A while ago (before we went on holiday and I got distracted) I had a very interesting Twitter conversation about the specifics of using synthetic phonics to sound out certain words/syllables.
The central issue was that I wondered how to explain, without using the concept of a ‘magic e’, how words like ‘mice’, ‘lace’, ‘page’ and ‘huge’ should be understood in terms of SSP.
My question was whether such words should be conceptualized as
- Vowel_e + soft Consonant’ [eg a_e + soft g = age]
- ‘Long Vowel + (soft Consonant + e)’ [eg long a + ge = age]
- Something else.
The responses were really thoughtful and interesting, and the second conceptualization was considered to be the most practical.
I was also given very sensible warnings that SSP is meant to be pragmatic rather than an exact science, which I understand.
Over the last few weeks, though, I’ve been mulling this over, partly because the logic of either option still seems hard to explain to a child (for reasons I’ve set out below), but also because I’m interested in other ways of thinking about explaining these particular sounds, and I wanted to take a particular (and relatively tricky) piece of SSP to its logical conclusion, to see where it led. I hoped that this might help me understand the inner workings of SSP itself.
To clarify, my understanding is that for the pragmatic purposes of SSP, written words are considered to work like this:
i. Letters form graphemes
ii. Graphemes are code for specific sounds
iii. Sounds as represented by graphemes are put together to make words
iv. Each grapheme is distinct, so a single incidence of a letter in a word can only be used in one of the graphemes which make up that word (that is, the ‘e’ in ‘ice’ must be either part of ‘i-e’ or ‘ce’ but not both).
So… the problem I see with option (1), a_e+g, is that if you understand ‘a_e’ (or ‘i_e’ or ‘u_e’) as split digraphs, which at first sight, of course, they would appear to be in SSP terms, then there is the problem of why the consonant is softened. If ‘a_e’ and ‘g’ are treated as *discrete* building blocks which do not affect each other, if each grapheme can only be code for one sound, and no letter can simultaneously form part of two graphemes in the same word, then there is no explicit reason, according to SSP’s explanatory logic, for the softening of the ‘g’. Yet the softened ‘g’ is needed, and since the ‘g’ is followed by an ‘e’, the child will notice this and so the likely result is that they will read it as ‘ge’. How to explain, then, using the specific logic of SSP, that the ‘e’ in fact affects both letters, if ‘a_e’ and ‘g’ must be considered as separate digraphs? Or how to explain that the g is softened but not because of the ‘e’ as would usually be the case? It’s not satisfactory, which I assume is the reason my respondents chose option (2).
A similar problem arises with option (2) though: if you see ‘ge’ as a digraph for ‘j’, and the ‘a’ as a separate grapheme, then how to explain the long ‘a’? Respondents to my question suggested that it could be explained that ‘a’ is sometimes, on its own, code for long ‘a’. This is true, but the problem with this, I feel, is the potential for confusion on the child’s part. It is clear that there is an ‘e’ at the end of the word, with one consonant in between it and the ‘a’, so it *looks* like a split digraph with a consonant in the middle. If it is *not* to be understood in this way, the logic ends up being:
‘‘a_e’ should generally be read as a split digraph representing code for long ‘a’ (as in ‘rate’), but when the consonant in between the parts of the digraph is softened by the ‘e’ at the end of the word, ‘a-e’ should not be read as a split digraph but as an ‘a’ as code for long ‘a’ plus a digraph representing the softened consonant.’
This seems to be to be a rather complicated little tangle for a learner reader to come to terms with.
I experimented on the boys to see what they made of it. Of course a sample size of two identical twins from the same family is not exactly sound, but they’re my only available resources…
I tried both (1) and (2) and they were rather dubious because, to them, in both cases there’s a common and easily recognizable form of digraph which has to be ignored in order for the explanation to work.
One idea I had was to use transparencies to put the ‘ge’ over the top of the ‘a_e’ so that the ‘e’s matched up; I liked it, but they weren’t too keen. After various attempts to explain myself, the version that they liked most was this one:
‘The a-e is a split digraph, and when you pop the ‘g’ into the middle it’s affected by the ‘e’ too.’
This means the a-e is conceptualized as the most common thing that it’s code for, but it also allows for an explanation of the modifying function of the ‘e’, which of course really works, orthographically speaking, in place of two diacritics – one on the ‘a’, and one on the ‘g’.
The problem with this is that it does not conform to the SSP approach as I understand it, where the letters form distinct graphemes which are code for one sound only.
Am I correct in my understanding of SSP’s logic? SSP seems to have been constructed deliberately in a way which avoids the concept of a modifier, so that ‘-ge’ and ‘-ce’ are presented simply as code for ‘j’ and ‘s’ sounds, and the orthographic purpose of the ‘e’ is not discussed. My assumption is that this was done to avoid the potential for confusion.
If my understanding is correct, what are the implications of the fact that a lot of very common words follow this pattern of a dual-purpose ‘e’ in a word, and can only be explained either by breaking the SSP rules and using the concept of a modifier (the ‘magic e’ idea is of course one way of expressing the modifier concept to children, and SSP rules do not allow for this approach), or by weakening the relationship between grapheme and phoneme (by saying that, for instance, a-e is frequently actually *not* a digraph representing long ‘a’, even when this is the sound which the word requires and the a-e pattern is present)?
All rules for beginners turn out, of course, not to apply to advanced practitioners – the ‘forget everything you thought you knew’ aspect of getting deeper into any field – but given the commonness of the kinds of words discussed above, they are bound to play a large part in a child’s reading long before they are at a stage when a SSP-based literacy curriculum could accommodate learner readers abandoning its rules.
I know this is a fiddly little question, but I think it’s really interesting. How might SSP incorporate a robust approach to these words which does not run the risk of tying teachers and children in knots trying to avoid the obvious fact that the ‘e’ modifies both the vowel and the consonant in these cases?