Friday, May 29, 2015

Scripps Spelling Bee 2015

I haven’t posted on the Scripps Spelling Bee event for some time now - since 2011 in fact. As in so many previous competitions, this was a nail-biter to the end and finished in a tie. It's the second tied win in two consecutive years. Prior to 2014, there hadn’t been a tie for fifty-two years. In the end, the two demon spellers were Gokul Venkatachalam from Missouri and Vanya Shivashankar from Kansas, beating off 283 other, determined competitors from the length and breadth of the USA. 
Thanks to FallingFifth.com
for 'The Alphaphones'

As soon as I looked at the names of this year’s winners, I thought there was something familiar about Vanya’s family name. And there was! Her sister Kavya won the event in 2009 with the word 'laodicean', about which I wrote at the time here.

At the end of 2015’s gruelling competition, Shivashankar correctly spelled ‘scherenschnitte’ meaning ‘scissor cuts’. Easy if you happen to speak German! Venkatachalam sealed his place on the winner’s rostrum with ‘nunatak’, an Inuit word meaning an exposed, rocky part of a ridge surrounded by glacial ice. 

If you’re wondering how the Spelling Bee could end up in a draw, here’s how Scripps explained it last year:
Once there are three spellers left in a round, the next round begins with a 25-word list. Ordinarily, a winner is declared if one speller misspells and the remaining speller correctly spells two words in a row. If no winner is declared before the list has been exhausted—or there are not enough words left for two consecutive spellings—co-champions are announced.

These days the spellings are unquestionably tougher than when an eleven-year-old Frank Neuhauser, the son of a Kentucky stone mason, won the competition in 1925 with the word ‘gladiolus’. In those days the prize was $500 in gold, a bicycle and a visit to the White House to meet the then president Calvin Coolidge. Today’s champions each take home an enormous trophy and $35,000 in cash. And, while I don’t expect this year’s winners will get the ticker tape reception and the enthusiastic crowds bearing bouquets of gladioli that Frank received, it’s certain that there is no lessening in the intensity of interest many people in the United States feel for this annual, fascinating encounter.

Thursday, May 21, 2015

Our wonderful testimonial from Jan Hilary of St George's C of E Primary School

Sounds-Write is very proud to be able to make public this testimonial from Mrs Janet Hilary, headteacher of St George's Church of England Primary School in London.

I recommend Sounds-Write to every teacher and school leader I meet. At St. George’s, where deprivation levels are extremely high, we achieve consistently outstanding results in all phases. Our Phonics Screening Check has been 96%+ every year.
Sounds-Write is a brilliant phonics programme for all pupils from Nursery but it also enables us to teach pupils with no English, or with specific learning difficulties so that no gaps exist in performance within year groups. It is a superb scheme for teaching polysyllabic spelling right through to Year 6.
The training is top quality and all staff use consistent methods to teach phonics, reading and spelling effectively. Skills are taught explicitly and pupils demonstrate confidence and success from the outset. Progress is rapid and the quality of writing from Reception onwards astonishes visitors from other schools. 

Janet Hilary, Headteacher and National Leader of Education

St George in her new incarnation


Wednesday, May 20, 2015

Sound to print: the appliance of science

‘Phonics,’ wrote Diane McGuinness, in her superb book Early Reading Instruction  ‘is a problematic word.’ Never was there a truer thing said! Why? Because ‘phonics’ is an umbrella term for all kinds of approaches, some good and some-fair-to-middling-grim.  According to McGuinness, the ‘classification is unsatisfactory because it does not identify the critical difference in logic between programs that teach the code backward from print to sound, and those that teach it forward from sound to print (linguistic phonics)’.

In McGuinness’s summation, the problem is deeply embedded in the orientation of the particular programme: is its orientation from print to sound, or from sound to print? In the former, the orientation is primarily visual; in the latter it is phonemic.

In this post I want to point to a fundamental, logical inconsistency in the approaches used by many specialists in the teaching of literacy here in the United Kingdom, in Australia and in the USA. The inconsistency is this: many teach some version of ‘traditional phonics’, which is to say that their orientation is ultimately graphemic or visual.

What is meant here by 'graphemic’ is that teaching goes from print to sound, rather than sound to print, as it does in linguistic phonics approaches. Why, you may wonder, is this an important issue? The simple answer is that, if our alphabet code were as straightforward and simple as it is with, say, languages such as Spanish or Italian, it wouldn’t be important at all.

Simple codes are relatively unambiguous: sounds are mostly represented by single-letter spellings and spellings represent one sound. So, teaching simple codes is easy and with languages like Spanish and Italian, which have such transparent codes, it matters much less if the orientation of the teaching is graphemic or phonemic. In other words, it’s hard to get the teaching wrong! However, in cases in which the code is much more complex, as is the case in English, it matters a lot!

Now, here’s the rub! Most phonic programmes start by teaching the simple one-letter spellings to one sound. This often masks whether they are truly phonic or merely graphemic because, at this early stage of teaching, it doesn’t make too much difference which way round the code is taught. The crunch comes when the complexities of the code are introduced. It is at this point that graphemic programmes suddenly become caught up in their own contradictions and have to resort to all sorts of nonsensical and, frankly, illogical explanations.

Here’s a perfect example of addled thinking. Don’t get me wrong, in almost everything I read by the man, I think Daniel Willingham is superb, but, when it comes to the teaching of phonic decoding, he is no teacher. The example is from his latest book Raising Kids Who Read. On page 10, he writes the following: ‘When gh appears at the start of a word, it’s pronounced as hard g (e.g. GHASTLY, GHOST).’ Okay, so far, so good!  But then, he writes, ‘In the middle of a word, it’s silent (e.g. DAUGHTER, TAUGHT)…’

What this tells us is that: a) that Willingham can’t segment words into sounds and then link the sounds to the spellings. [There are four sounds in ‘daughter’, /d/ /or/ (or /o/ in US English) /t/ /schwa/; and, three in ‘taught’.] b) he doesn’t 'get' that spellings are representations of sounds, or that spellings can be represented by one, two, three or four letters. As I've stated many times before on this blog, letters don’t ‘make’ or ‘say’ sounds. We do! There are no silent letters because ALL letters (I prefer ‘spellings’ a much more accurate term!) are silent. The only constant in the spelling system is the sounds.

This is extremely important precisely because as soon as the code moves from simple to more complex, unless teaching is anchored in the sounds of the language, confusion can (and does) quickly set in. For example, if the single-letter spelling o can be /o/ in ‘hot’, /oe/ in ‘go’, /oo/ in ‘do’ as well as /u/ in ‘mother’ and a teacher tells a young child, who is likely take such things very literally, that the letter o can ‘say’/’make’/o/, but then it can also ‘make’/oe/, and /oo/ and /u/, the child is almost certainly going to wonder whether, if it can ‘say’ or ‘make’ these sounds, it can ‘say’ or ‘make’ any sound. A ‘magical’ code that appears to operate like this is not a code and is unworkable.

On the other hand, if the child is taught to understand that the sounds s/he utters in words and sentences all the time are written/spelt by the spellings we introduce from the start, the code is always anchored in the forty-four sounds in speech. It is also makes perfect sense psychologically. Then, we don’t need ‘curly k’ and ‘kicking k’, ‘hard gee' or 'soft gee’, ‘short vowel sounds and long vowel sounds’. All of this becomes completely redundant. In fact, young, literal-minded children often have no idea what ‘hard’ and ‘soft’ sounds are. Nor indeed do they understand a lot of the completely gratuitous verbiage that goes with much of the teaching they get.

As for rules, I have never met a teacher in my life who knows what the rules are in their entirety, never mind the exceptions to the rules. In Early Reading Instruction, McGuinness explains why they are superfluous, but that’s the subject of another posting.


Saturday, May 16, 2015

Why doesn’t The Literacy Blog advocate the use of flash cards?

There are a number of reasons why I think that phonic programmes that advocate the use of flash cards are barking up the wrong tree. The use of flash cards is a legacy of old fashioned phonics programmes, which emphasise the visual/graphemic at the expense of the aural.

Presenting children with flash cards, which are by the very way in which they are presented, decontextualised, will 'work' only to a certain extent for some children, particularly those with some prior learning and for those with strong visual memories. For these children, some would, as the number of new words increases, reach the limits of their visual memories fairly quickly and go on to struggle; others would begin to connect spellings and sounds implicitly and go on to teach themselves to read. But for all children, I would argue, flash cards are a waste of time and even send children down the wrong path, especially when, as I believe, there’s a better way of teaching reading and spelling.


Both John Hattie and Daniel Willingham argue that learning is much more effective if it is contextualised. Not only are flash cards decontextualised in that they are not presented in the context of words, they are, and more importantly, also decontextualised in terms of how they function. In other words, they are not linked to where in relation to other sound-spelling correspondences they appear in words.

What are these dependencies?
Ø  Is the spelling comprised of more than one letter?
Ø  Are there alternative ways of spelling a particular sound?
Ø  Can a particular spelling represent more than one sound? For example, if the single letter spelling a is presented to children, it will invariably be presented as /a/. Although it is sensible to present it as /a/ to begin with, whether it is /a/ or not will be heavily dependent on what precedes and what follows it: if the spelling a is preceded by the sound /w/, spelt as w or u, the chances are it will represent the sound /o/, as is ‘swan’ or ‘squash’. In other words, context is vital.

The idea that a spelling can represent more than one sound is not best taught by rote activities such as flash cards. It is taught through a combination of carefully structured and presented activities that teach conceptual understanding of how the writing system works in relation to the sounds of the language, the precise correspondences in context (!), and the skills needed to be able to use this understanding and factual knowledge fluently. In other words, the fundamental flaw in the idea behind using flash cards is that there is no coherent, conceptual backbone to the approach.

Spellings, which are being presented on flash cards, are very often not grouped according to whether they represent single letter spellings, or two-, or three-, or four-letter spellings of sounds [sh i p, l igh t, eigh t). They aren’t grouped to be anchored in individual sounds so that it is made clear that there are different ways to spell the target sounds [sh e, s ea, m ee t, k ey, h a pp y, f ie l d]. Moreover, the fact that many spellings represent more than one sound is more often than not completely ignored [h o t, g o, d o, m o th er].

Another problem with the use of flash cards is that the approach relies heavily on paired-associate learning: associating one thing with another. Associating random isolated sounds with random isolated spellings is very hard to remember. Try it with an invented, symbolic alphabet. It is extremely difficult! And, if adults find it hard, how much harder is it for young children?

Linguistic phonics programmes, such as Sounds-Write, present a completely different alternative and one that is underpinned by a clear and coherent rationale supported by cognitive psychology. We begin with word building. Word building introduces sound-spelling correspondences in the context of whole words right from the start and it anchors them in what all children learn naturally: the sounds of their own language.

Advocates of linguistic phonics make absolutely explicit that the squiggles on the page we call spellings (comprised of letters of the alphabet) stand for the sounds in words. Even if, intellectually, young children don’t grasp this idea immediately, they begin to ‘cotton on’ very quickly as they operate the procedures of blending, segmenting and phoneme manipulation.

Our methodological approach also presents new information in small steps and from simple to gradually more complex. In other words, we build into the minds of the children a schema for sounds and spellings! It is this that helps children to remember.

Admittedly, as acknowledged by experts such as Hattie, the formation of schemata is a long process and it requires extended feedback and practice from those who know how the writing system is structured. We know that getting information into long-term memory is highly dependent on how we present the information. With adequate practice – more for some children, less for some others – it is possible to build a wholly integrated schema for all the major complexities of the alphabet code over the three years (KS1) we believe it takes to teach children to read and spell to a very high level of proficiency.

We don't need to drill children by rote. Reinforcement comes with consistent exposure in the activities of reading and writing in context.

Thursday, May 14, 2015

i.t.a: a great idea but a dismal failure

Talk to anyone today who was taught to read through i.t.a. (Initial Teaching Alphabet) and they will almost invariably tell you how they’ve never been able to spell correctly since. 
As i.t.a. was more or less abandoned in the sixties/early seventies (though it did cling on for much longer in some places), many of today’s generation of teachers will never even have heard of it except from their parents or grandparents! So why write a blog posting about it?

I’m writing about it because it did, at first sight, appear to be a great idea. At the same time, as the title of the post suggests, it was a disaster – because so many children were left floundering it its wake.

Starting with the ‘great idea’ bit, it was conceived by James Pitman, grandson of Isaac Pitman, developer of the famous shorthand system of note-taking still in use today. What James Pitman thought was not dissimilar from the ideas of Stephen Linstead, Chair of the English Spelling Society Spelling reform. Pitman thought that if he could produce a single symbol for every one of the forty-four sounds in English, children would have a simplified and very easy system to learn. Doing so would give us a writing code not unlike Italian or Spanish.

Most of the single letter spellings, the one-to-ones remained the same. So, the spelling in ‘bat’ remained the same as in our accepted orthography. Where the system differed was in many of the two-letter consonant and vowel spellings. Thus, Here’s an example: /th/ (unvoiced) in the word ‘thin’ was spelt q; /th/ (voiced) in the word ‘this’ was spelt d; the sound /ae/ as in ‘gate’ was spelt æ; and, the sound /oe/ in ‘goat’ was spelt œ.


If we had a system for spelling the forty-four sounds in English (forty-five in some accents) with one symbol, learning to read and spell would be easy. For example, the sentence ‘I have a goat’ would be written: ‘I hav a gœt’. At the time, Ladybird produced books to support the approach. To give you an idea of what they looked like, here is an example from a book called The Fisherman: 

The most obvious problem with such a system is that, at some point, the transition to our accepted orthography must be made . In the sentence 'I hav a goet.' above, the spelling of /v/ in ‘have’ is commonly spelt ve at the ends of words and the spelling of the sound /oe/ in ‘goat’ is oa. For children to make the transition, the teacher has to make explicit to children that, in English:
·      we spell sounds with one, two three or four letters
·      sounds can be spelt in multiple ways
·      many spellings represent more than one sound

The teacher also has to teach all the various common ways of spelling sounds for reading and spelling, and they need to know how to teach that many spellings represent different sounds and the skills to enable them to use this knowledge when reading and writing.

Because hardly any teachers knew how the transition to accepted orthography should be taught, many children were left struggling to work out the logic of the alphabet code. Teachers in (the then) junior schools (KS2) found themselves confronted with children writing what seemed to them like gobbledegook.

The next problem with i.t.a. was that it presented the spellings for the sounds of someone with a Received Pronunciation (RP) accent, which is not the accent of many speakers of English. So, it didn’t make sense to speakers of other varieties of English. In addition, aside from also violating the principle that it is never a good idea to teach what later needs to ‘un-taught’, no-one for a moment believed that all existing written materials should be re-written in i.t.a. This meant that after learning to read and write i.t.a., children had to be taught how the code works.

Today’s would-be spelling reformers peddle what is essentially the same line: simplify spelling and learning to read and write English will be much easier. As I’ve pointed out here, the idea is a pipe dream. Many previous attempts have been made and they all founder on the rocks of different accents of English and on establishing an agreed system of spelling the forty-four sounds in the language, including the most common vowel sound, the schwa.

There is one reason and only one reason for the spelling reformers’ confusion – instead of starting with the sounds of the language and teaching children the different ways of spelling those sounds, they start from spellings. Spellings, they seem to think, ‘make’ or ‘say’ sounds. They don’t. We are dealing with a symbolic system: spellings are symbols for sounds. Once this becomes your starting point, you have an anchor for all your subsequent teaching.

Below is the i.t.a. chart, which you'll also find on Wikipedia here.


Friday, April 17, 2015

The strange case of the word ‘yacht’

The strange case of the word ‘yacht’. This old chestnut comes up on a fairly regular basis and is cited as an example of how not all English words are decodable.

In truth, the word presents us with more of a challenge than many others. However, holding to the notion that every word incorporated into the English language is comprised of sounds and that all sounds have been assigned spellings, ‘yacht’ contains three sounds /y/ /o/ and /t/. How then can sounds be linked to spellings in a way that would enable young learners to remember how to spell it?

Well, the word came up in the round in a school I was working with a few years ago. We 'embarked' on an exercise in seeing how best to link the sounds of the word to the way it is spelled. I asked the class, a Year 4 who had been doing a good quality phonics programme since they had begun school, to work in small groups and to see what they could come up with. They proposed two choices: first, y  ach  t; and, y  a  cht. They then voted on which of the two they preferred. The answer was the second choice. [My vote went to the first!] This was, they argued logically, because the sound /o/ can be represented by the spelling a in lots and lots of words, when it follows the sound /w/. Having completed this short exercise, to implant it into their memories, they all wrote it and said all the sounds, before reading it back again. [This little activity should be repeated a week or so later.]

Of course I wouldn’t expect a young child to be able to read or spell the word without some support. And, if it came across the bows during a lesson, I’d deal with it there and then, or, if it proved to be too intrusive, save it to discuss later. This is how best to extend code knowledge in KS2/3/and 4.

Talking of it coming across the bows, where does the word come from? In actual fact, it is derived from Dutch. In early modern Dutch, a ‘jachtschip’ was a pirate ship. Readers of German might also recognise the word from ‘jӓger’ or ‘hunter’. In Dutch, the word sounds remarkably like the English word, except that the ch represents a separate sound, which sounds a bit to my ear like some Liverpudlians would say /k/ in the middle or at the ends of words: thus, /y/ /a/ /k+/ /t/.


So, what’s my point? Well, I have two actually. The first is that not only is the word decodable and encodable, it is also an example of how, even at a bit of a stretch, English is comprehensive. That is to say that it can easily incorporate pretty much any loan word from any language even when the loan word is a challenge for us to pronounce and, as a result, forces us to anglicise it. My second point is that, having analysed the word in the way suggested above, children are far more likely to remember how to spell it in the future. And, if as a teacher you find that you're stumped as to the derivation of a word, ask the children to look it up online.

Wednesday, April 01, 2015

Decoding, comprehension and muddled thinking

We know from a variety of different studies that the same regions of the brain are activated when we read as are activated in speech comprehension. As we are reading, our brains are hunting for meaning and, as long as a word is in our vocabulary, we understand it as we read. Of course, in the case of highly proficient readers, all of this happens in milliseconds and under the level of our conscious attention. For the beginning reader, things aren’t quite as straightforward.

The simple view of reading contends that efficient reading is the sum* of decoding ability and comprehension. These are the core competencies. To make sense of text, the reader must first be able to ‘lift the words off the page’, or turn spellings into sounds and blend the sounds together to produce words. Once decoding has taken place, only if the word or words are within the reader’s spoken comprehension will the message be understood, even if the reader is a super-efficient decoder. For example, I can read anything in Italian or Spanish or German but I don’t always know the meanings of the words.

It’s also true that our definition of what constitutes good comprehension will vary according to age and according to the amount of cultural background knowledge any particular reader possesses. For beginning readers (YR to Y2) and for students who can't read fluently, we should, for the most part be presenting texts that focus on practising fluency. Those texts should also, for most students, be fairly literal – as opposed to the kinds of more sophisticated, interpretive (figurative) meaning we expect of more mature readers.

This is quite different from the kinds of texts we might ourselves as teachers or carers, read to young children, which can and should contain all kinds of figurative meaning. The reason for the focus in teaching beginning reading and writing to be on literal meaning is that literal meaning is innate. For L1 speakers, literal meaning is straightforward and doesn’t need to be taught because children’s vocabularies vastly exceed what they are, as yet, able to read. So, in the beginning, teachers need to focus attention on teaching children to decode fluently.

Why should decoding be our prime concern? It’s important because to understand textual meaning, we must be able to hold sufficient amounts of text in working memory. If decoding ability is insufficiently automatic, the strain on working memory is enormous. Being unable to read fewer than about 60 words per minute is going to result in the reader being unable to remember the beginning of the sentence or paragraph by the time they get to the end. If reading speed does not increase, children won’t be able to understand more complex text.

Therefore, in the beginning of learning to read, automatising the decoding process must be our primary goal and teachers need to be aware that, aside from the issue of including large numbers of infrequently encountered words, the complexity of the text being presented should also be a major consideration. For example, syntactical complexity and the introduction of embedded relative clauses will place further strain on working memory.

In other words, the focus should be on learning to read, which subsumes a number of different aspects which I have explained a number of times before on this blog. Once children are out of the starting blocks, the kinds of texts they should be practising on should be commensurate with what they are learning in their phonics programme. And large amounts of practice needs to be available: 'readathons' and other encouragements are vitally important in improving fluency and comprehension, as well as reading for expression. To emphasise the point once again, if reading speed doesn’t increase, children won’t be able to understand what they read.

In developed countries, fluency without understanding is relatively rare. The number of children unable to understand text requiring a literal understanding is likely to be very low. In these relatively few instances, lack of understanding can be attributed to several things, among which are: poor language knowledge, especially for non-L1 speakers at home; shorter working memory capacity; intellectual problems; and, limited attention span.

Learning to read a transparent language, such as Spanish or Italian, is very much easier and enables the reader to make the transition from learning to read to reading to learn in a relatively short period of time. Leaning to read a more opaque and complex language such as English takes much more time (three years). For this reason, the training of teachers charged with teaching children in the early years needs to be done properly. Providing trainee teachers with a few hours here and there doesn’t cut it. If teachers don’t have a clear understanding of how the writing system works in relation to the sounds of the language and understand issues to do with working memory, they simply won’t be able to teach reading well and they will always be prone to being sidetracked into activities that don’t correlate to teaching reading and spelling successfully and waste time.


What’s the solution: (read my lips) train the teachers.

* When I wrote 'sum', I was writing metaphorically. If I had been quoting Gough and Tumner (1986) and referring to the formula R = D x L, I would have written 'product'. However, I think the point I'm making is clear enough.