Diane McGuinness · Early Reading Instruction · linguistic phonics · visual phonics

Graphemes and phonemes, or how NOT to teach reading and spelling

Although I’ve written about the differences between linguistic and traditional (graphemic) phonics a number of times to date, I’m often being asked for further clarification. This I am more than happy to give because it’s in the detail of what we do at Sounds-Write that makes it so effective. So, how do the two orientations differ from one another? Let me count the ways!
Here, adapted from Diane McGuiness*, is a classic example of ‘graphemic’ practice from a generic phonics programme, whose name I’ve invented, called Graphemes and Phonemes. As the name of the programme implies, G and P presents from letters to sounds, what McGuinness describes as an ‘entirely visual logic’. In other words, it does not make explicit from the start that the finite number of sounds in our speech is represented in print by spellings of those sounds.
In week 2 of G and P, the letter [ c ] is introduced. This, the children are told is the letter ‘see’ and the letter ‘see’ ‘says’ (or ‘makes’) the sound ‘kuh’. We’ll leave aside the fact that now young children have several things to remember: the visual symbol on which they are focused, the letter name, and the imprecisely rendered ‘kuh’ (two sounds, not one!). Some teachers will also link the spelling to a character and/or an action, making yet more things to learn and remember – a serious case of cognitive overload for young children. There are several things in this kind of practice that could easily confuse a young child. The first is what they are to focus on? Is it the letter name or the sound, or both? Is the addition of a character (a snake for /s/, for instance) or an action (weaving your hand like a snake) likely to help, or are these yet more things to hold in the working memory of a young child?

Even if the letter [ c ] is presented on its own without introducing the letter name and the teacher starts to say /k/ /k/ /k/, the sound-spelling correspondence is being taught out of context and, in that sense, the sound /k/ has no meaning. Doing this for all the letters of the alphabet and then expecting young children to remember the associations is extremely hard for them.
You might at this point object that not all teachers say sounds imprecisely. In fact, I readily concede that if there has been one singular improvement in the nearly fourteen years we have been running Sounds-Write trainings, it’s that our participants are much more likely to be aware that saying sounds accurately is very important. Fair enough! You might also object that not all programmes include characters or actions to go with sounds. Fine!
What you do have is teachers telling children that letters ‘make’ or ‘say’ sounds. They don’t! And I’ve explained why here and here. But let’s explore this further in the context of how things progress. Having been taught [ c ], the children might be asked to bring in some item from home that has a ‘kuh’ in it and they’ll turn up with a car, a cup, a picture of a cat and/or a caterpillar, and so on. That’s if they don’t bring a koala or a picture of some games kit!
A few weeks later, the class is introduced to the letter [ k ] and told that it’s the letter ‘kay’ and that ‘kay’ says ‘kuh’. So now we have two letters, ‘see’ and ‘kay’, and they both ‘say’ ‘kuh’, or /k/ if the teacher is articulating the sound precisely. Remember that these children are four years old and there is huge potential for getting confused. Do they try and use sounds or letter names or both? In the child’s mind, do the sounds these letter names ‘make’ change? Is the sound /k/ that ‘see’ is ‘making’ different from the one ‘kay’ is ‘making’? Although the idea that the sound /k/ can be spelt in different ways may be obvious to an adult, it will probably be far from obvious to a young child.
There is also the issue that highly acoustically perceptive children will notice that the sound /k/ will differ very slightly depending on where the sound happens to appear in a word. The /k/ at the beginning of a word is likely to differ slightly to the /k/ in the middle or at the end of a word, and this may further confuse the issue.
So now, the programme moves on a bit more and a few more weeks later the teacher introduces [ ck ], almost certainly without making explicit that the spelling is two letters but that it is one sound. ‘These’, too, the children are told ‘make’ the sound /k/.
By this time, many children will be completely baffled and probably assume that they have misheard or misunderstood what is being taught. Bear in mind also that some programmes will be teaching these letters out of context as paired associate learning, something adults, as well as young children, find difficult to do.
What should happen? Sound-spelling correspondences should be taught in the context of words, initially through word building exercises as I explain here, and not through the teaching of letters in isolation. Presenting sound-spelling correspondences in this way makes it psychologically ‘real’ to children – the sounds in speech are represented by spellings. Once word building has been completed, the children write the word. This process quickly builds up to hundreds upon hundreds of words, beginning first as isolated words but then quickly building into sentences and then whole texts. The simple one-to-one sound-spelling correspondences should be taught first, giving children the opportunity to practise the vital skills of segmenting, blending and phoneme manipulation to mastery level before embarking on teaching the greater complexities of the code. Children who can blend and segment with aplomb do not read words like ‘strom’ as ‘storm’!
When spelling alternatives for a sound are taught, the teacher should be making the point explicitly through practical activity that makes crystal clear how the code works: “This [ c ] is the way we spell /k/ in this word. Later, I’m going to show you some other ways of spellings the sound /k/.” A few weeks afterward, we can introduce the spelling [ k ] and say, “In this word, we spell /k/ like this [ k ]. Now we have two ways of spelling the sound /k/!” and so on. As long as instruction is grounded in the 44 sounds of the language, adding more alternatives, such as [ ck ] (‘stick’), [ ch ] (‘chemist’), or [ qu ] (‘mosquito’), is then much easier to do later as children’s ability improves with practice. We can also expand on the idea that we can spell sounds with one letter by introducing two-letter spellings, such as [ ck ], by saying “This,” pointing to the [ ck ], “is two letters but it’s one sound.”
Children can’t be expected to learn what is a highly complex (opaque) code if they are not presented with a programme that is taught explicitly, systematically, and from simple to more complex:
  • sounds are spelled one at a time from left to right across the page
  • a spelling can contain one, two, three or four letters
  • we spell sounds in more than one way and teach over time the different ways of spellings all the sounds
  • most spellings can represent more than one sound.
None of these ideas is difficult to teach if presented carefully and from simple to more complex.
*McGuinness, D., (2004), Early Reading Instruction, MIT Press, (page 42).