Episode 4: Sound to Print with John Walker

In the fourth episode of The Sounds-Write Podcast, John Walker discusses the sound to print approach to teaching reading and writing. He talks about the rationale behind using this orientation, how this approach differs to print to sound methods and the results we’ve seen from schools using the sound to print approach. Listen to the end of the episode for an exciting sneak peak into one of our upcoming projects. Enjoy!

Some helpful links:
Our case studies book
John's Twitter
John's blog
Learn more about Sounds-Write

Transcript

Laura:  00:00
Hello, and welcome to the Sounds-Write podcast. I'm the host, Laura, and in this month's episode, we'll be hearing from Sounds-Write's founder, John Walker. John discusses the sound-to-print approach to teaching reading and writing. He talks about the rationale behind using this orientation, how this approach differs from print-to-sound methods, and the results that we've seen from schools using the sound-to-print approach. I hope you enjoy the episode. Hello, John. Welcome to the podcast again.

John:  00:34
Yes, hi Laura, and thanks for having me back on the podcast. It's great to be here.

Laura:  00:39
Always great to have you. So, in this episode, we're going to be discussing the sound-to-print approach to teaching reading and writing. This is something that a lot of people seem to be talking about at the moment, actually. So it'd be great to hear from you, John, on the rationale behind why exactly Sounds-Write uses the sound-to-print approach and what this means in practice.

John:  01:00
Yes, absolutely. No, you're dead right. It is a hot topic of discussion at the moment, as you say. The scientific study of learning to read and write, is about mastery of a human invention. We need to teach the alphabet code the way it was intended, from sound-to-print.

Laura:  01:19
Right. And at this point, it's worth mentioning that our writing systems were, in fact, developed from sound-to-print, in that order. So, it's thought that spoken language emerged between 50 to 150,000 years ago, roughly, and it developed naturally as a way for humans to communicate with each other, whereas it took much longer for written language to emerge. And that happened around 5000 years ago in Mesopotamia, which is now modern day Iraq. And people took spoken language and they developed methods of recording that. So it literally went from sound-to-print in the way that it was developed. And almost all children learn spoken language with great ease. They grow up around spoken language. They pick it up, as you've said before, I think it's a quote from Steven Pinker, 'as easily as a spider learns to weave its web'. Written language comes after that and most children will begin learning to read and write once they've achieved some level of mastery of spoken language at their level. So, children pick up spoken language very naturally, whereas it's clear that in order to understand and produce written language, children do actually need explicit instruction. You always begin each Sounds-Write course with this information, and I think it's a great preface to the episode as well. Not only is it really interesting to understand where written language comes from, and maybe we should do another episode on it more specifically, because it's just a fascinating topic, but it's also really important to frame this discussion of sound-to-print with that understanding that actually, that is how writing systems developed.

John:  03:09
Yeah, no, you're dead right. It is really important for teachers to understand that speech is biologically primary, along with quite a number of other things that we learn naturally, whereas writing is biologically secondary. It's a human invention, as I've just said before, and all human inventions, such as mathematics, learning to read and write, musical notation, physics and so on, need to be taught systematically and explicitly. Peter Daniels and William Bright make this important point in their book 'The World's Writing Systems', when they say, 'language skills are a natural product of the human mind, while writing is a product of the human intellect. No infant illiterate absorbs its script along with its language. Writing must be studied.' So, they're making the point there that although of course it's perfectly natural to learn to speak and to listen, reading and writing need to be taught. So I want to start by saying that if young children are to learn how to read and write successfully, they need to understand the logic of the alphabet code from the beginning. And that logic is most easily understood if we teach it as it was developed from sound-to-print. And we can do this. The question is, how? Well, by teaching them activities first developed in the 1970s by a Russian child psychologist, Daniil Borisovich Elkonin, in which they build common CVC words that are already part of their oral repertoire, words like 'mat', 'sat' and 'sit'. So these are words that they're already very familiar with in their spoken language. And it makes perfect sense, of course, to be teaching them simple words like this made up of these simple consonant-vowel-consonant structures.

Laura:  05:10
Yeah. And it always makes me think of, I've heard you say before, it's like if a child is learning to read, say the word 'mat', you can say 'em' 'ay' 'tee' till the cows come home, but you're never going to get the word 'mat' from saying 'em' 'ay' 'tee', 'em' 'ay' 'tee', 'em' 'ay' 'tee'.

John:  05:32
Yeah, no, that's exactly right. Whereas, of course, /m/ /a/ /t/, you can hear 'mat', and that makes perfect sense to them.

Laura:  05:40
And so, spellings are always introduced, in the Sounds-Write approach, spellings are always introduced in the context of words, right?

John:  05:49
That's right. What becomes immediately obvious to children learning through activities such as word building is that words are comprised of sounds, and that to begin with, we spell those sounds with single-letter spellings. Now, the beauty and simplicity really, of a sound-to-print approach is that there are a finite number of sounds in English which all linguists agree is between 43 and 45, depending on accent. This limited number of sounds gives you an anchor for the spelling system in the sense that those sounds don't suddenly change. We don't suddenly drop sounds out of the language or bring new ones in. They're constant. In fact, that is the constant. The sounds of the language are what every child everywhere grows up to acquire naturally. As I've said before, the spelling system, on the other hand, of course, needs to be taught, and it needs to be taught from simple to more complex. Even in an opaque writing system such as English, reading and writing is perfectly straightforward to teach as long as it is approached systematically. And of course, as long as teachers are taught how to teach it. So, as I said, we begin with simple one-letter spellings of sounds in words that children are very familiar with. The amazing thing is that after learning to build, read and spell four or five words with five sound-spelling correspondences such as 'mat' and 'sit' and 'sat' and so on, they're really excited to be able to do the same for more than 30 new words with the addition of only another three sound-spelling correspondences, and in a very short space of time, a matter of weeks at most. So, children get really excited about this, it shows them exactly what they can do. And after that, the process, of course, is generative. The more sound-spellings we teach, the greater the number of words children can read and write. The process builds to increase exponentially.

Laura:  08:00
And you get children who within a month or two of starting Kindy or Reception, being able to read books with one or two sentences per page because, yeah, this process is exponential. They start to be able to read many more words as long as they have this CVC structure and contain the sounds that they've already covered in class.

John:  08:23
Yeah. In fact, what I find really exciting is that after three or four months or so, many children are able to write anything they want to write. Now of course those words won't be spelled correctly, but they can make a plausible go at it. They can read it back, their parents can read it back, the teacher can read it back, and they can read it back a week later because it's obvious what they're trying to write. So, for instance, a word like 'uncle', I've seen spelt as < uncl >. Well, that's perfectly plausible. Now, of course, it's going to be some time before we teach them the way that we spell /l/ at the end of 'uncle'. But in the meantime, they have a tool in their hands that they can use all the time, and I absolutely love that.

Laura:  09:13
And that's indicative of the fact that that child understands that the sounds that they are saying when they say the word 'uncle' are represented by spellings. And for now, maybe they haven't covered the spelling < le > of /l/, but they will represent that with a spelling that they have already covered to represent that sound.

John:  09:34
Yeah, absolutely, exactly.

Laura:  09:37
So could you describe what a print-to-sound approach might look like and what the difference means, in practice, to a sound-to-print approach?

John:  09:47
Yeah. In fact, there are lots and lots of differences between a sound-to-print approach and a print-to-sound approach, not least of which are all these rules that children are supposed to learn. And in fact, many of these rules are pretty unlearnable, because if you learn the rule, you've got to learn all the exceptions to the rule, then you've got to learn all sorts of funny things, like you've got to learn that hard < c > or soft < c >. But one of the things that I think I probably most object to is the fact that young children are often presented with letters on flashcards and asked to parrot sounds that are cued by their teacher. So, they've got to say things like /m/ /s/ /a/ et cetera, et cetera. And that's even if they're saying those sounds as they should be said. And this is less than ideal for a whole number of reasons. The first thing is that by presenting arbitrary letters like this, it can quickly become incredibly boring for young children, especially if they don't know what the purpose is of learning these, making the connection between these sounds and these flashcards. And again, that's very, very difficult to do. It's called paired associate learning in psychology, and most humans find that very hard to do. So, by presenting children with letters and asking them to produce these sounds represented by the letters, teachers are reinforcing this idea that language goes from print-to-sound, which is incorrect. It also presents letters outside of the context of real known words. And young children have, very often, they've got no idea what they're doing it for. So, as we said earlier, letters are there to represent the sounds in words, and it can be very confusing for a child to be taught sounds in isolation. So, this is something that we don't do in Sounds-Write. There are also huge differences in the language that print-to-speech teachers use. So, for example, many teach children that letters or spellings say or make sounds, when in fact it's humans who say and make sounds. And when they do this, the orientation of the code is completely lost.

Laura:  12:09
So, I just want to unpick that a little bit. So, letters don't make or say sounds, that's something that you talk about time and again. Let me just check that I've got this right. It's important to avoid saying that because it's us, it's humans that say sounds, not the letters. Language goes from sound-to-print, so it's important to make that distinction when we say that letters represent sounds rather than the common misconception that letters make or say sounds. Have I got that right?

John:  12:44
You have, exactly. And it's an important distinction. Why is it important? Because, as professor Diane McGuinness used to remind us, children are really magical thinkers. And if the spelling of a single letter, for example, the spelling < a > can be /a/ in 'mat', it can be /ae/ in 'April', it can be /or/ in 'tall', it could be /o/ in 'wash', it could be /e/, in 'any'. Well, you get the idea. They can easily think that these letters can say anything and that the code is impossible to learn.

Laura:  13:21
So, you mentioned Diane McGuinness there. Could you talk about some of the research on the difference between sound-to-print and print-to-sound approaches? We know, obviously, that writing systems were developed from speech-to-print, so, in that sense, it does make a lot of sense to teach it in that orientation. But is there any evidence that suggests that sound-to-print is a more effective approach?

John:  13:47
Okay. The truth is really that there's very little research on sound-to-print approaches. Nobody seems to be terribly interested in doing it. So, one of the things that we're trying to do is we're trying to do some of this research ourselves. But I mean, the problem with that is, of course, if you do it yourself, then people are going to call it into question, I suppose. But having said that, we do encourage schools doing Sounds-Write to give us their Phonics Screening Check results. The Phonics Screening Check, for listeners who don't know, is a test that's run annually for all students in England at the end of their second year of schooling in Year 1. And we also get results too, from many schools that are using a properly normed and standardised test, the Young's Parallel Spelling Tests. I'd like to give you a particular case in point from a school in England, in the UK. In this school, 40% of children are on free school meals. So, of course, as you can gather from that, it's by no means a school in a leafy suburb. So, it's a challenging environment. The school has provided its Phonics Screening Check scores at the end of Year 1. And you could read in the study - you can read that in the study that we provided. I think you're going to provide them in the notes, aren't you, at the end of this?

Laura:  15:16
Yeah, I'll put them in the show notes.

John:  15:18
Okay, lovely. So, during the pandemic, the PSC was disrupted, so that instead of doing the check in June 2020, when they were unable to do it, they did it in October 2021, 90% of the students passed and 90% of the students eligible for pupil premium also passed. And this compares very favourably with pre-pandemic national results of 82% for all students and only 71% for pupil premium students. And again, pupil premium students are those coming from disadvantaged backgrounds.

Laura:  15:57
That's amazing that this school was getting better results than the pre-pandemic national average, even post-pandemic, that's amazing.

John:  16:09
Yeah, no, it's fantastic. And probably, I mean, it's a real tribute to the teachers in the school and to the fact that they teach Sounds-Write with tremendous fidelity. So in the school, as in so many Sounds-Write schools, the students who also take the Young's Parallel Spelling Tests, as I've said, and again, of the 23 children that they tested, 20 had a spelling age above their chronological age, and of the other three, one was absent during the test, one had a spelling age equal to their chronological age, and one was only a month behind their chronological age, which is actually insignificant, really, statistically. Your listeners might want to know why, as well, we use the Parallel Spelling Tests. And we think that spelling is actually, it gives you a much more accurate indication of what children's literacy is like than simply using a reading test. And that's because reading, of course, is recognition memory. You have the cue in front of you. Whereas in terms of spelling, they have to retrieve the sounds of the word and represent them onto paper from their memory and that's very much more difficult. McGuinness used to say that it's a much deeper kind of memory.

Laura:  17:36
That makes sense. And it's like when you're learning a language, it's so much easier to read it or listen and understand rather than to speak or reproduce the language yourself.

John:  17:49
Yeah, that's right. It's the difference between receptive and expressive language, of course, isn't it? Yeah.

Laura:  17:55
So I wonder, this kind of lack of research on the topic. I think there's not that many approaches to teaching phonics that use the sound-to-print approach, and so maybe that's one of the reasons for the lack of research on it, which hopefully will change in the near future as people start to adopt sound-to-print approaches more. But, yeah, in the meantime, we do have our book of case studies that you can read in the show notes. But hopefully more people will come on board, especially universities.

John:  18:32
We hope so. We hope so.

Laura:  18:35
Yeah, hopefully. Because the distinction, it's a distinction that probably not a lot of people notice. And I think, especially when I first kind of started to get involved with Sounds-Write, it's not something that I thought about often, the distinction between sound-to-print and print-to-sound. But it is actually integral to the conceptual understanding of how language works, and if we don't understand that written language is a visual representation of spoken language, then we're missing a key part of it. And so it is really important that teachers understand that before they start teaching written language. So, you said earlier that there are roughly 44 sounds in spoken language, and we must map the different spellings onto these sounds rather than the other way around.

John:  19:30
Yeah, and as you mentioned, conceptual understanding there, and what we think is incredibly important is that you develop a conceptual framework which explains how the alphabet code works. And what I'd just like to talk about very briefly is that I'd like to give you the four, what we call the golden nuggets of conceptual understanding that underpin our whole approach to phonics. So a genuine sound-to-print approach teaches that we read and spell sounds one at a time, from left to right, across the page. I've talked about this on the podcast before, but it's worth making this clear to children in school, of course, especially if those children are learning to read a language script with a different orientation. The second thing is that it teaches that we can spell all the sounds in the language with one, two, three, or four letters. And perhaps we can come back to that again a bit later, too. It teaches that we can spell sounds in different ways. So learning the approximately 175 common spellings of the 44 sounds is what makes spelling so difficult for many people, why they have so much trouble learning to spell properly. And finally, it teaches that many spellings can represent more than one sound. I gave a more in depth description of these concepts in our podcast episode, 'What Is Sounds-Write', of course, that we recorded previously. These four points, though, offer a complete conceptual framework that is supplemented, of course, by teaching the necessary skills of segmenting, blending, and phoneme manipulation, and by teaching the code, as I said before, from simple to more complex.

Laura:  21:20
So, when we look at it like this and anchor the teaching of the code in our understanding that the 44 sounds of our language can be spelt in a number of different ways in these 175 common spellings, and others, of course, we can do away with a lot of the lingo that some phonics programmes use. Right?

John:  21:42
Yeah, we can. So, concepts like, for instance, silent letters become completely redundant. If you could map every spelling onto the sounds in a word. We don't need to use the idea of silent letters at all. All letters are silent. In fact, I commonly make the joke with children by putting my ear to the board and showing them that letters don't make sounds. They don't say sounds. It's us that say and make sounds. So whether the word is 'well', in which we spell the sound /l/ with two letters, or whether the word is 'pterodactyl', in which we spell the < pt > as a two-letter spelling alternative for /t/, or even perhaps 'psychology', in which the sound /s/ is spelt with two letters, the 'ps' spelling. There are no so-called silent letters, we don't need them.

Laura:  22:34
Of course. So, I've got one last question for you on this topic. When children start to read, and they're kind of being slowly introduced to one -letter, one-sound spellings such as the < m > in 'mat', they'll inevitably come across words that don't fit into this simple model. We've talked about this before on the podcast, I think, in our decodable readers episode, where we talked about the fact that children aren't just in this perfect phonics world, in which all of the words that they're exposed to fit into the sounds that they've already been taught. Right? So they are going to come across words like 'said', for example, in which the < ai > represents the sound /e/. So how does the sound-to-print model deal with high frequency words like these that children will inevitably come up against even quite early on, when they're just beginning to learn to read and write.

John:  23:33
Well, one of the things that they'll do immediately with the word 'said', if they're writing it, is they'd probably spell it < sed >, and, of course, we will see that all the time. What the teacher needs to do if they haven't taught that spelling for /e/ yet. And it's quite likely that they wouldn't be taught that until the second year of school. They'd simply write it on a board and say, 'this is the way we spell /e/ in this word. That is a way of spelling /e/ in the way that you've spelt it. But in this word, we need this spelling of /e/'. And already we're pointing them to the idea that there's more than one way of spelling a sound, and that's a very important concept, because after only a very short time, many children in schools are wanting to spell words that are more complex and they might be saying to their teacher, 'Miss, how do you spell the /ee/ in 'stream'?' And the teacher can say, either, 'It's the < ea > spelling', if they already know letter names, or they can write it on the board and say, 'This is the way we spell /ee/. It's two letters, but it's just one sound.' Very simple. But there is a clear dividing line between sound-to-print and print-to-sound approaches. Print-to-sound approaches mostly teach high frequency words as words that must be learnt by sight, meaning that the child is encouraged to remember the word. And then you set up a false dichotomy, I suppose, if you like. You're teaching them, on the one hand, of course, that all words are comprised of sounds and so on, and then you're teaching them a different method if you're asking them to try and remember words. So, they are taught to look at the shape of the word on the page and to memorise it, which I think for many children, that can be really damaging. Now, by contrast, a sound-to-print approach, high frequency words are treated as just exactly what it says on the tin. They're frequently occurring words found in children's books and other types of text, and many of these are very easily decodable, even in the early stages of learning to read and write. For example, 'mum' is a high frequency word. But once the sound-spelling correspondences for /m/ /u/ /m/ have been taught and the children have good segmenting and blending skills, the word is easily accessed by decoding. And in fact, I was looking at some of these high frequency words before we met today. And many of those words, like 'can' 'this' 'them' 'help' and so on, can be easily decoded in the earliest stages of learning to read, and yet you see them on people's walls all the time as words that children have to try and memorise from their shape or whatever. So, other high frequency words containing sound-spelling correspondences, the children haven't yet been taught. They can either wait until the teacher introduces them, of course, or the teacher can teach the yet to be taught sound-spelling correspondence. So, for example, in the example that you gave, 'said', it's a very commonly used high frequency word. The spellings < s > and < d > are commonly taught, of course, so it's likely that children are going to know those very early on. All the teacher has to do here is to point to the spelling, in the case of reading, of course, and tell the student, 'This is /e/, say /e/ here'. And they can also say that it's two letters, one sound, if they want to add that information. And the student says the sound /s/ /e/ /d/ 'said'. They're starting to be introduced to the idea, then, even before it's being taught that one sound can be represented by more than one letter, in this case, it's a two-letter spelling. So, asking students to memorise a whole long list of high frequency words is both extremely challenging for anyone, and it runs counter to teaching students that all words are decodable at some stage of their teaching instruction.

Laura:  27:56
Okay, thanks. So, it's been really good to hear more about the sound-to-print approach from you. And just quickly, before we finish off, I have a little sneak peek only for people listening to this episode. So, some of you might have attended our online symposium last year, we'll be holding our second online Sounds-Write symposium on the topic of sound-to-print next year. So stay tuned to hear more about that when the time comes. We're really excited about it. The one last year was a pretty big success, people really enjoyed it, so I can't wait for the one next year. It's very exciting.

John:  28:38
Yeah. And I think we're already thinking about some of the people that we can invite onto the symposium, aren't we?

Laura:  28:45
Yeah. Lots of fantastic people to hear from, I think. Yeah. Well, this has been great, John, thank you so much.

John:  28:55
Thank you. Thank you once again for inviting me, Laura, and I'm looking forward to the next one.

Laura:  29:02
Great. All right, thanks for listening. Bye.