Thursday, December 08, 2016

The know-nothing world of the academic opposition to phonics

If you want to know why so many Australian (and English) academics are so strongly opposed to a Phonics Screening Check, which really is a fig leaf for their hostility to phonics teaching itself, it is that, at bottom, they don’t understand the relationship between the sounds of the language and the writing system itself.

Five thousand years ago people were struggling to create a means by which nascent city states might be able to record everyday business.

For urban societies to be able to work, there have to be systems in place to enable them to function smoothly. Cities need legal, tax and business systems, which, in turn, require an accurate method of recording transactions: legal systems, for example, cannot rely on word of mouth. For a legal system to work, judgements have to be made and recorded so that they may be referred to in the future.

This is what made writing the basis for urban societies. In the words of Peter Daniels, ‘All humans speak; only humans in civilizations write*, so speech is primary, and writing is secondary: writing underpins the culture of cities’.

The reason initial attempts to record information, such as pictography, failed is that they were unable to represent the complexities of language: abstractions, the ambiguities, the intricacies of the tense system, and so on. But there was another and more elemental logic for the switch to the phoneticisation of writing: the limitations of human memory. People couldn’t remember more than a few thousand pictograms.

Phoneticisation liberated humanity from memory overload through the invention of the symbolic representation of sounds in language. There are forty-four or so sounds in the English language and roughly 175 symbols/spellings for representing them. The 175 or so common spellings are capable of representing any word and any new word in the language. In other words, it is extraordinarily generative. By the standards of most alphabetic languages, this is a lot to learn. The Spanish language, for instance, contains around twenty-two to twenty-four sounds and these can be represented by around thirty to thirty five spellings. For this reason, Spanish is relatively easy to learn; English, on the other hand and for reasons of its history, is more complex, less transparent, harder to learn. It makes sense then that it takes longer to teach.

What also makes sense is that because the English writing system is complex, teachers must be properly trained. Evidence that this is not happening abounds. It isn’t good enough to set up a screening check alone and hope that somehow this will ensure that phonics is taught well. It won’t be if teachers are not taught how our writing system functions and the steps required to teach it to mastery level.

So, apart from knowing which sounds are represented by which spellings, what are the levels of complexity which teachers and particularly university lecturers need to understand about the code? They are these:
  • A sound can be spelled with one, two three or four letters (mat, rain, night, through)
  • Sounds can be spelled in multiple ways (rain, play, great, gate, they, eight, vein, David)
  • Many spellings represent more than one sound. (hot, no, monkey, do)

I see large numbers of examples that disclose how many university teachers lack of understanding of these concepts. Take example three. The other day – as is typical in lines of questioning such as this – someone tweeted a demand to know how a child might deal with the word ‘read’, the assumption being that the recipient of the question wouldn’t know how to answer. As readers of this blog will know, the word can be read in two ways: as /r/ /ee/ /d/ to rhyme with ‘need’, or as /r/ /e/ /d/ (the simple past tense) to rhyme with ‘bed’. What could be easier! When I responded in this way  rather politely, I thought  I was blocked! This is amazing behaviour from someone allegedly interested in debating the pros and cons of the possible introduction of a phonics screening check, Australian style.

It is obvious from the sort of interventions they make that many academics have never tested their thinking by teaching a phonics lesson to young children, nor, from the level of ignorance they display, do have know how the code works or how best to teach it. Anthony Radice asks in his latest blogpost ‘Why don’t progressives want to debate?’ In the main, the answer, Anthony, is that they throw around half-baked generalisations and the moment you challenge them on the detail, they collapse like a deflated balloon.

Here’s my challenge to the ‘progressives’: set up an RCT, pitting whole language, traditional (graphemic) phonics and linguistic phonics (a la Sounds-Write) against one another. Any time, any place, anywhere! How the challenge will be denounced - but they won’t take it up!

*Daniels is not making a value judgement here. 'Civilisation' is meant in the archaeological sense and refers to the life and culture of cities.


Dick Schutz said...

The widespread ignorance of the history and substance of the English Alphabetic Code is indeed astounding. Just as most people don't understand that they are speaking prose, they don't understand that they are writing speech.

It seems to me though that you overly discount the importance and impact of an Alphabetic Code [Phonics] Screening Check. The data that have all the merits of an RCT are already "in" through the experience in England with such a measure over the last five years. Some classes and schools are now teaching "all" their kids how to handle the Code by the end of Yr 2, and some by the end of Yr 1. You've reported on some of these schools who use Sounds-Write, and there our other schools "out there" who are getting equally good results using other programmes.

The thing is, no one has yet analyzed the data at the school level. The analysis at the national level indicate gains over the years, and the "Local Authority" level analysis indicate wide variability. Although by statute all are teaching "systematic synthetic phonics," the variability can only be clarified by further analysis at the school level. The population of schools and LEAs is sufficiently large that random samples can be drawn ad infinitum to confirm any uncertainty in what the findings turn up--not only on the differences in performance but in the effect of the "cut score." (The present "cut score" of 36 is not unreasonable, but it's arbitrary. Anyone who "can read" can easily read all 40 items. The "pass" score of 36 may well be slipping some kids through the cracks--we don't now know.

Handling the Alphabetic Code [Phonics] is NOT "one factor" in reading; it is THE factor in reading. There are indeed work-arounds, and indeed other linguistic considerations involved, but as you sketch here in your brief summary of the history and complexity/simplicity, the Code is the link that makes written language possible.

Any child who can speak in whole sentences and participate in everyday conversation has the personal prerequisites to be taught how to handle the Code=to read. It's not in the kid, it's in the instruction. Allwegotta do is to look carefully at the instruction, and the Screening Check is a good instrument for making the probe.

John said...

Hi Dick and thanks for your comment (again) :)
Of the importance and impact of the screening check, I'm in no doubt. However, I am suspicious that, given the limitations of the test in terms of testing the code knowledge of the children, many teachers are whacking the kids through lots of last minute practice - not in itself a bad thing - only to drop phonics altogether at the end of Y1. I think that if the connection between the sounds we speak and the different ways of spelling those sounds is not firmly established, for the kids in that long tail of underachievement, things may fall apart later.
Based on what I've seen of schools teaching Sounds-Write (pretty much) as we would recommend, I strongly believe that Y2 (Year 3 of formal schooling) lends itself to enabling children to go very much further than current expectations would ever suppose. Teaching children of seven to be able to read and write words like 'subdermatoglyphic' and 'paucibacillary' (in context, of course!) at this level sets them up for being able to concentrate on reading for meaning almost exclusively.
We are a hundred percent agreed on the fact that if 'Any child who can speak in whole sentences and participate in everyday conversation has the personal prerequisites to be taught how to handle the Code=to read. It's not in the kid, it's in the instruction.'
I liked your contribution to Pamela Snow's blog and look forward to your interventions on other blogs. In fact, the Australian blogosphere and Twittersphere are going to make interesting reading in the coming months and years.
The 'pass' score for the PSC is 32, btw.
Best regards,

Dick Schutz said...

Yes, EdLand is a strange country. A "Screening Check" is not a "test." But "screening" is an unfamiliar practice in EdLand, so teachers convert the unfamiliar to the familiar. The DfE encouraged the glitch by establishing a "pass score" for the Check. (Thanks or correcting me on 32. That was a Freudian slip, but it supports your point that some kids will "slip later."

"Whacking the kids through lots of last minute practice" makes no sense whatsoever--unless you believe that the aim is to get kids to "pass the test." It's not hard to determine if a kid "can read." The kid knows; other kids know; the teacher knows. The hard thing to know is what to do if the kid can't read all 40 items on the Check. "Intervention" doesn't do the trick. All instruction is "intervention." It's in the instruction--the instruction that has been provided in the past and that will be provided in the future. The unintended flaws in the past can be untaught, but the "remediation" should be in the instruction, not in the kid. The kid is OK; the instruction isn't.

Yes, it will be interesting to see if AU learns anything from the UK experience. So far, not. But it's early in the game.