I recently came across Professor Kathy Rastle, who presented at ResearchED in Oxford last weekend, through an explosion of excitement on Twitter and photos of the slides she was showing and the one that struck me most said ‘Phonics: there is no other way’. An American academic now based at Royal Holloway and New Bedford College, she has made it her life’s work to improve literacy among children, young people and adults. You can find out more about her work at the Rastle Lab here: https://www.rastlelab.com/home.
Kathy says that computation of phonology is central to reading acquisition and skilled reading in alphabetic writing systems, and, as you may probably be aware, there are special problems between the association of spellings and sounds in English because the mapping is imperfect. She illustrates this with some research showing how British adults read 915 ‘non-words’, such as bamper, eluch, diswaps and congeave. The rationale was that it tells us what people have a learned about the relationship between spellings and sound. In the adults there were some words which were all pronounced in the same way such as bamper, and others, such as eluch, which had 22 different variations (out of 41 participants). The concept is very similar to the phonics check at the end of Y1 in English schools where children have to read a series of non-words such as fot, pib and zale.
The power of explicit instruction
Becoming a skilled reader in English and other languages requires us to learn this mapping between spelling and meaning - and it’s hard! And the less phonetic a language is, the harder it is, English being one of the hardest. How phonetic (how closely do the spelling and meaning rules stay consistent) across a range of languages can be found here in Seymour et al’s 2003 seminal list (https://vdocument.in/seymour-aro-erskine-2003.html?page=11)
|
Language |
Reading accuracy |
|---|---|
|
Austrian |
97% |
|
German |
96% |
|
Italian |
94% |
|
Spanish |
94% |
|
Dutch |
92% |
|
French |
72% |
|
Danish |
71% |
|
English |
34% |
English is ‘orthographically deep’ as it has one symbol to many sounds and many symbols-to-one sound and many exceptions. We can blame the fact that we were invaded over the centuries (by the Romans, the Anglo Saxons and not least by William the Conqueror). The upper classes spoke French for around 200 years after 1066, hence the influence of ‘higher order’ words such as ‘to descend’ (from descender) as opposed to the lower order ‘to go down’. This can be exemplified with children and young people in your teaching. If you ask pupils what words in English pronounce the ‘ch’ grapheme as a soft ‘sh’ sound, they will give you words such as machine, chandelier, chef, Charlotte and chalet. What do they all have in common? They come from the French word of course!
Kathy Rastle and colleagues undertook a really fascinating experiment about instruction in phonics. Over 10 days, 48 adults learned to read novel words printed in two artificial writing systems. One group learned spelling-to-sound and spelling-to-meaning regularities solely through experience with the novel words, whereas the other group received a brief session of explicit instruction on these regularities before training commenced which only took up 3% of total teaching time. Results showed that virtually all participants who received instruction performed at ceiling on tests that probed generalisation of underlying regularities. In contrast, despite up to 18 hours of training on the novel words, less than 25% of discovery learners performed on par with those who received instruction. These findings illustrate the dramatic impact of teaching method on outcomes during reading acquisition.
This has huge implications for the teaching of languages
If you could teach children the rules of pronunciation near the beginning of their language learning journey, and if you only spent around 3% of your total teaching time on it (if we think of the 5 year language learning course from Y7 to Y11 with 2 hours a week at KS3 and 3 hours a week at KS4 we have 468 hours). We only need to spend around 10 hours in year 1 (30 sessions of 15 minutes each followed by short 5-10 minute slots to review), and then recycle and practice them regularly particularly in Y8 and Y9, then it is an investment worth making.
As we can see from the table above, Spanish and German are very regular phonetically with very few exceptions. In Spanish with its 25 phonemes, ‘what you see is what you get’. You pronounce every single letter, so ‘chocolate’ would be pronounced differently from the English in the sense that the ‘o’ is pronounced ‘o’ as in ‘hot’ (and is always pronounced like that), and the ‘e’ at the end is pronounced like ‘eh’ as in ‘hay’. Spanish has a few challenges like the ‘ll’, ‘qu’ and ‘j’, ‘ge’ and ‘gi’ and ‘z’, ‘ce’ and ‘ci’.
German, with 44 phonemes, the same as English, is very regular, but we do need to be aware of some of its challenges, such as ‘ei’, ‘ie’, ‘w’, ‘j’, ‘ch’ and ‘z’. it has around 34 phonemes which are very different from English, which need to be taught. The others, such as ‘p’, ‘m’, ‘n’ and ‘b’ do not need to be taught to English native speakers as they are identical to English and will therefore be automatically pronounced correctly,
Although French, with 36 phonemes, comes lower down the table, it is very regular. However, it does have the ‘silent letter rule’ at the end of words, which makes it significantly less phonetic. But it does follow very clear and consistent rules of pronunciation. For example -s, -t, -x and -z are almost always silent at the end of words (apart from liaison, when followed by a word beginning with a vowel). In French the grapheme ‘oi’ is only ever pronounced the same way (‘wa’) and the nasal graphemes ‘in’, ‘on’, ‘an/en’ and the most high frequency word in French ‘un’ (which can be tricky for non-native speakers) only have one way of being pronounced.
Discovery learning is a myth
Kathy says:
“There is no other path, no “other approaches” to becoming a skilled reader of an alphabetic writing system, than through fluent decoding. Children must know what letters are and what they represent to progress in reading.”
As we saw in the experiment described about, learners were 75% more likely to unlock the written code with direct instruction. High quality phonics instruction allows learners in all languages to rapidly unlock code. As children are taught phonics in English schools (it has been the law since 2010) they will also understand all the terminology such as phonemes, graphemes, segmenting and blending and they know how to learn using a phonics methodology.
Want to find out more?
If this blog has given you a taste for more of what Kathy has to say, do what her Google Talk about Learning to Read, recorded in September 2022 here: https://www.youtube.com/watch?v=U_RUfYozxGY
Why not watch one of my free webinars, which demystify the terminology around phonics teaching (phonemes, graphemes etc) when, how and which phonemes to teach in French, Spanish, German which last around 35 minutes each here https://speakeasi.net/past-webinars/.
Wendy Adeniji