Clarifying the Code

WHO DECIDES WHAT THE ALPHABETIC CODE IS?
 

The English alphabetic code isn't whatever a phonics author decides it is

Search online for the English alphabetic code and you will find charts, tables, lists of GPCs and explanations that appear authoritative.

You may be told there are 44 phonemes.

You may be given a table showing “the code”.

You may be told that a particular letter or grapheme always represents particular phonemes.

You may be given a simplified phoneme notation created by the author.

You may even be given a test with predetermined “correct” answers to establish whether you understand the code.

But there is a fundamental question that needs to come first:

Who decided that this is the correct analysis?

And, perhaps more importantly:

How was it calculated?

The English alphabetic code is not a list invented by a phonics programme, author, publisher or training provider.

It describes relationships between spoken English and written English.

That means any attempt to map the code has to start with the speech being represented.

And speech varies.

“There are 44 phonemes in English”

This is one of the most familiar statements in phonics.

It is useful as an approximation for describing a particular reference accent.

It should not be mistaken for a universal inventory of the sounds every English speaker uses.

Even the Department for Education qualifies its description of English as having:

 

“44 or so sounds (phonemes)”

The DfE also explicitly recognises pronunciation variation and acknowledges that the alphabetic code can be analysed in different ways.

Source: Department for Education, The Reading Framework: teaching the foundations of literacy

https://www.gov.uk/government/publications/the-reading-framework-teaching-the-foundations-of-literacy

The problem isn't using a standardised phoneme inventory for instruction.

A shared reference can be extremely useful.

The problem arises when a reference system is treated as though it determines what somebody actually says.

The speaker comes before the symbols

A phonetic or phonemic symbol should describe a sound.

The symbol doesn't determine which sound the speaker produced.

That distinction matters.

Suppose a teacher says a word naturally in their own accent.

If their pronunciation differs from a predetermined answer, there are at least two possibilities:

the teacher has analysed their speech incorrectly or the predetermined transcription doesn't represent the teacher's pronunciation.
 

Those are not the same thing.

Yet a test based on one fixed list of symbols and one predetermined answer can easily conflate them.

Before asking whether somebody has selected the “correct” symbol, we need to know:

What sound did they actually produce?

“Write the sounds you actually hear”

That sounds straightforward.

But there is a methodological problem.

A highly literate adult already knows how the word is spelled.

Research has demonstrated that orthographic knowledge can influence how readers consciously conceptualise the phonemic structure of spoken words.

So when a teacher looks at a written word and is asked to identify its sounds, we cannot automatically assume the answer has arisen purely from listening to speech.

They may be influenced by the letters they can see.

They may recognise familiar graphemes.

They may retrieve GPCs they have previously been taught.

They may report the phonemes they expect those graphemes to represent.

That is why phonological analysis and orthographic analysis need to be distinguishable.

Source: Ehri, L. C., & Wilce, L. S. (1980). The influence of orthography on readers' conceptualization of the phonemic structure of words. Applied Psycholinguistics, 1, 371–385.

https://doi.org/10.1017/S0142716400009802

And what if the symbols themselves look like spellings?

There is another complication.

Sometimes systems designed to avoid the complexity of the International Phonetic Alphabet replace IPA symbols with familiar letters or letter combinations.

It may appear easier.

But consider what is happening.

The teacher is being asked to think about speech independently of spelling, while the notation being used to represent those speech sounds may itself look like English spelling.

That raises an important question:

Are we separating phonological knowledge from orthographic knowledge, or building orthography into the way phonology is represented?

This matters particularly if the aim is to find out whether teachers can identify the constituent sounds in speech without allowing their knowledge of letters to determine what they think they hear.

“Schwa isn't really a separate sound”

This is another example of why a personal analysis shouldn't automatically become an instructional standard.

Whether a sound appears distinct “to my ear” isn't sufficient grounds for establishing a reference analysis for teachers.

Schwa /ə/ is an established phonetic symbol used to represent a central vowel sound and is highly relevant to the pronunciation of unstressed syllables in English.

The DfE itself discusses schwa in the Reading Framework.

Source: Department for Education, The Reading Framework: teaching the foundations of literacy

https://www.gov.uk/government/publications/the-reading-framework-teaching-the-foundations-of-literacy

This doesn't mean every unstressed vowel produced by every English speaker must automatically be transcribed /ə/.

It means the decision should be based on the speech being analysed and an explicit reference framework, rather than what an individual author believes they hear.

Be careful with “always”

Statements such as:

<qu> always represents /k/ + /w/

or

<x> always represents /k/ + /s/

make attractive teaching rules.

They are also easy to disprove.

Consider:

quiche

quay

unique

xylophone

exam

anxious

The pronunciation and mapping depend on the word.

English orthography is complex precisely because grapheme–phoneme relationships are not universally one-to-one and invariant.

A useful pattern is not necessarily a universal rule.

What counts as a grapheme?

This question becomes even more interesting when a system defines a grapheme as representing one phoneme and then encounters written units that appear to correspond to two phonemes.

Are they no longer graphemes?

Or does the framework need to accommodate a more complex relationship?

The DfE explicitly acknowledges this issue:

 

“The number of graphemes in a word usually corresponds to the number of phonemes…”

Usually.

It then recognises exceptional cases in which one grapheme represents two phonemes.

Source: Department for Education, The Reading Framework: teaching the foundations of literacy

https://www.gov.uk/government/publications/the-reading-framework-teaching-the-foundations-of-literacy

So even seemingly basic questions such as:

How many phonemes?

and

How many graphemes?

require an explicit method of analysis.

They cannot always be answered by applying a memorised rule.

A code chart is an analysis of the code

This distinction is fundamental.

A chart headed:

THE ENGLISH ALPHABETIC CODE

can easily create the impression that the chart is the code.

It isn't.

It is somebody's representation and analysis of relationships within the code.

The DfE makes this particularly clear in Appendix 5 of the Reading Framework:

 

“The chart is not definitive because the alphabetic code can be analysed in various ways.”

Source: Department for Education, The Reading Framework: teaching the foundations of literacy

https://www.gov.uk/government/publications/the-reading-framework-teaching-the-foundations-of-literacy

That one sentence has enormous implications.

If the code can be analysed in various ways, teachers need more than a chart.

They need to understand how to undertake the analysis.

A finite GPC list isn't the whole English alphabetic code

Another common approach is to select the correspondences considered necessary to get children reading independently and omit rarer correspondences.

There can be perfectly legitimate instructional reasons for doing that.

But the resulting list should not then be confused with the English alphabetic code itself.

Children encounter words outside instructional lists.

Teachers encounter them throughout the day.

A child may want to read or spell a word containing a correspondence that isn't on the programme's chart.

That correspondence hasn't ceased to exist because it wasn't selected for explicit teaching.

This distinction matters:

A teaching sequence is not the code.

A programme's GPC inventory is not the code.

A chart is not the code.

They are instructional representations and selections from a much larger system.

Reading and spelling are bidirectional

One thing many accounts of the alphabetic code get absolutely right is that reading and spelling involve
movement in opposite directions.

For reading:

GRAPHEMES → PHONEMES

For spelling:

PHONEMES → GRAPHEMES

The Australian Government's Literacy Hub describes mapping words in early reading and spelling as involving identifying the phonemes heard, identifying how many there are and representing those phonemes with letters.

It also describes alphabetic-code knowledge as connecting how a word is said, how it is spelled, and what it means.

Source: Australian Government, Literacy Hub, Introduction to SSP: Module 3

https://www.literacyhub.edu.au/video-assets/introduction-to-ssp-module-3-slide-outlines/

The DfE's Writing Framework similarly states that children learning to spell first identify the constituent phonemes in spoken words and then write the graphemes representing them.

Source: Department for Education, The Writing Framework

https://www.gov.uk/government/publications/the-writing-framework

This is why teacher knowledge needs to be bidirectional too.

Knowing a phonics programme isn't the same as knowing how to calculate the code

A teacher can learn:

this grapheme represents this phoneme

and

this phoneme can be represented by these graphemes

without necessarily being able to analyse an unfamiliar example independently.

That distinction matters enormously.

What happens when the word isn't on the chart?

What happens when the correspondence wasn't taught?

What happens when a child pronounces the word differently from the teacher?

What happens when the teacher's own pronunciation differs from the instructional reference?

What happens when the spelling tempts the teacher to identify a phoneme they didn't actually produce?

What happens when two apparently knowledgeable teachers map the same word differently?

Telling teachers the answer doesn't establish whether they understand how that answer was reached.

The question isn't “Have you learned this code chart?”

It is:

Can you independently map the relationship between the speech and print?

That means being able to ask:

What word was actually said?

Which constituent phonemes were actually produced?

How many are there?

Which letters constitute the graphemes in this written word?

Which phoneme value does each grapheme represent in this word?

Does the analysis change if the pronunciation changes?

And, where an instructional reference pronunciation is required:

Can the teacher distinguish that reference pronunciation from their own?

Why does this matter in classrooms?

Because children don't spend the whole school day reading and spelling words selected by a phonics programme.

They write:

sed

wont

enee

woz

bickoz

A teacher has to decide what those attempts reveal.

Which phonemes has the child perceived?

Which graphemes have they selected?

Which relationships are already plausible?

Where is the mismatch?

Is it actually a mismatch given the child's pronunciation?

What does the child appear to understand?

And:

What should the teacher do next?

A programme cannot make that judgement for every word every child encounters throughout the day.

The teacher needs the underlying knowledge to do it.

So who tests whether teachers have that knowledge?

This is the issue.

Teachers can complete phonics training.

They can learn programme terminology.

They can follow a programme sequence.

They can memorise a GPC chart.

They can learn somebody else's analysis of English words.

None of those things, by themselves, establishes that a teacher can independently:

identify the constituent phonemes in speech distinguish their own pronunciation from an instructional reference identify the graphemes in written words map speech and print bidirectionally

and explain the decisions underlying that mapping.

Previous research gives us good reason not to assume that knowledge.

Stainthorp's study of graduate trainee primary teachers in England found that they:

 

“did not use a consistent system for segmenting words into component sounds.”

Source: Stainthorp, R. (2004). W(h)ither Phonological Awareness? Literate trainee teachers' lack of stable knowledge about the sound structure of words. Educational Psychology, 24(6), 753–765.

https://doi.org/10.1080/0144341042000271728

Connelly's UK research similarly found difficulties when adults were required to relate phonemes to the letters and letter groups representing them.

Source: Connelly, V. (2002). Graphophonemic awareness in adults after instruction in phonic generalisations. Learning and Instruction, 12(6), 627–649.

https://doi.org/10.1016/S0959-4752(01)00034-2

We don't need another definitive-looking code chart

We need to be able to explain how the analysis was reached.

We need to distinguish:

speech from spelling

actual pronunciation from reference pronunciation

phonemic awareness from orthographic knowledge

the English alphabetic code from a programme's selected teaching sequence

and

demonstrated mapping knowledge from familiarity with somebody else's answers.

A reference framework is necessary.

Instructional simplification can be useful.

Code charts can be useful.

Phonics programmes can be useful.

But none should become unquestionable simply because somebody has labelled their analysis:

“THE CODE”

The more important question is:

Can the analysis be justified?

And for teachers:

Can you calculate it yourself?

Because ultimately, the English alphabetic code isn't whatever an author, programme or chart says it is.

It is the relationship between the speech being represented and the written system representing it.

And if teachers are expected to help children navigate that relationship throughout the day, we need to know whether they have the underlying knowledge to do so.