupil must arrive at the correct pronunciation and recognise the word. There is no doubt that phonics is a useful tool contributing to this process, but not the only tool.


  6. Tami Reis-Frankfort says

    Reblogged this on and commented:
    Why not teach children the phonic code if we know they will need it?


  7. Pingback: SEND the right message | thinkingreadingwritings

  8. Pingback: Why We Can’t Remember How We Learned | thinkingreadingwritings

  9. Pingback: Spring Cleaning | thinkingreadingwritings

  10. Pingback: Choosing the right tools for the job | thinkingreadingwritings

  11. Pingback: Reflections on Hidden Potential | thinkingreadingwritings

  12. Pingback: Reflections on the Hard Road to Success | thinkingreadingwritings

  13. Pingback: Leveraging Literacy at Secondary School | thinkingreadingwritings

  14. Pingback: What do you do when you don’t know the code? | Tutoring Colorado

  15. Pingback: No Excuses Left | thinkingreadingwritings

  16. Pingback: Looking Back – old problems, new challenges | thinkingreadingwritings

  17. Pingback: Looking Past the Masks | thinkingreadingwritings

  18. Pingback: Does phonics help or hinder comprehension? | thinkingreadingwritings

  19. Pingback: Wars and Waste | thinkingreadingwritings

  20. Pingback: Doors to Opportunity | thinkingreadingwritings

  21. Pingback: Does it matter if some can’t read? | thinkingreadingwritings

  22. Pingback: The Stubborn Gap | thinkingreadingwritings

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.