Reading Simplified has identified a real weakness in conventional phonics instruction. Teaching one correspondence at a time, in a carefully rationed sequence that can stretch across months or years, is painfully inefficient. Children encounter ordinary written English every day, yet instruction often acts as though they should not see a spelling until it has reached its appointed place in the scope and sequence.

Reading Simplified rejects that slow drip. Instead of teaching one spelling for a sound and withholding the alternatives until later, it groups several possibilities together. A child working with /ee/, for example, might see <e>, <ee>, <y>, <e_e>, and <ea> at the same time and sort words according to those spellings.

That is certainly more efficient than pretending English contains one predictable spelling for every sound. It gives children access to more of the written language sooner and acknowledges from the beginning that a phoneme can be represented in several ways.

But the organizing principle has not changed. The sound is still the starting point, and the spellings are arranged underneath it.

Reading Simplified describes this explicitly as focusing on sounds rather than spellings because children arrive at school already knowing the sounds of their language. The instructional sequence then uses those sounds as the framework for organizing print.

This is linguistic phonics delivered faster. The slow drip has been replaced by a bigger bucket.

Knowing spoken language is not the same as consciously analyzing phonemes

Children certainly arrive at school knowing spoken language. They know thousands of words. They understand sentences and grammatical relationships. They recognize familiar expressions, make inferences from context, learn new meanings, and use language to tell stories, negotiate, joke, request, argue, and explain.

None of that means they consciously experience speech as a sequence of isolated phonemes.

Phoneme awareness requires a child to abstract away from continuous speech and deliberately identify units that are not delivered as neatly separated segments. Some children learn to do this fairly easily. For others, particularly many of the children who eventually land in intensive intervention, it is extraordinarily difficult.

Yet sound-first instruction takes this highly abstract metalinguistic task and makes it the entrance point to literacy.

For a learner with dyslexia, a receptive and/or expressive language disorder, limited phonological working memory, executive-function difficulties, or some combination of these, being shown five possible spellings for /ee/ may not feel particularly simplified. It may simply create five competing possibilities attached to a category the child already has difficulty holding and manipulating.

This kind of instruction can work. It may work quite well for many children, and it may move some children through early decoding more efficiently than a traditional phonics sequence.

It does not work for everyone.

Those of us working in SWI intervention know this because we inherit many of the children for whom it did not work. They have already had phonics. Many have had Orton-Gillingham. Some have had speech-to-print or other linguistic phonics approaches. They have blended, segmented, manipulated phonemes, sorted spellings, memorized correspondences, and completed increasingly intensive versions of the same basic instructional proposition.

Then they arrive in our practices still unable to make sense of written words.

That experience makes me very wary of claims that a reading approach can quickly resolve reading, spelling, vocabulary, and comprehension difficulties for virtually every learner. Human language and human learning are far too variable for promises like that to pass without scrutiny.

A phoneme sort is not an explanation of English orthography

Reading Simplified uses a sentence such as “He sees many of these each fall” to help children remember several spellings associated with /ee/. As a memory device, perhaps that is useful.

It still does not explain why those particular words are spelled as they are.

Why is see spelled with <ee>? Why does each contain <ea>? Why does many end in <y>? What happens when pronunciation changes across a family of related words? What role does morphology play in preserving spelling relationships? What can a word’s relatives tell us that its isolated pronunciation cannot?

A phoneme sort organizes words according to one feature. It does not necessarily reveal the principles governing the spellings themselves.

Consider:

heal
health
healthy
healing
healer

The pronunciation represented by the vowel letters changes across this family, while the spelling relationship remains visible. If the words are organized primarily according to phoneme, heal and health are pulled into different categories. If they are investigated as relatives, the stability of the spelling becomes evidence.

English orthography routinely preserves meaningful relationships across changes in pronunciation. A framework centered on phonemes can acknowledge that fact, but it cannot make full use of it while continuing to treat sound as the primary organizational structure.

Adding morphology is not the same as integrating morphology

This is where Reading Simplified begins to look very much like the broader evolution of Structured Literacy.

As criticisms of phonics-first instruction accumulate, more components are added: morphology, syntax, semantics, vocabulary, comprehension. The resulting framework looks increasingly comprehensive because all the appropriate words appear somewhere in it.

The components are still largely siloed.

Phonology remains foundational. Morphology becomes another instructional area. Syntax gets its own activities. Vocabulary gets its own routines. Comprehension occurs around text. A program can include every one of these domains while still failing to teach children how they interact inside the same writing system.

English does not operate in those silos.

A word carries meaning, structure, grammatical information, pronunciation, spelling, and relationships to other words simultaneously. Those pieces constrain and explain one another.

This is exactly what the recent movement toward word families exposes.

Reading Simplified now points to examples such as muscle → muscular and hymn → hymnal to show why letters that appear puzzling in an isolated word become understandable when related words are examined.

SWI has been teaching words through their morphological families for years. We have long asked children to investigate why an apparently unpronounced letter is present by looking at relatives in which the relationship becomes audible or otherwise visible. The family is not brought in as a mnemonic after the spelling has caused trouble. It is part of the evidence used to understand the word.

So yes, we see phonics-oriented approaches beginning to use ideas SWI has been foregrounding for a long time.

I am glad they are learning from them.

If muscle/muscular and hymn/hymnal are legitimate evidence, then the lesson is larger than “word families can sometimes help explain silent letters.” These examples demonstrate that English spelling represents relationships among words and that pronunciation alone cannot account for the orthography.

Once that is acknowledged, it should change the framework.

Morphology cannot remain an enrichment layer added after the supposedly basic code has been taught. Word families cannot be pulled out only when a spelling seems irregular. Meaning cannot be confined to vocabulary instruction while phonology continues to organize beginning word study.

Authentic integration requires changing the questions we ask from the beginning.

Inclusion is not integration

A program can say that it includes phonology, morphology, syntax, semantics, vocabulary, and comprehension. That tells us what is somewhere in the curriculum. It does not tell us how the curriculum understands language.

If phonology is taught first and the other components are subsequently added around it, the architecture remains phonics-first even if the program eventually contains quite a lot of linguistics.

This is the same limitation I see as Structured Literacy tries to expand beyond its earlier phonics-heavy framework. The response has often been additive. Morphology is added. Syntax is added. Semantics is added. More boxes appear on the diagram, but the relationships among them remain poorly developed.

The resulting curriculum can look comprehensive while still asking children to assemble the language system from pieces that adults deliberately pulled apart.

SWI begins from a different premise. The word itself is available for investigation.

We can ask what it means, how it is built, which words belong to its family, what those relatives reveal, and what jobs the letters are doing within that structure. Phonology is absolutely part of that investigation, but it does not get to erase the rest of the evidence before the investigation begins.

A word like sign does not require separate weeks for phonics, morphology, vocabulary, and syntax. Its relatives signal and signature immediately give us information about meaning, structure, pronunciation, and spelling. The child can encounter those relationships together because they actually exist together.

That is much closer to the system the child is trying to learn.

What happens to the children who do not self-teach the rest?

Reading Simplified makes ambitious claims about how quickly children can acquire enough of the sound-spelling system to begin teaching themselves the rest.

Some will.

Many children are remarkable statistical learners. Give them enough print experience and enough partial information, and they will infer relationships that were never explicitly taught. Plenty of reading programs have survived for decades partly because typically developing children are very good at learning despite the limitations of the instruction they receive.

The children who concern me are the ones who do not make those leaps.

They do not automatically infer the rest of the orthographic system after several months of phoneme-centered instruction. They do not effortlessly hold multiple possible graphemes in working memory while decoding an unfamiliar word. They do not reliably discover morphological relationships after years of being trained to attend primarily to sounds.

Those are often the children who eventually reach clinicians like me.

By then, some are convinced that English spelling is arbitrary. Others believe they are simply bad at reading. Many have learned elaborate routines for attacking words without gaining much understanding of why the words are written the way they are.

Repeating the same organizing principle with greater intensity does not necessarily solve that problem.

Sometimes we need another way in.

The framework has to change

I agree with Reading Simplified that conventional phonics instruction often moves too slowly. I agree that children should encounter alternative spellings much earlier than many traditional programs allow. I agree that word families can illuminate spellings that make little sense when words are considered in isolation.

Those observations point toward a more interesting conclusion than accelerated phonics.

They suggest that the sound-first framework itself is too small.

English orthography represents pronunciation, but it also represents meaning and structure. It preserves relationships among words even when pronunciation changes. Morphology, etymology, syntax, semantics, and phonology are not separate curricular accessories attached to a phonics core. They are interacting sources of evidence within the same language system.

If we truly accept that, instruction cannot continue to begin with phonemes and simply bolt everything else on later.

The entire architecture has to change.

Teaching several spellings for /ee/ instead of one may speed up phonics instruction. Adding word families may make some spellings easier to explain. Adding morphology, syntax, vocabulary, and comprehension may make a program look much broader than phonics programs of the past.

But adding more rooms does not change what sits at the center of the house.

If phonology still owns the house, we have not yet moved beyond phonics-first instruction.

Scaffolding should reduce the distance between the learner and the language. If the learner has to learn the scaffold itself, the scaffold can become part of the problem.

That is something I think about often in literacy intervention, because we have a tendency to add things. We add colors, cards, characters, hand motions, tapping routines, boxes, circles, underlines, manipulatives, symbols, games, and scripted procedures. Each addition is usually intended to help. It may make the lesson look more structured or more multisensory, and sometimes it really does make something easier for a particular learner to notice.

But every added element also asks something of the learner.

A color has to be interpreted. A card has to be manipulated. A gesture has to be planned and coordinated. A character and its role have to be remembered. A sorting routine has to be learned. A symbol has to be connected to whatever linguistic concept it is supposed to represent.

That may be worthwhile if the added layer is helping the learner access the language more easily. If it is not, we have simply added noise to the system.

The scaffold is not the thing being learned

One of the most useful questions we can ask ourselves during intervention is deceptively simple:

What do I actually want the learner to notice here?

If I am working with the word jumped, I may want the learner to notice the relationship between jump and jumped, understand what < -ed> is doing, examine its spelling, and connect that spelling to pronunciation.

The learner does not also need to master an elaborate procedure for demonstrating that knowledge unless the procedure itself is doing useful work.

They do not need to become good at tapping in order to understand phonemes. They do not need to remember a permanent color code in order to understand morphemes. They do not need to know which cartoon character represents a suffix before they can understand what a suffix contributes to a word. They do not need to circle, scoop, box, slash, slide, or gesture unless that action is actually helping them attend to the linguistic information we are trying to teach.

We should not confuse an instructional routine with the language itself.

There are two things I do consider essential when we are investigating morphological structure: word sums and the Matrix.

Those are different from an added color system, character, gesture, or mnemonic because they represent the linguistic relationships we are actually studying. A word sum makes the morphological construction of a word explicit:

< jump + ed—> jumped >

A Matrix organizes a family around a shared base and allows us to see how that base combines with other morphemes to build related words. We can use it to generate words, test possibilities, compare structures, and see relationships that are much harder to hold in working memory if they exist only in conversation.

I do not think of either one as decoration added on top of the language. They are ways of representing the language.

So when I talk about removing unnecessary scaffolds, I am not talking about removing the Matrix or word sums. Those stay. What I can change is how the learner interacts with them. I can write the word sum while the learner tells me what to write. The learner can type it. We can build a Matrix together on paper or on a screen. A learner with limited motor control can point, select, or use an AAC device while I do the physical writing.

The linguistic thinking stays intact even when the response mode changes.

That does not mean I never use other visual or physical supports. I do. Sometimes I will print a poem in block letters and have a learner color the morphemes while we talk about them. We might put one kind of design over digraphs and another over trigraphs while we investigate what the letters are doing. A visual mark on the page can make a particular relationship easier to see.

The difference is that the marking is serving the investigation. The child does not have to learn an entirely separate system before gaining access to the word.

Sometimes even a familiar scaffold adds too much

Tapping is a good example.

I sometimes use tapping while we are announcing the sounds in a word. For some learners, that small motor action can help them hold the sequence in mind.

For others, it makes the task harder.

Many of the children I work with have more going on than dyslexia. Some also have dysgraphia, fine-motor difficulties, coordination challenges, or motor-planning difficulties. For a child with mild apraxia or poor fine-motor control, coordinating a tapping sequence while simultaneously thinking about speech sounds, letters, spelling, and the word itself may be an additional burden rather than a support.

In that case, we do not tap. We write, or I write while the learner talks, or the learner types. Nothing has gone wrong because we abandoned the routine. The routine was never the goal. The language was.

Response mode should remove barriers without removing useful work

The same principle applies to digital activities.

Clicking and dragging a morpheme onto a base can be useful. Sorting words on a screen can be useful. For a learner with limited fine-motor control, significant motor-planning difficulties, or someone who accesses language through an AAC device or other assistive technology, clicking and dragging may be exactly what allows them to participate fully in the linguistic task.

The response mode should fit the learner.

But if a child is capable of writing or typing, I generally want them producing the language themselves rather than only manipulating pieces that someone else has already created.

There is a difference between dragging< -ed > onto jump and writing:

< jump + ed—> jumped >

When the learner writes or types the morphemes, they have to retrieve the spelling, sequence the letters, and produce the written forms. If they announce the letters as they write, they are attending directly to the orthography while constructing the word.

The same applies when we build a Matrix. I can hand a learner a completed Matrix and ask them to find words in it, and there may be times when that is useful. But whenever possible, I want them involved in constructing it. Deciding where a morpheme belongs, writing or typing it, testing a word sum, and determining whether the resulting word belongs in the family requires much more linguistic thinking than simply recognizing a prepared answer.

That gives us a useful clinical rule:

Choose the response mode that removes an access barrier without removing useful language work.

If motor demands interfere with the linguistic task, change the response mode. Do not change the thinking.

More multisensory is not automatically better

The term multisensory has become so positively loaded in literacy education that it can be difficult to question whether every sensory addition is actually helping.

But using more modalities does not automatically make an activity more effective.

A child can be looking, listening, moving, touching, and talking while still devoting very little attention to the thing we actually want them to learn. In fact, the more actions we layer onto a task, the more careful we need to be about where the learner’s attention is going.

Are they thinking about why two words are related, or are they trying to remember which card to pick up?

Are they noticing what a suffix contributes to a word, or are they trying to remember which color represents the suffix?

Are they thinking about what the letters are doing, or are they concentrating on getting the hand motion right?

A busy lesson can look wonderfully engaging from the outside while placing a tremendous amount of unnecessary demand on the learner.

Sometimes the deepest learning looks surprisingly plain. A child and an adult are sitting together with a word on a piece of paper. They compare it with another word. They talk about what the words mean. They write a word sum. They add relatives to a Matrix. They test a possible structure. They notice a spelling. They change their minds and try again.

There may not be much to photograph for social media, but there may be a great deal of thinking happening.

The lightest scaffold that works

I have become increasingly interested in using the lightest scaffold that allows a learner to see, understand, and work with the linguistic information.

Sometimes that means a printed sheet.

Sometimes it means a highlighter.

Sometimes it means typing rather than handwriting.

Sometimes it means dragging pieces around a screen.

Sometimes it means tapping.

Sometimes it means eliminating all of those things and simply talking about a word while we write it.

The Matrix and word sums remain because they are part of how we make morphological structure and relationships explicit. The optional layers around them can come and go according to the learner.

The decision should come from the learner and the linguistic target, not from loyalty to a routine.

A scaffold earns its place when it makes something easier to perceive, remember, understand, or express. When it requires the learner to devote additional attention to a system that is not itself the object of learning, we should at least ask whether it is helping enough to justify the cost.

That question becomes especially important for students with working-memory limitations, motor-planning difficulties, dysgraphia, language disorders, attention difficulties, or other challenges that make every additional demand more expensive.

For those learners, “one more helpful thing” may not feel helpful at all.

You do not need fancy materials to teach language deeply

There is a tendency to associate sophisticated instruction with sophisticated materials. The more pieces on the table, the more intentional the lesson can appear.

I do not think that is where the sophistication lies.

The sophistication is in knowing the language well enough to recognize what is worth investigating. It is in knowing which question to ask next. It is in listening carefully enough to notice what the learner understands and where the misunderstanding begins. It is in knowing when to add a scaffold and when to remove one.

Sometimes all you need is the text, the learner, a Matrix, some word sums, and a conversation.

Sometimes you need a computer or an AAC device because that is how the learner accesses the conversation.

And very often, if the learner can write or type, a pen and paper or a keyboard will do quite a lot.

Keeping instruction simple does not mean making the thinking simple.

It means leaving room for the thinking to happen.

One of the assumptions that has shaped reading intervention for decades is fairly straightforward: if a learner has difficulty with phonological processing, we should spend more time strengthening phonological processing.

There is evidence supporting the importance of phonology in learning to read, but after years of working with students who have severe and persistent reading difficulties, I have become increasingly interested in another question. What is the learner already doing when phonological processing is difficult? Are there other parts of the language system that the reader can use, and may already be using, to recognize written words?

A fascinating study by Jeremy Law and Pol Ghesquière gives us a window into that question.

Their 2022 paper, Morphological Processing in Children with Developmental Dyslexia: A Visual Masked Priming Study, examined how fifth-grade children with dyslexia processed the morphological structure of written words. The study included 24 Dutch-speaking children with dyslexia and 46 typically reading children of the same age.

The children with dyslexia showed intact, rapid morphological processing. Even more interestingly, the way they used morphological information differed from the way the typically reading children used it, and meaning appeared to play an important role.

This was not a test of whether children could explain morphology

The researchers were not sitting across from children asking them to identify prefixes, explain suffixes, make new words, or consciously manipulate morphemes. The children did not have to know the word morpheme, nor did they have to explain how a word was structured.

Instead, the researchers used a masked priming task. A word appeared on the screen so briefly that the child was not consciously aware of seeing it. Immediately afterward, another word appeared, and the child had to decide whether that second item was a real word.

If the first word activates information that helps the reader process the second word, response time becomes faster. For example, if teacher flashes briefly before TEACH, the reader may recognize teach more quickly because teacher and teach are morphologically related.

This allowed the researchers to investigate rapid morphological processing during visual word recognition rather than a child’s ability to perform an explicit metalinguistic task.

A child’s ability to explain morphological structure is not the same as the reading system’s use of morphological information. We see this throughout language development. Children use syntax long before they can identify a clause. They use grammatical morphemes long before they can define a suffix. A test of explicit morphological awareness therefore does not tell us everything about how morphological information is functioning during reading.

Real morphological relationships and apparent ones

The researchers did not simply compare related and unrelated words. They created different kinds of prime-target relationships so they could separate the effects of spelling, meaning, and genuine morphological structure.

One condition involved a true morphological relationship, comparable to:

teacher → TEACH

The two words share spelling, meaning, and morphological structure.

Another condition contained what looked like the same structure on the surface but did not involve a real morphological relationship:

corner → CORN

The ending <er> is certainly a real suffix elsewhere in English, and corner visually contains corn, but corner is not a morphological relative of corn. The researchers also included semantic-only and orthographic-only relationships so they could determine whether meaning or shared spelling by themselves explained the priming effects.

This design allowed the researchers to ask not merely whether children with dyslexia process morphology, but what information makes the reading system treat two written words as morphologically related.

The children with dyslexia showed a substantial priming effect in the genuinely morphological condition: 132 milliseconds. In the pseudo-derived condition, the effect was essentially absent at negative 5 milliseconds. The semantic-only and orthographic-only conditions did not produce significant priming either.

The typically reading children showed a different pattern. They demonstrated priming for both the genuinely morphological and pseudo-derived forms. Their reading systems responded rapidly to something that looked morphologically decomposable even when the apparent pieces were not meaningfully related.

The children with dyslexia did not show that same pattern. Law and Ghesquière concluded that morpho-orthographic information alone was not sufficient to produce the same facilitation for these readers. Their morphological processing depended more heavily on the semantic interpretability of the relationship.

The finding is difficult to square with the idea that more segmentation is always the answer. In this experiment, children with significant reading difficulty were rapidly and automatically sensitive to larger meaningful units, and their reading systems were not similarly facilitated by a merely plausible-looking segmentation.

Earlier phonological difficulty was related to later morphological processing

Because the children were part of a longitudinal research project, Law and Ghesquière also had phonological-awareness data from second grade. They were therefore able to examine whether earlier phonological performance was related to morphological processing three years later.

Among the typically reading children, there was no significant relationship.

Among the children with dyslexia, there was. Poorer phonological-awareness performance in second grade was associated with a larger genuine morphological priming effect in fifth grade. The relationship appeared on both phonological-awareness measures used in the study.

This is correlational evidence, so it does not establish that phonological difficulty caused the children to become more reliant on morphological information. The authors nevertheless offer a plausible interpretation: children who have difficulty processing written language efficiently at the phoneme-grapheme level may increasingly make use of larger linguistic units during word recognition.

Morphemes provide one such source of information because they carry meaning, spelling information, grammatical information, and connections to related words. Law and Ghesquière describe their findings as consistent with morphology functioning as a possible compensatory route for readers who have greater difficulty working at a smaller phonological grain size.

This complicates the familiar instructional sequence in which larger linguistic units are treated as something children reach only after they have mastered smaller ones: first phonemes, then graphemes, then syllables, and eventually morphemes and meaning. Language itself is not organized in that sequence. A written word carries its pronunciation, spelling, morphological structure, grammatical possibilities, meaning, and relationships to other words simultaneously, and the reader’s brain is not required to use those sources of information in the order we choose to teach them.

In this study, the children with dyslexia were already processing morphological information rapidly despite substantial phonological difficulty. The authors explicitly concluded that their morphological representations were intact.

They were not waiting for phonology to be repaired before morphology became available.

Meaning changes what counts as a useful “chunk”

The contrast between teacher and corner also has implications for the way morphology is taught.

A learner can be taught to find pieces in words, circle a suffix, underline something that looks like a root, or break a word into visible chunks. None of those procedures guarantees that the resulting analysis represents actual morphological structure.

<er> is a suffix in teacher because teacher is built from teach and <er>, and the relationship is supported by meaning and by other members of the same morphological family. Seeing the same letters at the end of corner does not establish the same structure.

The dyslexic readers in this study responded differently to those two kinds of relationships. Their processing was facilitated by the real morphological relationship, not by a letter sequence that merely resembled one.

For instruction, that raises a practical question about what we mean when we say we are teaching morphology. If morphology becomes another exercise in spotting chunks, memorizing lists of roots, or stripping apparent prefixes and suffixes from words, we may be leaving out the very feature that appears especially relevant for these readers: the meaningful relationship among words.

Morphology is not simply visual segmentation. It is structure that carries meaning.

What the study does and does not establish

Law and Ghesquière did not conduct an intervention study. They did not teach morphology to one group and phonics to another, and they did not demonstrate that morphological instruction should replace phonological instruction.

The study involved Dutch-speaking fifth graders, so it would also be inappropriate to assume that every feature of the finding transfers identically to English-speaking readers of every age.

The relationship between earlier phonological difficulty and later morphological processing also remains correlational. The study cannot tell us exactly how that developmental pattern emerged.

The authors themselves call for additional longitudinal and experimental research, including research examining whether morphological instruction changes this kind of rapid morphological processing.

What the study does show is that these children with dyslexia were already processing genuine morphological relationships quickly and automatically despite significant phonological difficulty. The children with weaker earlier phonological skills also showed greater later sensitivity to meaningful morphological relationships.

That is not evidence that morphology can simply be trained as a compensatory strategy. It is evidence that some children may already be making use of morphology in precisely that way.

The weakness does not always have to be the entrance point

When a learner has severe phonological difficulty, intervention often begins by identifying the weakest part of the system and making that weakness the primary route into reading instruction. There are times when direct work on that weakness is appropriate, but it does not follow that every other useful source of linguistic information should wait until the weakness has been mastered.

A child who struggles intensely with phonological processing still has a language system. The child knows words, understands meanings, recognizes relationships, uses syntax, and brings knowledge about the world to print. Law and Ghesquière’s findings suggest that morphological processing may remain available even when phonological processing is substantially compromised.

For instruction, that opens another route. We can work with the whole word and investigate its meaning, its structure, its relatives, its history, and the jobs its graphemes are doing within that structure. Phonology remains part of the investigation, but it does not have to function as the entrance exam for everything else we know about the word.

For some readers who have spent years struggling with phoneme-by-phoneme approaches, the meaningful structure of words may not be an advanced skill waiting for them at the end of intervention. It may already be part of how they are reading.

Reference

Law, J. M., & Ghesquière, P. (2022). Morphological Processing in Children with Developmental Dyslexia: A Visual Masked Priming Study. Reading Research Quarterly, 57(3), 863–877.

A mom in one of my Structured Word Inquiry Foundations classes was looking at three words:

prefer

preference

preferential

We were talking about how the pronunciation changes across the family.

preFER

PREFerence

preferENtial

The stress moves, and the vowels change with it. In preference, many speakers reduce the medial vowel dramatically; in my own pronunciation, I don’t produce a separate vowel there at all. The spelling remains.

Then this mom realized what she had been asking children to do with written syllable types. Her jaw dropped.

The words are clearly related. Their spelling shows the relationship. Yet if she analyzed each word according to written syllable types and surface pronunciation, she would have to keep changing the analysis as she moved through the family.

Morphologically, we can investigate:

<pre + fer —> prefer>

<prefer + ence> —> preference

<prefer + ent + i + al> —> preferential

The <i> in preferential is a connecting vowel letter before the suffix <al>.

The spelling gives us stable information while the pronunciation changes. Once you see that, a question becomes difficult to avoid: Why are we teaching children to organize written words around syllable categories that keep changing with pronunciation when the writing itself is preserving meaningful structure?

I have been asking that question for years.

The more I have studied English orthography, worked with struggling readers, and read the research on multisyllabic word instruction, the less justification I can find for asking children to memorize and apply written syllable types. Increasingly, the literature coming from within Structured Literacy itself is acknowledging many of the same problems.

First, I am not saying syllables don’t exist

Syllables are real units of speech.

Children can hear syllables. We can count them. We can compare stress. We can investigate what happens to vowels in unstressed syllables. Syllable awareness is part of phonological awareness.

That is not the same thing as teaching children that written English is organized into six categories generally called closed, open, vowel-consonant-e, vowel team, r-controlled, and consonant-le syllables, and then teaching procedures for drawing division lines through written words.

Those are invented instructional categories. A child can know that a spoken word has three syllables without learning to mark, scoop, divide, and label three pieces of its written form.

There is also research showing benefits from some kinds of flexible syllable-level chunking. That is not evidence that children need the six written syllable types or rigid syllable-division rules. Timothy Shanahan, who still recommends limited syllabication instruction, has acknowledged that approaches based on rigid spelling rules for syllabication have not shown the same benefit as more flexible approaches.

My argument here is specifically about the written-syllable apparatus that has become deeply embedded in Orton-Gillingham and many Structured Literacy programs: identify the vowels, determine where to divide, label the written syllable types, retrieve the associated vowel rules, pronounce each chunk, put the chunks back together, and then adjust when the result doesn’t sound like a word.

That is the system I stopped using.

Look at what happens with two short words

Start with hoping and hopping.

Written syllable analysis: ho | ping hop | ping

Morphological analysis: <hope + ing> → hoping

<hop + ing> → hopping

In hoping, ho can be labeled an open syllable. The child retrieves the rule that a vowel at the end of an open syllable is generally “long.”

That can get the child to the pronunciation, but what is ho doing in the word?

It has no meaning in hoping. It does not explain why the word is spelled with one <p>. It does not connect hoping to hope. It does not explain what happened to the final <e>.

The morphological analysis does all of that:

<hope + ing> → hoping

Now compare:

<hop + ing> → hopping

We can investigate why the final consonant is doubled in one word and why the final <e> is replaced in the other. We can build the families:

hope, hopes, hoped, hoping, hopeful

hop, hops, hopped, hopping, hopper

The child is not learning less phonology. The phonology is now being studied inside an analysis that also gives information about spelling, meaning, structure, suffixing, and word relationships.

Then the words get longer, and the problem gets worse

Written syllable types look tidiest in carefully selected examples. Longer English words are much less cooperative because English is a stress-timed language.

Vowels in unstressed positions routinely reduce. They may move toward schwa. They may become very short and indistinct. In some pronunciations, a vowel disappears from speech altogether.

The Structured Literacy literature acknowledges this explicitly. In Structured Literacy Interventions, Kearns, Lyon, and Kelley explain that reduced vowels occur throughout polysyllabic English and that even when the individual parts of a word have been decoded, those parts may not sound like the naturally pronounced whole word.

Louisa Moats, writing in the same volume, calls schwa an “unavoidable problem” in multisyllabic words and warns teachers not to rely exclusively on spelling words by sounding them out. She also explicitly notes that syllable boundaries and morpheme boundaries often differ, and that when syllables are the focus, students may not simultaneously recognize the morphemes.

That is a significant admission. The instructional literature itself recognizes that focusing on written syllables can interfere with recognizing the meaningful structure of the word, yet written syllable work remains part of the instructional sequence.

That is where I part ways.

What is the syllable type of a vowel you don’t say?

Return to preference.

Depending on dialect and speaker, the vowel in the middle may be substantially reduced or elided altogether. That is normal English pronunciation.

So what is a struggling reader supposed to do with a written syllable procedure when the spelling contains a vowel letter but speech does not provide a nice, fully realized vowel phoneme for it?

Which written syllable type tells the child what to do with a vowel that has been reduced almost beyond recognition? What if the child’s pronunciation differs from the teacher’s? What happens when the stress moves again in preferential?

The problem is not with English. The problem is expecting surface pronunciation to provide a stable organizing structure for a written word.

The morphological family is still sitting there:

prefer

preference

preferential

The spelling is doing exactly what we should want a writing system to do: preserving relationships even while normal spoken pronunciation changes.

This happens everywhere

Consider family.

A worksheet can present:

fam | i | ly

But plenty of English speakers ordinarily say something much closer to fam-ly. The medial vowel can reduce so much that it disappears.

The child may know the spoken word perfectly well. Then we ask the child to pronounce three carefully articulated written chunks, hold those chunks in memory, combine them, and somehow recognize the natural word they already knew.

When that becomes difficult, we can end up treating the child as though the problem is blending when the procedure itself may be creating part of the problem.

Connecting family with familiar and familiarity gives us information about the spelling that does not depend on pretending every vowel letter must be strongly represented in every pronunciation.

Or consider intensive.

A syllable division can give us:

in | ten | sive

But its meaningful structure is:

<in + tense + ive> → intensive

One analysis creates temporary pronunciation chunks. The other shows the meaningful structure of the written word while still allowing us to investigate its pronunciation.

Sometimes you have to know the word before you can divide it

This is one of the strangest problems with written syllable division.

Look at:

tiger

and

river

If you already know the words, you can divide them so that the resulting written syllable classification appears to explain the vowel pronunciation.

But suppose you don’t know the word. Where does the consonant go? Which pronunciation should you try?

Shanahan uses words such as tiger and camel to illustrate the ambiguity. Divide tiger one way and you get the expected pronunciation. Apply the same decision to camel and you get something like came-el. His solution is to teach children to try alternatives and remain flexible.

But follow what the child actually has to do.

The child sees an unfamiliar word, tries one division and one vowel pronunciation, and decides whether the result sounds like a known word. If it doesn’t, the child changes the division or changes the vowel and tries again. Perhaps the child tries schwa. Eventually, the child’s oral vocabulary has to tell them when they have arrived at a real word.

That is not a rule reliably determining pronunciation. It is a trial-and-error procedure constrained by vocabulary.

This was not obvious to me until I really tried to make written syllable division make sense.

Look at:

robin, focus, camel, pilot, table, baby, lemon, tropical

If you know the pronunciations, you can go backward and place division lines that appear to explain them. If you do not know the words, the lines do not reliably tell you which pronunciation to choose.

Researchers were pointing out this problem decades ago.

In Emerald Dechant’s 1969 Linguistics, Phonics, and the Teaching of Reading, Dechant summarizes Glass’s work on syllabication. Glass observed that readers appeared to use known sounds to determine syllable divisions rather than using syllabication rules to discover the sounds. He asked the obvious question: Why syllabicate?

Dechant went on to note that a good reader does not need to consciously divide every word or know syllabication rules, and that continuously doing so could slow reading and interfere with comprehension.

This objection is more than half a century old.

I recently saw the problem described in real time

A practitioner was discussing a student who had received years of specialized dyslexia intervention. The student had become quite capable of recognizing many words, but she was still expected to perform the program’s syllable-division procedures.

The practitioner described what the student actually did: she figured out the word first and then divided it.

Once she knew what the word was, she knew its pronunciation. Once she knew the pronunciation, she could work backward and perform the required syllable analysis.

But if she had already figured out the word, what work had the division done?

Reading was the achievement. The division was a second task performed after reading.

This becomes especially concerning when mastery of the procedure itself is used as a gate a student must pass before receiving more advanced instruction. A child can end up spending additional weeks practicing how to demonstrate the instructional method rather than expanding what they know about words.

Now let’s talk about cognitive load

This is where the case against rule-based written syllable division becomes particularly difficult for me to ignore.

Imagine what a child may be asked to do when reading one unfamiliar multisyllabic word. The child may need to find the vowels, identify vowel teams, inspect the consonants between vowels, remember a division rule, decide where the division line belongs, label each written piece by syllable type, retrieve the expected vowel pronunciation associated with each type, pronounce the pieces, hold those pieces in phonological working memory, combine them, notice when the result does not sound like a known word, remember which vowel can “flex,” try an alternative pronunciation, perhaps try schwa, reassemble the word, check it against oral vocabulary, and then return to the sentence and remember what the sentence was about.

That is a substantial amount of executive and working-memory work to impose in the middle of reading.

For a child who already has weaknesses in working memory, language processing, attention, sequential processing, or executive control, the cost may be even greater.

This is not simply my clinical impression.

A chapter specifically devoted to long-word intervention in Louise Spear-Swerling’s 2022 book Structured Literacy Interventions was written by Devin Kearns, Cheryl Lyon, and Shannon Kelley. Their summary of rule-based syllabification is remarkable.

They report that these approaches:

• do not work very consistently;

• may involve rules that are too numerous and complex to be very helpful;

• may tax students’ working memory because the child must remember the rules and parts of the word simultaneously; and

• have been examined directly in very few intervention studies.

Their conclusion is quite plain:

“Overall, the data do not strongly support teaching rule-based syllable division.”

That statement appears in a book called Structured Literacy Interventions. These are not SWI practitioners criticizing Structured Literacy from outside the framework. These are authors working within Structured Literacy summarizing the limitations of the evidence themselves.

Even the alternatives are designed partly to reduce the burden

The same chapter discusses more flexible approaches to breaking up longer words. In describing Bhattacharya and Ehri’s graphosyllabic work, the authors specifically note that students were encouraged to be flexible about where they split words in order to reduce working-memory load and make the strategy more fluid.

That tells us something important about the original problem. When strict divisions become too cumbersome or unreliable, the response is to loosen the divisions.

Other approaches simplify the rules. Others teach children to try alternative vowel pronunciations. Others tell children to “flex.” Some have children divide a word where the resulting pieces “look OK.”

The field keeps trying to make the procedure lighter.

I think we should also be willing to ask whether the child needs the procedure in the first place.

These routines take instructional time to learn

Teaching a child a procedure for breaking longer words into syllable-based chunks requires instructional time before that procedure can even be used independently.

In the same Structured Literacy chapter, Kearns, Lyon, and Kelley describe research in which students practiced one multisyllabic word-reading procedure every day. They report that it took students approximately a week to fully process the method and another two weeks before they began applying it independently. The chapter recommends approximately 10–15 minutes per day for at least four weeks to give students enough practice to internalize and use the strategy.

Before we devote that much intervention time to teaching a child a procedure, we should ask what the procedure is buying us.

Ten to fifteen minutes a day could also be spent reading, writing, building vocabulary, studying meaningful word families, investigating bases and affixes, comparing pronunciations across relatives, learning why spellings remain stable, and using those words in sentences and connected text.

It is not enough to establish that children can learn a procedure. We need to know whether the procedure is useful enough to justify the instructional time and cognitive effort required to learn it. The amount of time is longer for many learners of course and many never learn to apply syllable types at all.

We finally have data on the rules/types

In 2020, Devin Kearns tested two commonly taught syllable-division patterns against 14,844 words from texts used in Grades 1 through 8.

For the VC|CV pattern, the familiar rabbit pattern, the expected short-vowel pronunciation occurred in 70.6% of applicable words.

For the V|CV pattern, the familiar tiger/mason pattern, the expected long-vowel pronunciation occurred in only 30.5%.

The results became even less encouraging when the words became longer.

For words with more than two syllables:

VC|CV: 62.5%

V|CV: 18.8%

That 18.8% deserves attention because longer words are precisely the words for which children are often told syllable division will help them.

Kearns concluded that the unreliability of the VCV pattern “may not justify the effort required to use the strategy.”

There is another important point in his analysis. When reduced vowels such as schwa and other vowel pronunciations are removed from consideration, the apparent accuracy of the patterns improves.

But children reading English do not get to remove vowel reduction from English.

Schwa is not an inconvenient statistical contaminant. It is a normal feature of English phonology, and the words containing those reduced vowels are the words our children actually have to read.

Kearns studied specific syllable-division patterns. These percentages should not be generalized to every one of the commonly taught written syllable types because equivalent corpus percentages have not been established for all of them.

What the data do show is that at least some of the rules children are taught with considerable confidence are much less dependable than that confidence suggests.

The newest research makes this even harder to defend

Kearns returned to the question with Christina Novelli in a study published in Annals of Dyslexia in July 2026. They investigated whether children in Grades 2–4 actually develop sensitivity to the subtle VCV long-vowel tendency.

Children showed only limited sensitivity to the pattern. Older children showed more sensitivity than younger children, and students with reading difficulties were less likely than their peers to use the long-vowel tendency.

Kearns and Novelli then raise the instructional question directly. There are few empirical tests of the value of syllable-division strategies, and the statistical patterns may not be consistent enough to justify explicit instruction. Even when children can learn subtle tendencies, educators must consider whether the benefit is worth the cognitive demand of storing and retrieving pattern information that applies to a limited number of words.

Their study was not an instructional trial, so their conclusion is appropriately cautious. Still, they explicitly question whether teaching apparent syllable structures provides enough benefit to justify the time required to model, practice, and master them.

The question is not simply whether children can memorize these patterns. It is whether we can justify asking them to.

Jennifer Buckingham raises the same question about instructional time

In 2024, Jennifer Buckingham wrote about what teachers actually need to know from the Science of Reading literature. Her discussion of syllables is notable because she calls them the least stable of the subword units she examines and asks how useful it is to spend instructional time teaching rules based on syllable division.

She points to Kearns’s findings and also resurrects an observation Edward Dolch made in 1945:

“Rules require an extra step between seeing print and thinking sound.”

Dolch’s point was that the extra step should not be inserted when it can be avoided.

Buckingham then moves directly to the issue of opportunity cost. Whenever instructional time is spent on one thing, that time is unavailable for something else.

This is exactly the calculation I want us to make for struggling readers.

The question is not whether we can teach children written syllable types. Many children can learn them.

The question is whether this is the best use of their time and cognitive resources.

The contradiction is already visible inside Structured Literacy

This is the pattern I keep noticing in books and articles.

Authors describe many of the problems accurately and then continue recommending some version of the same framework.

Kearns, Lyon, and Kelley report that rule-based syllable division is inconsistent, complex, potentially taxing to working memory, weakly studied, and not strongly supported. Their answer is to recommend more flexible approaches to syllable-level chunking.

Shanahan acknowledges the ambiguity of English syllable division and that rigid rule-based approaches have not produced convincing reading gains. His answer is to retain limited syllable work and teach children to flex between alternatives.

Buckingham questions the instructional time devoted to syllable rules and describes syllables as the least stable subword unit while still leaving room for some basic syllable instruction.

Moats describes the problem of schwa, warns that sounding out becomes limited in multisyllabic words, and explicitly acknowledges that syllable boundaries may interfere with recognizing morpheme boundaries, yet written syllable types remain in the instructional sequence.

I don’t think these authors are being dishonest or deliberately contradictory. I think they are confronting a genuine problem inside an inherited instructional framework.

Once we assume that longer written words must first be managed through pronounceable syllable chunks, the available solutions are better syllable rules, fewer syllable rules, flexible syllable rules, or syllable rules plus morphology.

I no longer share that assumption because another structure is already available to us in the word.

Morphemes can give us fewer pieces to hold

One of the most interesting arguments in the Kearns, Lyon, and Kelley chapter actually favors morphological analysis.

The authors point out that words generally contain fewer morphemes than syllables, which can make a morphological approach more efficient and reduce how much information a student must maintain in memory.

They illustrate the point with unpresentable. In their analysis, the word has five syllables but three morphological pieces:

un + present + able

The important point here is their own comparison: fewer meaningful pieces may mean less strain on phonological memory.

That deserves attention.

If we are concerned that long words overwhelm struggling readers, and one analysis gives the child more chunks to hold while another gives fewer, meaningful chunks that also contribute to vocabulary and spelling knowledge, we should at least question why the syllable chunks are automatically given priority.

Kearns, Lyon, and Kelley also give an example of syllable division creating the problem

The same authors use reappear as an example of a word in which syllable division can lead a reader toward a misleading pronunciation.

This is their example, not mine.

They explain that without recognizing the morpheme re-, a reader might divide the written word in a way that leads toward a pronunciation something like reep-pare. Their point is that recognizing the morpheme prevents the misleading division.

In other words, the Structured Literacy text itself gives us an example in which morphology supplies information that syllable division does not.

That is worth noticing because it is not an argument being imposed on their work from outside. It is part of their own explanation of why morphological information is useful in reading long words.

Peter Bowers makes the structural problem explicit

Peter Bowers has made this problem particularly clear with words such as action.

A syllable division may give:

ac | tion

But the morphological structure is:

<act + ion> → action

Now we can see why the <t> is there and connect the word to:

act, actor, acting, react, reaction, action

The pronunciation of the base changes, but its spelling preserves the relationship.

Because syllable boundaries are not constrained by morpheme boundaries, dividing words into syllables can cut directly through the structures that explain their spelling and connect them to their relatives.

English spelling is not merely a transcription of pronunciation. It represents language, and language includes meaning, morphology, phonology, syntax, and history.

The spelling often remains constant precisely when pronunciation does not.

Consider another family

Now look at:

tribe

tribal

tribute

contribute

contribution

distribute

distribution

redistribute

A syllable-type routine may treat trib as a closed written syllable in one member of the family. Then we get to tribal.

Now what? Is tri open?

Did the meaningful structure suddenly change?

No. The pronunciation changed.

The family did not.

Morphological investigation allows us to follow the relationships:

<tribe + al> → tribal

<tribe + ute → tribute

<dis + tribe + ute> → distribute

<re + dis + tribe + ute> → redistribute

We can then investigate related formations such as distribution, contribute, and contribution.

The vowels shift as stress changes. Some reduce substantially. The spelling continues to carry information about the family.

A written-syllable routine has to keep relabeling pieces according to the pronunciation of each individual word. The morphology gives us a structure that travels across the family.

There is direct dyslexia research worth knowing

This is not only an argument from linguistic structure.

In a classic study by Carsten Elbro and Elisabeth Arnbak, Danish adolescents with dyslexia received morphological training and were tested under different text-presentation conditions.

One particularly interesting finding was that the readers with dyslexia were helped when text was presented morpheme by morpheme. They were not similarly helped when it was presented syllable by syllable.

Later reviews have discussed morphological processing as a possible compensatory resource for readers with dyslexia.

That does not prove that syllables are useless, nor does it prove that every morphological intervention is superior to every syllable-based intervention.

It does tell us something important about what kind of chunks can carry useful information for a struggling reader. A syllable gives information about pronunciation. A morpheme can contribute information about pronunciation while also carrying spelling and meaning.

The broader morphology research points in the same direction

A 2010 systematic review by Bowers, Kirby, and Deacon examined 22 studies of morphological instruction. They found benefits for literacy learning, with particularly strong benefits for less able readers, and found no evidence that morphology needed to be postponed until children were older.

Goodwin and Ahn subsequently synthesized 30 studies and found a moderate overall effect of morphological instruction, with significant effects across morphological knowledge, phonological awareness, vocabulary, decoding, and spelling.

Their separate meta-analysis focused on children with literacy difficulties also found benefits from morphological intervention.

A newer preregistered meta-analysis by Colenbrander and colleagues, published in 2024, found small-to-moderate positive effects of morphological instruction on children’s reading and spelling while also appropriately noting limitations involving transfer, heterogeneity, and certainty of evidence.

None of these studies proves Structured Word Inquiry as a complete instructional approach. That is not what I am claiming.

They do show that meaningful morphological structure is not an optional decorative addition that must wait until after children have mastered written syllable types.

Children can learn from morphology, including struggling readers, and morphology is already part of the writing system they are trying to understand.

“Teach both” is not automatically the safer choice

A common response is: Why not teach syllable types and morphology?

For some learners, you can.

But adding two analytical systems does not automatically make instruction more comprehensive. For many children, its overload and causes confusion.

Consider what we are asking the child to do:

ho | ping

versus

<hope + ing>

Or:

ac | tion

versus

<act + ion>

Or:

in | ten | sive

versus

<in + tense + ive>

These analyses divide the written word differently because they are answering different questions.

The child then has to learn which boundary to attend to for which task and when one analysis should temporarily take priority over the other.

For children with strong working memory and cognitive flexibility, that may be manageable. For a child with language-processing difficulties, limited working memory, executive-function weaknesses, sequential-processing difficulty, or years of accumulated instructional confusion, two competing maps of the same word can create an additional burden.

More instruction is not automatically better instruction. Sometimes it simply gives the child more procedures to manage.

And what happens after all the rules?

This is one of the hardest parts of the system for me to defend.

After the child has identified the vowels, divided the word, labeled the syllables, remembered the rules, and pronounced the pieces, what do many modern approaches tell the child to do when the result is wrong?

Flex the vowel.

Try another pronunciation. Perhaps try schwa. See whether one of the alternatives produces a real word.

In one example in Structured Literacy Interventions, a student is explicitly taught that a vowel such as <i> may have a long pronunciation, a short pronunciation, or schwa, and that the child should try the alternatives until a recognizable word emerges.

After all of the rules, the child still needs vocabulary. The child still needs flexibility. The child still needs to recognize which pronunciation produces a word that makes sense.

The rules have not eliminated uncertainty. They have created a procedure for navigating it. It’s ultimately, “guess and check.” However to know if it makes sense when you get to the right pronunciation, you must already know what the word means.

Some children may find that procedure useful.

For children who repeatedly become lost inside the procedure, I see no reason to insist that they remain there.

Pronouncing a word is not the same as learning a word

Suppose the child finally arrives at the pronunciation of preferential.

Do they know what it means?

Do they recognize its relationship to prefer and preference?

Can they spell it?

Can they explain why much of the spelling remains stable even while the pronunciation changes?

Can they use it in a sentence?

Do they recognize <al> as a suffix that forms an adjective?

Do they understand what the connecting <i> is doing?

Can they recognize the same pieces in other words?

Reading is not successful simply because a child managed to produce a plausible string of sounds.

The purpose of reading is language, and the written word gives us access to meaning.

What I do instead

I do not replace six syllable types with six morphology rules. I do not want another list for the child to memorize.

I begin with the word.

What does it mean?

How is it built?

What words are related to it?

What can its history tell us?

Then we investigate what the letters are doing within that structure.

With hoping:

<hope + ing> → hoping

With hopping:

<hop + ing> → hopping

With preference:

<pre + fer + ence> → preference

With preferential:

<prefer + ent> + i + al> → preferential

With action:

<act + ion> → action

The phonology has not disappeared. Now we can investigate it more accurately. We need to know the morphemes before we can ID the graphemes. Graphemes don’t cross morpheme boundaries.

Where is the stress? Which vowels have reduced? Which graphemes represent which phonemes in this particular word? Has the pronunciation of a morpheme changed across the family? Which letters help preserve the identity of that morpheme?

The child is studying the writing system rather than memorizing a procedure for attacking it. They don’t need strategies, they need to understand how words work.

I would rather teach the stable information

This is what finally ended written syllable types for me.

English pronunciation moves. Stress moves. Vowels reduce. Sometimes vowels disappear from speech entirely. Morphemes can change pronunciation substantially across a family.

But English spelling often preserves meaningful relationships:

sign → signal → signature

heal → health

act → action

prefer → preference → preferential

tribe → tribal → distribute → distribution

That stability is not an unfortunate complication getting in the way of phonics. It is part of how English writing works.

Once I understood that, I could no longer justify teaching children to treat pronunciation-driven written chunks as the fundamental architecture of words.

So what does the evidence actually allow us to say?

It allows us to make a much more precise argument than “syllables don’t work.”

Syllables are real units of speech, and some forms of flexible syllable-level chunking have been useful in intervention research.

That is not the same thing as evidence for the six written syllable types and traditional rule-based syllable division.

The direct research base for rigid syllable-division instruction is limited. Mainstream Structured Literacy authors themselves report that the rules may be inconsistent, overly complex, and burdensome to working memory.

Kearns found that one frequently taught pattern predicted its expected vowel pronunciation only 30.5% of the time overall and 18.8% in words longer than two syllables. Those percentages apply to the patterns he studied; we do not have equivalent corpus percentages for every one of the commonly taught written syllable types.

The newer Kearns and Novelli research found limited sensitivity to the VCV tendency and explicitly raises the question of whether teaching such statistical information is worth its cognitive and instructional cost.

The literature also acknowledges that syllable boundaries and morpheme boundaries can conflict, that morphology can provide fewer chunks to maintain in memory, that morphological structure provides meaning information syllables cannot provide, and that vowel reduction creates unavoidable complications in multisyllabic pronunciation.

Meanwhile, morphological instruction has a substantial evidence base of its own.

For me, those findings were more than enough to justify changing my instruction over ten years ago. But the biggest finding for myself as a clinician was the change in reading, spelling, and reading comprehension for my clients.

What I would ask if this were my child

If my child were spending intervention time marking vowels, drawing division lines, labeling open and closed syllables, memorizing syllable-type definitions, reciting division rules, and then “flexing” vowels when the rules did not produce a recognizable word, I would want answers to some very specific questions:

• What evidence shows that learning this specific procedure improves the reading and spelling outcomes my child needs?

• How much intervention time is being spent learning and practicing the procedure itself?

• Can my child use it independently while reading real text?

• Does my child have to know the word before knowing how to divide it?

• What happens when the syllable boundaries cut through meaningful morphemes?

• How are reduced vowels and schwa handled?

• How is dialect handled?

• How are morphology and word families incorporated?

• Is my child learning why words are spelled as they are, or mainly learning how to label them?

• Is the procedure helping my child understand English, or is my child becoming better at performing the procedure?

I don’t want struggling readers doing more work than the writing system requires

Children with dyslexia have already spent enough time feeling as though English is an endless collection of things they failed to remember: another rule, another exception, another marking system, another list, another step, another reminder to “flex.”

Then we tell them English is irregular.

I don’t think that is an accurate description of what they are trying to learn.

English is complex, but complexity is not the same thing as chaos. Words carry histories. They belong to families. Morphemes retain spellings across changes in pronunciation. Affixes carry grammatical and semantic information. Stress affects pronunciation. Vowels reduce. Graphemes operate inside morphological structure.

There is plenty for children to investigate without adding another layer of written categories for them to memorize.

I am not asking children to do less language work. I am asking them to spend that effort learning how the language and the writing system actually work.

I don’t want a child using precious cognitive resources to remember where someone decided to draw a line through a word.

I want that child learning why the word is spelled the way it is.

Once you have shown a struggling reader the difference between:

<hope + ing> → hoping

and

ho | ping

it becomes very difficult to look at written syllable types the same way again.

References

Bowers, P. N., Kirby, J. R., & Deacon, S. H. (2010). The effects of morphological instruction on literacy skills: A systematic review of the literature. Review of Educational Research, 80(2), 144–179. https://doi.org/10.3102/0034654309359353

Buckingham, J. (2024). What is at the heart of the Science of Reading for teachers? Nomanis, 16.

Colenbrander, D., von Hagen, A., Kohnen, S., Wegener, S., Ko, K., Beyersmann, E., Behzadnia, A., Parrila, R., & Castles, A. (2024). The effects of morphological instruction on literacy outcomes for children in English-speaking countries: A systematic review and meta-analysis. Educational Psychology Review, 36, 119. https://doi.org/10.1007/s10648-024-09953-3

Dechant, E. (1969). Linguistics, Phonics, and the Teaching of Reading. Charles C Thomas.

Elbro, C., & Arnbak, E. (1996). The role of morpheme recognition and morphological awareness in dyslexia. Annals of Dyslexia, 46, 209–240.

Goodwin, A. P., & Ahn, S. (2010). A meta-analysis of morphological interventions: Effects on literacy achievement of children with literacy difficulties. Annals of Dyslexia, 60, 183–208.

Goodwin, A. P., & Ahn, S. (2013). A meta-analysis of morphological interventions in English: Effects on literacy outcomes for school-age children. Scientific Studies of Reading, 17(4), 257–285. https://doi.org/10.1080/10888438.2012.689791

Kearns, D. M. (2015). How elementary-age children read polysyllabic polymorphemic words. Journal of Educational Psychology, 107(2), 364–390. https://doi.org/10.1037/a0037518

Kearns, D. M. (2020). Does English have useful syllable division patterns? Reading Research Quarterly, 55(S1), S145–S160. https://doi.org/10.1002/rrq.342

Kearns, D. M., Lyon, C. P., & Kelley, S. L. (2022). Structured Literacy interventions for reading long words. In L. Spear-Swerling (Ed.), Structured Literacy Interventions: Teaching Students with Reading Difficulties, Grades K–6. Guilford Press.

Kearns, D. M., & Novelli, C. (2026). Are early readers sensitive to the patterns in bisyllabic English words? Annals of Dyslexia. https://doi.org/10.1007/s11881-026-00371-4

Shanahan, T. On Eating Elephants and Teaching Syllabication. Shanahan on Literacy.

Tortorelli, L. S., Strong, J. Z., & Anderson, B. E. (2024). Multisyllabic decoding achievement and relation to vocabulary at the end of elementary school. Journal of Experimental Child Psychology, 246, 106018.

Bowers, P. N. (2025). Peter Bowers on Syllabication and Morphology—Response to Feb. 22, 2025 Blog.

The Phoneme Do-Si-Do

Why advanced phoneme manipulation is not a prerequisite for learning to read

I came across a post introducing a “phoneme warm-up” with this description:

Sixty seconds. No letters on the table.

Six phonemic tasks in difficulty order. Run them until the group breaks. Where they break is your next two weeks.

If a card hits the table, it stopped being a phoneme task.

Which one does your group fail first: deletion or substitution?

And I thought: Warm-up?

For many of the dyslexic learners I work with, this is not a warm-up. This is the main freaking endurance workout.

Look at what they are being asked to do before the literacy lesson has supposedly even begun:

1. Isolate the final sound.

2. Blend four phonemes with no pause.

3. Segment with chips.

4. Delete the first sound.

5. Substitute the medial vowel.

6. Reverse a two-sound word.

I can practically hear the square-dance caller.

Isolate that final sound!
Four phonemes, blend ’em round!
Segment left and segment right!
Take that first one out of sight!
Swap that vowel, here we go!
Turn ’em backwards, do-si-do!

Swing your phoneme, round you go. Hold it, move it, don’t be slow. Drop that /s/ and keep the rest. Change the middle. Pass the test. Reverse those sounds and off you go.

And if the child breaks down?

Apparently that is the lesson plan for the next two weeks.

For children who are already very good at this particular kind of metalinguistic manipulation, the routine probably looks terrific. They can hold a sequence of phonemes in working memory, consciously isolate one, inhibit the original response, remove or replace a unit, maintain the rest of the sequence, and produce an answer quickly.

Those are the kids who can dance.

They also tend to be the children for whom learning to read and spell comes more easily.

That should make us ask a different question.

Are these tasks building the capacity that produces literacy, or are children who are developing literacy successfully also becoming increasingly proficient at consciously manipulating the phonological abstractions represented by alphabetic writing?

Those are not the same claim.

A predictor is not necessarily a prerequisite

This confusion has caused a great deal of trouble in reading instruction.

When performance on a task predicts later reading achievement, we have a tendency to turn successful performance on that task into something children must acquire before they can become successful readers. But prediction does not establish an instructional sequence.

Height predicts which children are more likely to become good basketball players. We would not conclude that basketball instruction should consist of making children taller.

Phonemic-awareness performance is associated with reading development. That is well established. What does not follow is that children must become proficient at progressively more difficult phoneme-manipulation tasks in order to become proficient readers. Castles and Coltheart (2004), in a review devoted specifically to the question of causality, concluded that the evidence did not establish the simple causal sequence in which phonological awareness develops first and then produces successful reading and spelling.

In fact, the developmental relationship goes in both directions. Perfetti and colleagues (1987) found a reciprocal relationship between phonemic knowledge and learning to read in first graders, including evidence that reading development contributed to growth in phoneme deletion.

Other studies make the role of literacy especially difficult to ignore. Morais and colleagues (1979) found that adults who had never learned to read alphabetically had considerable difficulty with phoneme addition and deletion, while adults from similar backgrounds who had acquired alphabetic literacy could perform the tasks. Read and colleagues (1986) found a similar pattern among Chinese adults: those literate only in Chinese characters had much greater difficulty consciously adding and deleting individual speech sounds than adults who also knew alphabetic writing.

That makes sense.

Phonemes are not little beads sitting inside spoken words waiting for children to discover them. Speech is continuous. Learning to consciously analyze speech as a sequence of abstract phonemic units is itself a metalinguistic achievement, and alphabetic writing gives us a representational system that helps make those units available for conscious analysis.

Explicit phoneme manipulation does not simply develop first and produce literacy. The relationship is reciprocal. Alphabetic literacy itself helps develop the conscious phonemic awareness we then turn around and treat as its prerequisite.

A diagnostic task is not necessarily an instructional task

This one deserves to be written on the wall of every intervention room:

What we use to diagnose does not necessarily belong in treatment.

A task may tell us something useful about how a learner is performing without becoming something we should spend instructional time practicing. Yet reading intervention repeatedly makes exactly that leap.

A child performs poorly on phoneme deletion, so we practice phoneme deletion. A child performs poorly on phoneme substitution, so we practice substitution. A child performs poorly reading nonsense words, so we give the child nonsense words to read.

Hello, nonsense words.

Pseudowords are widely used in reading research and assessment because they remove the contribution of a known word and ask the reader to deal with an unfamiliar letter string.

But even if we accept a task as diagnostically useful, it does not follow that practicing the task is treatment.

And I would go further: I am not particularly persuaded by nonsense-word assessment either. English words are not meaningless strings of grapheme-phoneme correspondences. Real words have meanings, structures, relatives, histories, frequencies, pronunciations, and orthographic patterns that interact. Removing all of that may create an interesting experimental or assessment condition, but we should be cautious about treating performance in that artificial condition as a pure measure of an underlying reading capacity.

An assessment task is not automatically an instructional target.

We have confused those two things far too often.

Getting better at phoneme gymnastics is not the same thing as getting better at reading

If I train children intensively on deletion, they may get better at deletion. If I train substitution, they may improve at substitution. If I repeatedly ask them to reverse two-phoneme words, they may become faster at reversing two-phoneme words.

But that is not my treatment outcome.

I care about reading and writing.

Clemens and colleagues (2021) examined the claims surrounding advanced phonemic-awareness training, including deletion and substitution, and questioned whether proficiency on these tasks is necessary for skilled reading or whether training them improves reading outcomes. Their paper is a critical review and preprint, not an intervention trial, but its central criticism is important: association between advanced phonemic-awareness performance and reading does not establish that advanced phonemic-awareness training is a necessary route to reading.

Castles and colleagues (2003) provide another reason to be cautious about interpreting these tasks as some independent measure of a child’s “phonological layer.” They found that orthographic knowledge influenced performance on phoneme deletion and reversal tasks. Even performance on supposedly oral phonological manipulations can be affected by what the learner knows about spelling.

And the broader phonemic-awareness research does not require us to deny that phonemic-awareness instruction can be useful. Ehri and colleagues (2001) found benefits of phonemic-awareness instruction for reading and spelling. Rice and colleagues (2022) also found positive effects of phonemic-awareness instruction on phonemic-awareness outcomes. Rehfeld and colleagues (2022), looking specifically at children suspected of reading disability, found different effects across phonemic-awareness skills, with a relatively small effect for deletion.

None of that establishes a required hierarchy culminating in deletion, substitution, and reversal.

Every minute I spend teaching a child to reverse up is a minute I am not spending helping that child understand an actual English word.

But what if we put letters on the table?

Someone will inevitably respond that the real problem with this routine is the instruction to use “no letters on the table.”

There is good reason to question that instruction. Written letters can provide concrete, stable support for phonological analysis.

But that is not my central objection.

Put the letters on the table.

I still do not want to spend intervention time having dyslexic children perform a phoneme do-si-do.

Doing deletion with print does not suddenly make deletion an important literacy goal. Moving letter cards around while substituting phonemes does not establish that substitution practice is necessary for learning to read. The presence of graphemes may reduce the memory demands of the task, but it does not answer the more important question:

Why are we spending instructional time on this task in the first place?

With or without print, I am not interested in training children to become increasingly proficient at isolated phoneme gymnastics.

For the children I work with, I can examine phonology in something much richer:

a word.

A real, meaningful word that we can see and say. A word whose meaning we can establish, whose morphological structure we can investigate, whose relatives we can connect, and whose graphemes we can analyze within that structure. A word we can put into sentences, encounter in text, read, spell, and write.

The phonology is still there.

I have not abandoned it. I have given it somewhere to live.

The child who “breaks”

The wording of the original post keeps bothering me:

“Run them until the group breaks. Where they break is your next two weeks.”

Think about the logic of that for the learner on the receiving end.

The child struggles with phoneme deletion.

Therefore, deletion becomes the intervention.

The child continues to struggle.

Therefore, we practice more deletion.

The task exposing the child’s greatest vulnerability becomes the activity we repeatedly ask the child to perform.

For some children, perhaps they improve quickly and everyone feels reassured that the hierarchy worked. But for many of the severe dyslexic learners who eventually arrive in my practice after years of intervention, this has become an instructional treadmill. They have spent enormous amounts of time listening harder, tapping sounds, segmenting sounds, blending sounds, deleting sounds, substituting sounds, and manipulating sounds because their continued difficulty performing those tasks has repeatedly been interpreted as evidence that they need more of them.

At some point, we have to be willing to question the task rather than the child.

If a learner cannot enter literacy through increasingly abstract manipulation of speech, perhaps the answer is not another two weeks of increasingly abstract manipulation of speech.

Perhaps we offer another entrance.

The phoneme do-si-do

So here comes the caller one more time.

Find the ending!
Blend ’em fast!
Pull ’em apart!
Drop the first!
Switch the middle!
Turn ’em round!

Around the phoneme floor we go.

For the children who can already do the dance, it looks beautifully efficient. For the children who cannot, we call their difficulty diagnostic and prescribe more dancing.

I think we have mistaken proficiency at a metalinguistic task for a prerequisite to literacy.

A predictor is not necessarily a prerequisite.

What we use to diagnose does not necessarily belong in treatment.

A diagnostic task is not necessarily an instructional task.

And getting better at phoneme gymnastics is not the same thing as getting better at reading.

The children who perform these manipulations easily are often also the children who learn to read and spell relatively easily. That does not prove the gymnastics caused the literacy. Research on the reciprocal relationship between literacy and phonemic awareness gives us good reason to consider the other direction as well.

For the children who find this work extraordinarily difficult, I see no reason to make successful completion of the dance the price of admission.

I would rather put a word on the table.

And begin.

References

Castles, A., & Coltheart, M. (2004). Is there a causal link from phonological awareness to success in learning to read? Cognition, 91(1), 77–111. https://doi.org/10.1016/S0010-0277(03)00164-1

Castles, A., Holmes, V. M., Neath, J., & Kinoshita, S. (2003). How does orthographic knowledge influence performance on phonological awareness tasks? The Quarterly Journal of Experimental Psychology Section A, 56(3), 445–467. https://doi.org/10.1080/02724980244000486

Clemens, N. H., Solari, E. J., Kearns, D. M., Fien, H., Nelson, N. J., Stalega, M. V., Burns, M. K., St. Martin, K., & Hoeft, F. (2021). They say you can do phonemic awareness instruction “in the dark,” but should you? A critical evaluation of the trend toward advanced phonemic awareness training [Preprint]. PsyArXiv. https://doi.org/10.31234/osf.io/ajxbv

Ehri, L. C., Nunes, S. R., Willows, D. M., Schuster, B. V., Yaghoub-Zadeh, Z., & Shanahan, T. (2001). Phonemic awareness instruction helps children learn to read: Evidence from the National Reading Panel’s meta-analysis. Reading Research Quarterly, 36(3), 250–287. https://doi.org/10.1598/RRQ.36.3.2

Morais, J., Cary, L., Alegria, J., & Bertelson, P. (1979). Does awareness of speech as a sequence of phones arise spontaneously? Cognition, 7(4), 323–331. https://doi.org/10.1016/0010-0277(79)90020-9

Perfetti, C. A., Beck, I., Bell, L. C., & Hughes, C. (1987). Phonemic knowledge and learning to read are reciprocal: A longitudinal study of first-grade children. Merrill-Palmer Quarterly, 33(3), 283–319.

Read, C., Zhang, Y.-F., Nie, H.-Y., & Ding, B.-Q. (1986). The ability to manipulate speech sounds depends on knowing alphabetic writing. Cognition, 24(1–2), 31–44. https://doi.org/10.1016/0010-0277(86)90003-X

Rehfeld, D. M., Kirkpatrick, M., O’Guinn, N., & Renbarger, R. (2022). A meta-analysis of phonemic awareness instruction provided to children suspected of having a reading disability. Language, Speech, and Hearing Services in Schools, 53(4), 1177–1201. https://doi.org/10.1044/2022_LSHSS-21-00160

Rice, M., Erbeli, F., Thompson, C. G., Sallese, M. R., & Fogarty, M. (2022). Phonemic awareness: A meta-analysis for planning effective instruction. Reading Research Quarterly, 57(4), 1259–1289. https://doi.org/10.1002/rrq.473

The latest round of public argument about Reading Recovery has me thinking less about Reading Recovery itself and more about what happens when professionals disagree in public. Lyn Stone recently wrote about the silence that often surrounds these debates. People will tell you privately exactly what they think. They will agree in messages, emails, conference hallways, and conversations after presentations. Then the argument moves into public view and they disappear.

That hesitation is understandable. Speaking publicly can carry reputational risk, strain professional relationships, and, for people who teach courses, consult, write, present, or operate businesses, create commercial risk as well. It is considerably safer to wait quietly and see which way the professional winds blow.

But silence has consequences too. When knowledgeable people who have legitimate concerns refuse to say what they think publicly, contested ideas begin to look uncontested. A small number of people end up carrying the burden of questioning them while everyone else watches from the sidelines. That is not particularly healthy for a profession that claims to be evidence-informed.

Professional Courage Is More Than Speaking Loudly

Professional courage is not simply the willingness to criticize an approach you already disagree with. The harder test comes when evidence challenges something you have previously defended, especially when the uncomfortable question concerns your own theoretical camp, your own colleagues, your own training, your own course materials, or an approach into which you have invested considerable time, money, reputation, and professional identity.

That is when principles actually become principles. It is relatively easy to demand that Reading Recovery proponents reconsider Reading Recovery if you have never been invested in Reading Recovery. It is considerably harder to apply exactly the same standard of scrutiny when someone begins questioning an approach you believe in.

The only thing worse than remaining silent in the face of legitimate criticism may be becoming an apologist when that criticism lands too close to home. If we believe instructional practices should withstand scrutiny, then they all have to withstand scrutiny, including the ones we prefer.

Public Argument Can Be Excellent Professional Development

One of the most valuable things about substantive public disagreement is that it forces precision. When someone knowledgeable challenges your argument, you cannot rely on vague language or familiar terminology and assume everyone knows what you mean. You have to figure out exactly what you are claiming and where the disagreement actually lies.

Is the disagreement about evidence? Is it about the nature of English orthography? Is it about what counts as linguistic information, instructional sequence, or what learners need to know? Are two people using the same terminology while operating from fundamentally different assumptions?

Those are the kinds of questions that sharpen professional thinking.

If you spend enough time publicly debating knowledgeable proponents of an approach that differs from your own, you begin to understand your own position much more clearly. You have to explain what you do, how it differs from what they do, and why those differences matter. You may discover that two approaches you once considered similar are actually based on very different assumptions. You may realize that an argument you have been making is too broad, that your language is imprecise, or that you need better evidence for a claim.

Sometimes you walk away thinking, “I need a better way to explain that.” Sometimes you go looking for research because someone has identified a weakness in your argument. Sometimes you change your wording. Sometimes you change your thinking.

That is professional development.

The purpose of disagreement, however, cannot simply be to make us increasingly skilled at defending whatever we already believe. If every challenge merely becomes an opportunity to strengthen the fortress around an existing position, we are not really engaged in inquiry. We are engaged in advocacy. Genuine professional dialogue requires the possibility that the person disagreeing with you may have identified something you have overlooked.

Debate Is Not the Same Thing as Trolling

Valuing public disagreement does not mean treating every argument as equally worthwhile. There is an enormous difference between good-faith disagreement and rhetorical gamesmanship.

If someone challenges you in good faith, they engage the argument you actually made. If you raise three substantive points, they address those points. They tell you where they think your evidence is inadequate, where your reasoning fails, or what you have overlooked. They make their own case, and they allow you to respond to it. You may strongly disagree with one another, but both sides are participating in the same conversation.

Those exchanges can be tremendously valuable, especially when the person on the other side knows the subject well enough to make you work.

Bad-faith argumentation looks very different. You raise a substantive point, and the response ignores it and substitutes a different question. You answer that question, only to have the discussion shift again. Evidence already provided goes unaddressed while new demands for evidence appear. An appeal to authority replaces an argument. A false equivalency is introduced. A straw man becomes easier to attack than what you actually said. The goalposts move. A stream of ostensibly polite questions begins to function less like genuine inquiry and more like sealioning.

If the substantive argument is not going well, the conversation may eventually shift from the claim to the person making it. Motives are questioned. Tone becomes the issue. Labels appear. Sometimes the name-calling comes next.

At that point, the problem is no longer disagreement. The problem is that one side is not actually participating in the disagreement.

If this happens in your own professional space, you are not obligated to provide unlimited access to your time, your audience, or your platform in the name of “open dialogue.” Removing someone who repeatedly misrepresents what you have said, refuses to address repeated answers, or uses your comment section primarily to provoke and derail is not the same thing as avoiding criticism.

You can welcome disagreement without volunteering to be trolled.

Professional Investment Can Make Good-Faith Discussion Harder

Professional and commercial investment make these conversations even more complicated. If you have built a course, program, consultancy, organization, or professional reputation around a particular position, there is naturally more at stake when that position is challenged.

Having an investment does not make your argument wrong. Most professionals have investments of one kind or another. We teach particular approaches, recommend particular practices, build expertise, create materials, and sometimes earn income from that work.

The problem begins when protecting the investment becomes more important than examining the criticism.

You can often see this shift in the conversation itself. A difficult question goes unanswered while a safer one is substituted for it. Evidence that does not support the preferred conclusion is ignored or reframed. The discussion repeatedly returns to familiar talking points rather than addressing the actual objection. Instead of asking, “Is this criticism valid?” the conversation becomes an exercise in figuring out how to defend what has already been built.

That is precisely when professional courage matters most. If you expect someone else to reconsider an approach despite having years of training, professional relationships, status, or income tied to it, then you have to be prepared to do the same when the scrutiny reaches your own doorstep.

Being Wrong Is Not the Professional Failure

Professional identity can make reconsideration difficult because approaches rarely remain abstract. You invest years learning them. You teach them, recommend them, write about them, present about them, and become known for them. Eventually, criticism of the approach can begin to feel like criticism of you.

But an instructional approach is not a person, and changing your mind about an instructional practice does not mean that everything you previously believed or did was foolish.

Knowledge changes. Evidence changes. Experience changes our understanding. You encounter learners who do not behave the way a theory predicts they should. You learn things your original training did not teach you. You find better explanations. You discover limitations in practices you once believed were sufficient.

That should happen over the course of a professional life.

Being wrong is not the professional failure. Refusing to reconsider because too much of your identity depends upon being right is far more concerning.

Silence Can Distort a Field

There is another reason public disagreement matters: other people are watching these conversations because they are trying to learn.

Teachers, tutors, SLPs, parents, researchers, administrators, and interventionists are trying to determine what they think. When disagreement happens only privately, they never see it. They see the published position, the conference presentation, the influential organization, the popular program, or the person with a large professional following. What they do not see are all the knowledgeable people privately saying, “I don’t think that is right.”

Public criticism changes that informational landscape. It tells people that a claim is contested. It gives them competing arguments to examine. It exposes assumptions that might otherwise remain hidden and allows observers to evaluate how each side responds when its claims are challenged.

That last part may be particularly revealing. The quality of an argument tells you something, but so does the willingness to engage criticism directly.

Can someone answer the objection that was actually raised? Can they acknowledge a legitimate point made by someone on the other side? Can they distinguish evidence from authority? Can they say, “I was wrong about that”? Can they revise a position without treating revision as humiliation?

Those are not signs of weakness. They are signs that a person is still participating in inquiry.

Put Your Reasoning on the Table

The most useful professional opponents are often the ones who understand your argument and still think you are wrong. They force you to make the strongest version of your case. They ask you to show your evidence, explain your reasoning, and account for things your preferred explanation may not adequately address. In return, you should be able to ask the same of them.

That is how professional thinking develops. It does not develop through universal agreement, carefully curated conversations in which everyone politely avoids the difficult question, or private messages saying, “I completely agree with you,” followed by silence when the discussion becomes public.

Professional courage means putting your reasoning where other people can examine it, but it also means accepting the reciprocal obligation that comes with doing so. If you demand that other people reconsider their positions when evidence challenges them, you have to be willing to reconsider yours. If you expect Reading Recovery proponents to interrogate long-standing assumptions, you cannot suddenly decide that scrutiny is unfair when someone begins interrogating assumptions within an approach you favor.

No instructional approach should be protected by professional loyalty, including our own.

Give me a healthy debate any day of the week. A serious opponent who understands the argument and thinks you are wrong may teach you something. Someone who has no intention of engaging with the argument in the first place has very little to offer.

The goal is not agreement. The goal is intellectual honesty, and that requires more than being willing to speak. It requires being willing to listen, answer the question that is actually on the table, and change your mind when the argument against you turns out to be better than the one you brought with you

One of the questions I’m asked most often about Structured Word Inquiry is also one of the hardest to answer briefly: What kind of approach is it?

SWI is not a phonics program, but it is also not whole language or balanced literacy. It is not simply morphology instruction with phonics added later. Inquiry is central to it, but even “inquiry-based literacy” does not quite explain what distinguishes the approach.

The clearest description is this: Structured Word Inquiry is an explicit, analytical, language-based approach to literacy that begins with meaning and word structure, then examines how spelling represents pronunciation within that structure.

More simply, it is a meaning- and structure-first approach. Phonology is taught explicitly, but it is not required to carry the entire burden of getting the learner into the word.

That distinction is especially important for children with dyslexia.

Start with the word

Many beginning and remedial reading programs start by breaking language into increasingly small pieces. Children learn letter-sound correspondences, practice segmenting and blending, study syllable types and division procedures, and are then expected to use those procedures to identify unfamiliar words.

For some children, that sequence works reasonably well. For others, particularly those with significant phonological and sequential processing difficulties, the process itself becomes the obstacle.

SWI begins somewhere else: with the word.

The child is told what the word is. We talk about what it means and how it is used. We investigate how it is built, identify its relatives, write word sums, compare forms, and notice what stays stable and what changes. Once we have established the word’s meaning and morphological structure, we examine what the letters are doing.

That includes phonology. We can identify which graphemes represent which phonemes, compare pronunciations across a family, use IPA when greater precision is useful, and investigate cases in which pronunciation changes while spelling remains stable.

Phonology has not been removed from instruction. It has been placed inside a meaningful linguistic structure.

Giving phonology something to hang onto

For many children with dyslexia, phonological processing is one of the least stable parts of the system. Segmenting spoken words, blending phonemes, holding sound sequences in working memory, and coordinating those sequences with written forms can require enormous effort.

Yet conventional intervention often makes proficiency in those very skills the gateway to everything that follows. A child may spend weeks or months working on sounds, blends, syllable types, and decoding procedures before being given access to much of the vocabulary, morphology, grammar, and word history that could actually make the writing system more understandable.

When the child does not progress, instruction often becomes more concentrated around the same bottleneck. There are more flashcards, more blending practice, more segmentation, another syllable rule, another marking system, or a new verbal prompt for the same procedure.

At some point, it is worth questioning not only how the procedure is being taught, but whether mastery of that procedure needs to come first at all.

When sounding out becomes the bottleneck

Consider what we ask a struggling reader to do when we present an unfamiliar multisyllabic word and expect the learner to decode it through a syllable-based procedure. The child may need to locate vowels, determine syllable types, decide where to divide the word, assign pronunciations, hold those pieces in memory, blend them, and then determine whether the result is a word they recognize.

That is a considerable processing load, particularly for a learner whose phonological, sequential processing and working memory are already vulnerable.

Something revealing often happens with these students. Once someone tells them what the word is, they can sometimes go back and apply the taught syllable procedure successfully. They can mark it, divide it, label its syllable types, and explain the rule.

But that raises an important question: did the strategy allow the child to identify the word, or did knowing the word allow the child to apply the strategy?

Those are very different outcomes.

Written syllable-type instruction can also narrow attention to pronunciation while obscuring other information carried by the spelling. English words are not simply strings of graphemes arranged to encode sound. Spellings also represent morphological structure and relationships among words. When instruction begins and ends with the problem of how to pronounce the letter string, much of that information remains unavailable to the learner.

SWI does not require pronunciation to do all of the work.

Access to the whole writing system

Take a word such as action. A learner does not need to independently sound out action before we can teach anything useful about it.

We can establish what the word means and use it in sentences. We can investigate its relationship to act, acts, acted, acting, active, and activity. We can determine which morphemes are present and compare how the base behaves across the family. Only after establishing that structure do we examine the graphemes and the pronunciations they represent.

The learner is therefore doing far more than trying to recover a spoken word from a sequence of letters. They are using vocabulary, morphology, grammar, spelling, pronunciation, and prior knowledge together. They can compare evidence, recognize relationships, test ideas, and explain what they notice.

For a child who finds phonological processing difficult, this changes the entry point. The hardest part is still taught, but it is no longer the entrance exam for everything else.

Explicit instruction does not require a rigid sequence

SWI is sometimes mischaracterized as a form of whole language or balanced literacy because it does not organize instruction around a traditional phonics scope and sequence. That comparison misses a fundamental difference.

SWI is explicit. We explicitly teach what bases, prefixes, and suffixes are. We explicitly investigate grapheme-phoneme correspondences. We teach suffixing conventions, examine spelling changes, distinguish morphological structure from phonological structure, and require conclusions to be supported by evidence.

Children are not asked to guess words from pictures or context. If a learner does not know a word, we can simply tell them the word and then investigate it. Context helps establish meaning; it is not used as a substitute for analyzing the spelling.

What SWI does not require is a universal sequence in which every learner must demonstrate mastery of one isolated skill before being permitted to encounter the next one. A child can continue learning vocabulary, morphology, syntax, spelling, reading, and writing while phonological abilities are still developing. Difficult skills can be revisited repeatedly within increasingly rich language rather than functioning as gates that stop the rest of instruction.

Literacy as inquiry

One useful way to explain SWI to parents and teachers is to describe it as inquiry-based learning applied to reading and spelling.

Instead of treating the writing system as a collection of rules to memorize and procedures to perform, we give children real linguistic questions to investigate. They compare words, gather evidence, consider possible explanations, revise an analysis when necessary, and justify their conclusions.

Why is there a g in sign? How are please and pleasant related? Why can the pronunciation of a base change while its spelling remains stable? Where are the morpheme boundaries in a word, and what does that tell us about the graphemes? Why does a suffix have the spelling it does?

These are not peripheral questions. They are questions about how the English writing system actually works.

This kind of instruction also gives struggling readers access to intellectual work they are often denied. A child who cannot yet blend a difficult sequence efficiently may still be very capable of understanding a concept, identifying a meaningful relationship, recognizing a pattern, or reasoning from evidence. SWI allows those abilities to participate in literacy learning rather than placing the child’s weakest skill in charge of the entire lesson.

Why this can work so well for children with dyslexia

For many dyslexic learners, the advantage of SWI is not that it avoids phonology. It is that phonology no longer has to stand alone.

Meaning and morphological structure provide more stable information around the pronunciation. The learner knows what word is being investigated, understands something about what it means, and can see how it connects to other words before being asked to analyze the speech represented by the spelling.

That gives the child more than one source of linguistic information to work with and allows phonological knowledge to develop within meaningful, repeated encounters with real words.

It also means that a learner does not have to remain indefinitely in beginning-level material simply because segmenting or blending is still difficult. A ten-year-old can investigate sophisticated words and ideas while continuing to work on phonological analysis. An adolescent can study morphology, etymological relationships, vocabulary, syntax, and spelling without first proving mastery of a sequence designed for beginning readers that is based on how strong readers learn.

The instruction can meet the learner where the learner actually is: strong in some areas, vulnerable in others, and capable of continuing to learn across all of them.

Begin with something the learner can understand

For children who have struggled for years, the difference can be substantial.

Instead of beginning every lesson with another demand to perform the skill that has repeatedly failed them, we can begin with language they know. We can talk about a word, connect it to experience, investigate its family, notice its structure, use it in sentences, read it in connected text, and then examine its pronunciation and spelling with precision.

The child is not being asked to guess. The child is not being protected from difficult linguistic work. And the child is not being asked to memorize the writing system through exposure alone.

The child is being taught to study it.

That is the shift.

Start with the word. Establish its meaning and structure. Then investigate what the letters are doing.

For many children with dyslexia, that gives phonology something it has been missing all along: a structure to hang onto.

What If My Child Hasn’t Made Progress in Reading or Spelling?

What would truly different look like?

I recently came across a practitioner asking for help with a student who had been receiving the Wilson Reading System for years at a specialty school for dyslexia.

Years.

According to the practitioner, the student had been kept working on one-syllable words because she continued to miss one or two items on the program’s post-tests. Meanwhile, she kept reading on her own and learning words anyway.

The practitioner said it herself: “She basically figures out the words and then divides instead of the other way around.”

And I sat there thinking, Well, yes.

Because when I was a grown adult trying to understand how written syllable types and syllable-division procedures were supposed to help struggling readers figure out unfamiliar words, I came to exactly the same problem. How do you reliably know how to divide a word if you don’t already know what the word is?

Take a common vowel-consonant-vowel pattern. A child may be taught that the consonant can go with the second syllable, giving us a word like tiger, or with the first syllable, giving us a word like river. How do you know which one works? You try a pronunciation and see whether it produces a word you recognize.

In other words, you need to know something about the word—its pronunciation, its meaning, whether it makes sense. Once you know that, you can go backward and explain how the word should be divided.

And that is exactly what this student was doing. She figured out the word and then performed the procedure that was supposedly meant to help her figure out the word.

So what work was the syllable division actually doing?

When she encountered a word she genuinely did not know, according to the practitioner, she still sometimes stopped and said, “I have no idea what that word is.” After years of specialized dyslexia intervention.

And the question being asked was how to get her to practice syllable division more. Someone suggested making it more fun by “karate chopping” the syllables.

I kept thinking maybe we are asking the wrong question. Maybe the question isn’t how do we make her better at the procedure? Maybe it is why are we still so committed to the procedure?

When the procedure becomes the gate

This story becomes even more interesting when you look at the way Wilson handles progression.

Wilson says progress monitoring is used to ensure that students master the skills in each substep before moving on to new ones. Its End-of-Step Assessments include word reading, spelling, fluency and comprehension, but also the student’s ability to mark the concepts currently being taught. Students who do not meet the required mastery criteria receive more instruction before proceeding.

So a student can be reading words and still have program concepts she has not sufficiently mastered. And that is where I start getting uncomfortable, because reading and spelling are supposed to be the outcomes.

If a child can recognize the word but cannot correctly perform our analysis of how she was supposed to recognize the word, what exactly has she failed? If she figures out the word and then divides it, has she demonstrated a reading problem, or has she demonstrated difficulty with a particular metalinguistic procedure we have decided a reader should use?

Those are not the same thing.

A procedure designed to support reading can gradually become something the child must master in its own right. And sometimes the child ends up serving the program instead of the program serving the child.

Then parents are told to ask for Orton-Gillingham

Now imagine being the parent. Your child has struggled with reading and spelling for years. You have watched other children take off while yours continues to labor over words. Eventually someone says dyslexia.

You start researching. You join parent groups. You talk to other families. Maybe you get a private evaluation. And one recommendation comes up again and again: Orton-Gillingham.

Your child needs OG. Ask the school for OG. Find an OG tutor. Make sure the intervention is OG-based.

I understand why parents latch onto it. They are trying to find something that will finally be different.

But before asking for a particular program or approach, I think there is a much more important question: What has my child already spent the last several years doing?

Not what was the curriculum called. What did the child actually do?

Because once you look underneath the program names, there can be a lot of overlap.

What is already being taught in school phonics?

School phonics programs vary. I am not suggesting that every school teaches the same curriculum or that all classroom phonics is Orton-Gillingham.

But look at the kinds of things children may already be doing in contemporary foundational reading programs. They may spend substantial instructional time:

• segmenting spoken words into phonemes;

• blending phonemes into words;

• manipulating sounds;

• learning grapheme-phoneme correspondences;

• practicing consonant digraphs and blends;

• learning short- and long-vowel patterns;

• learning vowel teams;

• working with written syllable types;

• dividing longer written words into syllables;

• learning spelling patterns and generalizations;

• practicing irregular words;

• reading decodable or controlled text;

• and progressing through a systematic sequence with extensive review of previously taught material.

UFLI Foundations, for example, uses an eight-part lesson routine that includes phonemic awareness, visual and auditory drills, a blending drill, a new phonics concept, word work, irregular words, and connected text. Its first unit on reading longer words explicitly introduces syllables, closed/closed patterns, and open and closed syllables.

The programs differ. The sequences differ. The lessons differ. But these are familiar instructional targets in many schools.

Now compare that with Orton-Gillingham. The Orton-Gillingham Academy’s curriculum handbook includes phonemic blending and segmenting, sound-symbol associations, at least six syllable types, basic syllable-division patterns, spelling rules and generalizations, decoding and encoding, multisensory practice, cumulative instruction, and overlearning. Morphology, vocabulary and comprehension are included as well.

I am not saying school phonics and OG are identical. They are not.

I am asking parents to notice that a child can move from classroom instruction into OG while many of the central things being taught remain very familiar: segmenting, blending, sound-symbol relationships, vowel patterns, syllable types, syllable division, spelling generalizations, and repeated practice toward automaticity.

So if those are the very things your child has struggled to learn and use for years, it is reasonable to ask: What is actually going to be different?

More intensive is not the same as different

Orton-Gillingham may be taught one-to-one rather than in a classroom. That can make a tremendous difference. The practitioner may be much more knowledgeable. The pace may be slower. Instruction may be more diagnostic and responsive. There may be more review, more repetitions, more immediate correction, and far more opportunities to respond.

For some children, that is exactly what was missing. A child who received weak classroom instruction may need the same foundational content taught much better, much more carefully, and much more intensively.

But that is not the same child as the one who has already had good, explicit instruction and still struggles. And it is certainly not the same child as the one who has already spent years in intensive OG-based intervention and still struggles.

At some point, we have to distinguish between changing the delivery of instruction and changing the framework itself.

A smaller group changes delivery. One-to-one teaching changes intensity. A slower pace changes delivery. More repetitions increase intensity. A more responsive practitioner can dramatically improve the quality of teaching. All of those changes may help.

But if the child is still spending a substantial part of intervention segmenting, blending, tapping, identifying syllable types, dividing written words, applying pronunciation rules, practicing spelling generalizations, and demonstrating mastery before moving forward, then the basic way we are asking that child to understand written language may not have changed much at all.

Sometimes intensifying that framework works. But when it doesn’t, why is our answer so often simply more? More phonemic awareness, more blending, more syllable practice, more repetition. Another mnemonic, another game, another level, another year.

At some point, we have to allow another possibility: Maybe this child needs a different framework.

There are two very different questions we can ask

We can ask, How can I teach this procedure better? Or we can ask, Does this procedure need to organize this child’s intervention in the first place?

Those questions can lead us in completely different directions.

If a child cannot remember syllable types, we can find better ways to teach syllable types. If syllable division does not stick, we can clap it, scoop it, color-code it, tap it, karate chop it, and practice it another hundred times.

Or we can ask why syllable division is occupying so much instructional space for a learner who has already spent years struggling with it.

And this is not a criticism unique to SWI. SPELL-Links, another language-literacy intervention, explicitly contrasts its treatment of syllables in spoken language with what it describes as the “man-made conventions for printed words” used in OG-based programs, where substantial attention is given to marking letters and using visual patterns to determine where written words should be divided.

I keep returning to the practitioner’s observation: She figures out the word and then divides it.

That is not a minor detail. The procedure was supposed to help her get to the word. Instead, she gets to the word and then uses her knowledge of the word to perform the procedure.

Rather than immediately trying to fix that, I want to know something else: How did she figure out the word?

Start with what the learner knows

What did she know about that word? Had she heard it before? Did she know what it meant? Did the sentence help her recognize it? Did she recognize a meaningful part? Did she know a related word? Could she recognize a base, prefix, or suffix? Did something about its spelling connect with another word she already knew?

I want to know all of that, not because I am looking for a bigger collection of “strategies” she can use to attack an unfamiliar word. I am not interested in replacing syllable division with fourteen other tricks for figuring words out.

I want the learner to know more about the writing system itself.

That is a very different goal.

Structured Word Inquiry gives us another framework

One genuinely different framework is Structured Word Inquiry, or SWI.

And before I explain what that looks like, there is something important to understand. A child can have tremendous difficulty reading and spelling a word while knowing perfectly well what that word means.

Many dyslexic learners have plenty to work with at the level of concepts and meaning. They know words. They have ideas. They can use words in conversation that they cannot reliably read or spell.

For those learners, one of the hardest pieces is often precisely what phonics instruction tends to put near the beginning: working out the relationships between graphemes and phonemes.

So I keep coming back to a very simple question: Why would I always begin with the part that is hardest for the learner?

Why not begin with something more concrete, something the child already knows, something meaningful, something interesting enough to talk about, something they can actually grab onto?

That is one of the things I love about SWI. We can begin with a word, a real word, a meaningful word, often a word the learner already knows and uses.

Then we investigate it. What does it mean? How is it built? What other words are related to it? And then, what are the letters doing in this word?

That order changes a lot.

We begin with meaning and concepts. We investigate structure. We connect the word to relatives. We give the learner a network of knowledge about the word and the system it belongs to. Then we get to the sounds.

And now the sounds have somewhere to go.

The phonology has something to attach to.

Meaning is there from the beginning. Morphological structure is there. Related words are there. Grammar can enter the conversation. And yes, phonology is there too. We absolutely investigate grapheme-phoneme relationships.

For a learner who has spent years starting with sounds, struggling with sounds, practicing sounds, forgetting sounds, and then receiving still more sound-based practice because the previous practice did not stick, that is not a small change. It is a completely different way into written language.

I am not teaching word-solving strategies

This is an important distinction. I do not teach children a menu of strategies to pull out when they get stuck on a word: look for a chunk, divide it this way, try another vowel sound, cover this part, circle that part, try this trick.

That is still putting the learner in the position of asking, Which procedure am I supposed to perform now?

My goal is different. I want the learner to understand more and more about how the English writing system works.

I want them to know that written words have meaningful structure, that bases can occur across families of related words, that prefixes and suffixes contribute meaning and grammatical information, that spellings often remain stable across a morphological family even when pronunciation changes, that graphemes operate within morphemes, that the pronunciation of a spelling depends upon the word we are actually investigating, and that English spelling represents meaning and structure as well as pronunciation.

The more the learner understands the system, the less dependent they are on somebody standing beside them telling them which procedure to perform.

How I apply SWI in language intervention

I am a speech-language pathologist, and the struggling readers I work with often have far more going on than an isolated difficulty decoding words. They may struggle with vocabulary, morphology, sentence comprehension, syntax, word retrieval, verbal memory, inference, oral expression, reading comprehension, spelling, or some combination of those things.

So I do not use SWI as a little morphology activity added onto the end of a conventional phonics lesson. I apply SWI within broader language and literacy intervention.

We might begin with one meaningful word the learner already knows. We talk about what it means and connect it to concepts and experiences. We investigate its morphological relatives and build a family around a stable base. We look at prefixes and suffixes and notice how morphology changes meaning and grammar. We formulate sentences. We read. We write. We find those words in connected text. And we investigate the graphemes and phonemes in the actual words we are studying.

Meaning, structure, grammar, pronunciation, reading, and writing all have something to connect to.

The child does not have to master every one-syllable pattern before being allowed to investigate an interesting multisyllabic word. Knowing the word is not cheating. Using meaning is not cheating. Recognizing a related word is not cheating.

Those are things the learner knows about language, and we can build more knowledge from there.

Instead of beginning every lesson by confronting the learner with the thing that has been hardest for years, we can begin with something they understand and grow the investigation outward.

Morphology is not just the extra stuff

For a long time, the hierarchy has been treated as almost self-evident: first phonology, then phonics, then increasingly complicated decoding, and later we can get to morphology, vocabulary, and meaning.

But that hierarchy deserves scrutiny.

A 2026 study in Reading Psychology compared a morphology-focused intervention with a phonology-focused intervention for fourth graders with dyslexia. The morphology-focused instruction was built around Structured Word Inquiry. The morphology group showed decoding-accuracy and spelling gains comparable to the phonology group while showing stronger vocabulary growth; the authors also found that initial phonological awareness was related to differential response to the interventions.

One study does not establish a new gold standard, and I do not want a new gold standard. I do not want every child with dyslexia automatically prescribed SWI any more than I want every child automatically prescribed OG. Children are different.

But the study puts pressure on the assumption that phonology must always occupy the center while morphology waits at the edges, especially for the learner who has already spent years working from sounds outward without making the progress everyone expected.

Is Orton-Gillingham really the gold standard?

Parents are often given the impression that this research question has already been settled. It has not.

A 2021 meta-analysis examined Orton-Gillingham-based interventions for students with or at risk for word-level reading disabilities. The authors found that OG interventions did not produce statistically significant improvements over comparison conditions on the combined foundational outcomes of phonological awareness, phonics, fluency, and spelling. They were careful not to conclude that OG does not work; rather, they concluded that the existing research did not yet allow strong claims about its effectiveness and called for better studies.

That distinction is important.

Many children make excellent progress with OG. If your child is thriving, there is no reason to take successful instruction away.

But the evidence does not justify treating the words Orton-Gillingham as though they end the intervention conversation. A dyslexia diagnosis does not automatically tell us which instructional framework will work for an individual child, and failure to respond to classroom phonics does not automatically mean the answer is a more intensive version of many of the same instructional targets.

Maybe it is. But maybe it isn’t.

Ask what is actually going to change

If your child has not made the progress you expected, I would not begin with, “Does my child need Orton-Gillingham?”

I would begin with, “What has my child already been taught, and what are we going to do differently?”

Ask what your child has practiced over and over and what they actually know. Ask whether the procedures they have been taught are doing the work those procedures were supposed to do. Ask which things have remained persistently difficult despite years of good instruction and practice.

Then ask about the proposed intervention. Is it different primarily because the group is smaller, because the pacing is slower, or because there will be more repetition? Those may be exactly the changes your child needs.

But ask whether anyone has also considered changing the framework.

Will your child spend another year practicing procedures for attacking words, or will somebody teach your child more about words and about the writing system those words belong to? What role will meaning have? What role will morphology have? What role will vocabulary and grammar have? Can instruction begin with words your child already knows and build outward? Can we use areas of strength to support areas of difficulty?

And if another six months go by without meaningful progress, will somebody reconsider the framework, or will your child simply get more?

The program is supposed to serve the child

I keep thinking about that student. Years of specialized dyslexia instruction. Held at one-syllable words because she had not demonstrated sufficient mastery. Meanwhile, she kept reading on her own and learning words.

And now she can look at a word, think about what it means, figure out what it is, and then go backward and divide it according to the procedure. And we are still asking how to get her better at the division.

Maybe she has been giving us useful information all along.

Children do not owe our programs mastery. Our programs owe children useful instruction.

If an approach is working, keep going. But if your child has spent years segmenting, blending, tapping, practicing sound-symbol correspondences, learning syllable types, dividing written words, applying spelling rules, working through controlled sequences, and repeating skills until mastery—and reading and spelling are still not becoming what they should be—you are allowed to question whether the answer is simply a stronger dose.

A genuinely different way of understanding written language.

Maybe the question we should have been asking all along is not how do we make this child succeed within this framework?

Maybe it is: What if this child needs a different framework?

A recent discussion started with a very good question: Why does <baby> become <babies>? More specifically, why do we use the <es> form of the plural suffix rather than simply adding <s>?

We can represent the present-day spelling this way:

baby + es → babies

Or, if we look further into the structure of the word:

babe + y + es → babies

We know the <y>/<i> suffixing convention is involved. The question is why this word takes <es> rather than <s>.

One explanation I encountered was that the <e> in <es> is there to “preserve” or “confirm” the vowel phoneme represented by <y> after <y> changes to <i>. I kept thinking about that explanation because I couldn’t see anything happening to the phoneme that needed to be preserved.

The grapheme changed. The phoneme didn’t.

Consider what actually happens in the pair:

baby → babies

In <baby>, final <y> writes /i/. In <babies>, after <y> changes to <i>, the <i> writes /i/.

The grapheme changed. The phoneme didn’t.

We see the same thing in:

try → tries

The <y> in <try> writes /aɪ/, and the <i> in <tries> writes /aɪ/. Again, the spelling changes while the vowel phoneme remains the same. That shouldn’t surprise us. <y> and <i> overlap considerably in the phonemes they can represent in English.

There is an even cleaner way to see that the <e> in <es> is not needed to preserve the vowel:

happy + ness → happiness

The <y> in <happy> writes /i/. When <ness> is added, <y> changes to <i>, and the <i> in <happiness> still writes /i/. There is no <e> following it to preserve, confirm, or rescue anything. <y> wrote /i; now <i>` writes /i/.

The same thing happens with another vowel:

deny + al → denial

The <y> of <deny> writes /aɪ/, and the <i> of <denial> writes /aɪ/. Once again, the grapheme changes and the phoneme doesn’t.

Whatever eventually explains the <e> in <babies>, we don’t need that <e> in order for <i> to continue representing the vowel previously represented by <y>. We can see that from other words.

<i> does this elsewhere too

We don’t even have to stay within the <y>/<i> convention. Consider words such as <immediate>, <mediate>, and <mediation>. A single <i> is perfectly capable of writing /i/ in English without an accompanying <e> serving as a phonological marker.

The assumption that a single <i> naturally wants to represent /ɪ/, and therefore something else must intervene when it represents /i/, comes from approaching the spelling through a familiar phonics framework. But letters do not have one inherent sound. <i> can write more than one phoneme. So can <y>. Their shared ability to write /i/ does not create a problem that another letter has to solve.

The original question is still there

None of this actually tells us why we write:

baby + es → babies

rather than:

baby + s → *babis

That was the original question, and it is a good one.

What we need to resist is the idea that because someone has asked why, we are obligated to immediately produce a because.

A good question does not obligate us to invent an answer.

We can identify a pattern before we know exactly how that pattern arose. We can describe what the evidence shows without claiming to have established its historical cause.

In present-day English, we can observe the singular <baby> and the plural <babies>. We can identify the <y>/<i> change. We can identify <es> as the form of the plural suffix that occurs here. We can also observe that the vowel represented by <y> in <baby> is represented by <i> in <babies>.

Saying that the <e> is there to “confirm” or “preserve” that vowel is a different kind of statement. It proposes a reason for the spelling, and that explanation requires its own evidence.

When we look backward in the language, we get an even better reason to be cautious.

English did not always distribute <i> and <y> the way it does today

Present-day English has fairly strong positional preferences for <i> and <y>. We generally do not expect native English words to end in vocalic <i>. That is a present-day convention, however, not something that has always been true.

Historical research on English spelling shows that the distribution of <i> and <y> changed substantially over time. Marco Condorelli’s work on their standardization describes a major shift during the sixteenth and seventeenth centuries toward the distribution we recognize now: vocalic <i> primarily in initial and medial positions, with <y> used for vowels word-finally.

Before English spelling became more standardized, <i> and <y> were much more variable. We can see that directly in surviving Middle English texts.

One of my favorite examples I found comes from Sir Gawain and the Green Knight. The manuscript gives us the singular:

<ladi>

At line 2497:

“þe luf of þe ladi…”

Then, only a few lines later, at line 2515, we find the plural:

<ladis>

“þat lordes and ladis þat longed to þe table…”

The same manuscript gives us <ladi> and <ladis>. There is no <e> there at all.

John Gower likewise gives us singular <ladi> and plural <ladis> in Middle English, while other Middle English texts show still other spellings, including forms corresponding to <ladies>. English writers were using several spellings while these conventions were developing.

That historical variation gives us a very different way to think about the present-day pair:

lady → ladies

The spelling did not emerge from a timeless rule requiring an <e> to protect or confirm the vowel. Earlier English could write <ladi> and <ladis>. Later, the positional distribution of <i> and <y> became increasingly standardized, and present-day English eventually settled on <lady> and <ladies>.

The spelling we have now is the outcome of that history. We cannot assume that every feature of the modern form was placed there to perform a present-day phonological job.

English spelling accumulated over time

Nobody designed English orthography at a conference. Nobody sat down with <baby> and decided:

“We’re going to change this <y> to <i>, but children might pronounce the <i> incorrectly, so we had better put an <e> after it to confirm the vowel.”

Writing systems do not develop that way.

English spelling accumulated across centuries. Pronunciations changed. Inflections changed. Spellings varied. Scribal practices varied. Printing introduced new pressures. Some distinctions disappeared while others became conventionalized. Gradually, a much more stable written system emerged.

The history of <i> and <y> shows that process particularly well because we can watch their distribution becoming more regular over time. The present-day pattern is the result of that history, not evidence that the same pattern has always existed.

That is why we should be cautious about working backward from a modern spelling and assigning every letter a contemporary sound-based purpose. Sometimes the history is messier than the phonics story we would like to tell.

So why <es> today?

The question is still there.

Why do present-day words ending in a single <y> use the <es> form of the plural suffix?

baby → babies
lady → ladies
puppy → puppies
try → tries

while words in which final <y> belongs to a vowel digraph take <s>:

day → days
boy → boys
key → keys

That is a visible pattern in present-day English. We can teach it, investigate it, compare words, and discover the generalization with learners. We do not have to assign the <e> a phonological function simply because we would like a more satisfying answer for why the pattern exists.

Perhaps there is scholarship that traces precisely how the modern <y> → <ies> plural pattern developed. Perhaps further historical investigation will show how the earlier variation settled into the convention we use today. But if we have not established that history, there is nothing wrong with saying so.

This is the pattern the evidence currently shows us. We can describe it reliably even if we have not established its historical cause.

That is not the same thing as asking children to memorize an arbitrary rule.

Learners can examine:

baby → babies
lady → ladies
puppy → puppies
try → tries

and contrast them with:

day → days
boy → boys
key → keys

They can observe what changes and what stays the same. They can test the pattern against more words, look for counterexamples, investigate the word structure, and then ask why.

Sometimes we will know why. Sometimes history will help us find out. Sometimes the best answer available will be, “I don’t know yet.”

Knowing where the evidence stops

Inquiry is not the production of increasingly elaborate explanations. It is the examination of evidence and the willingness to stop when the evidence stops.

With <baby → babies>, we know quite a lot. We know <y> and <i> can write the same phonemes. We know the vowel remains stable when <y> changes to <i>. We know this can happen without a following <e> because <happy → happiness> demonstrates it plainly. We know the modern distribution of <i> and <y> developed over time, and we can look directly at Middle English and find <ladi> beside <ladis>, with no <e> being used to “confirm” the vowel.

What we may not yet know is exactly why the modern plural system settled on <es> in this particular environment.

We do not need to be embarrassed by that gap or cover it with a story.

A good question does not obligate us to invent an answer.

Sometimes we just have to describe what is there and keep looking.

A recent Education Week article on what is being called “linguistic phonics” has generated a lot of discussion about speech-to-print instruction.

Much of that discussion has centered on whether speech-to-print is better than more traditional approaches to phonics and I think that is a reasonable question.

I also think it is a much smaller question than the one we need to be asking.

In a response to the article, linguist Lyn Stone made two observations that I think are particularly important. She wrote:

“English orthography is not a transcription system in which phonemes simply wait to be assigned one of several possible spellings.”

And later:

“Listing possibilities is not an account of the writing system.”

I agree.

Those two statements lead me to a larger question:

If English orthography is not fundamentally a transcription system, why are we still organizing literacy instruction around phonology?

Speech-to-print can remove a tremendous amount of unnecessary cognitive load

I understand why speech-to-print approaches are attracting attention.

Compared with traditional phonics instruction, and particularly with many Orton-Gillingham-derived approaches, they can remove a tremendous number of instructional pieces that add unnecessarily to cognitive load.

A child may no longer be expected to remember multiple syllable types, syllable-division procedures, vowel rules, markings, exceptions, and a long sequence of previously “mastered” concepts before attempting to read or spell a word.

Reading and spelling may be integrated from the beginning. Practice may be cumulative and interleaved. Children may encounter variation rather than being protected from it until some predetermined point in a scope and sequence. There may be less time spent classifying words and more time actually reading and writing them.

Those are meaningful improvements.

For a child who has spent years being asked to remember what kind of syllable they are looking at, where they are supposed to divide it, which vowel rule applies, whether an exception overrides that rule, and which procedure comes next, taking away all of that instructional machinery may be an enormous relief.

That machinery is not the writing system.

It is machinery we added to the teaching of the writing system.

Removing it can reduce extraneous cognitive load.

But reducing extraneous cognitive load is not the same thing as creating an instructional model that works for every learner.

And it does not remove the cognitive demands of the speech-to-print task itself.

Speech-to-print still asks the learner to do a lot

Consider what the learner may have to do when spelling from speech.

They have to access the spoken word.

They have to attend to or segment its phonological structure.

They have to hold that sequence in working memory.

They have to retrieve possible graphemes for those phonemes.

They have to inhibit competing possibilities.

They have to coordinate those possibilities with what they have previously learned.

They have to sequence the letters.

And when more than one grapheme could plausibly represent a phoneme, they have to determine which spelling belongs in this particular word.

Suppose the learner hears /iː/.

Perhaps they know that /iː/ can be represented by <ee>, <ea>, <e>, <y>, and other graphemes in different words.

That’s useful information, but the phoneme /iː/ cannot tell the learner whether this particular word requires <ee> or <ea>.

So the child chooses a plausible spelling.

Does that look like the word?

Does it match something they remember?

If not, try another possibility.

Guess.

Check.

Revise.

For a learner with strong oral language, a robust vocabulary, well-established lexical representations, adequate working memory, efficient lexical access, and a lot of experience with print, this process may become quite efficient.

The child already knows the word. The word has a strong representation in memory. Once enough orthographic information is available, the child may recognize which spelling belongs to it.

But notice how much the method is relying on the learner to bring to the task.

The child must hold information in mind while considering alternatives. They must inhibit competing responses. They must compare choices. They must recognize the intended written word. They may need to revise an initial response.

So while speech-to-print can remove the unnecessary cognitive demands imposed by traditional phonics and Orton-Gillingham procedures, the guess-and-check demands involved in choosing among plausible graphemes can still place considerable demands on working memory, lexical access, and executive functioning.

And for some struggling readers, those are precisely the systems that are vulnerable.

Removing unnecessary rules reduces extraneous cognitive load. It does not eliminate the cognitive demands of the task itself.

Not every struggling reader brings the same language system to the task

This is where my clinical experience makes me much more cautious about broad claims for speech-to-print approaches.

I have spoken with parents, therapists, and tutors whose children or students received this kind of instruction and did not make the expected progress. I have also worked with learners whose language profiles make it clear that changing the direction of phonics, from print-to-speech to speech-to-print, does not address the primary barrier they are experiencing.

These children do not simply have difficulty matching sounds and spellings.

Some have weak vocabularies.

Some have significant receptive and expressive language weaknesses.

Some have poorly established lexical representations.

Some have substantial working-memory limitations.

Some struggle with lexical retrieval.

Some have difficulty with inhibition, cognitive flexibility, attention, sequencing, or other executive-function demands.

Some process language slowly.

Some have highly uneven language-processing profiles, with substantial strengths in one area and significant weaknesses in another.

For those learners, simplifying phonics may remove obstacles.

But it does not necessarily supply what is missing.

That distinction becomes especially important when speech-to-print is presented as the next answer for children who have already failed to make progress with other forms of phonics instruction.

The problem may not be that we have been traveling in the wrong direction between speech and print.

The problem may be that we have made the road between speech and print too much of the journey.

A list of possible spellings is still a list

This brings me back to Lyn Stone’s observation:

“Listing possibilities is not an account of the writing system.”

Exactly.

Knowing that /iː/ may be represented by <ee>, <ea>, <e>, or <y> tells a learner something true about English.

It tells them what English might use to represent that phoneme.

It does not tell them why a particular word contains a particular spelling.

Why is heal spelled with <ea>?

Why is heel spelled with <ee>?

Why does he end with <e>?

Why do heal and health retain the same <ea> even though the pronunciation changes?

Phonology alone cannot answer those questions.

Their spellings preserve information about those relationships even when their pronunciations differ.

Pronunciation matters enormously.

But pronunciation is one part of the system.

It is not the whole system.

And when phonology leaves several spellings possible, the learner either has to guess and check, or we can give them more linguistic evidence.

What if the word, not the phoneme, is the object of study?

This is where I think the current discussion gets trapped in an unnecessarily narrow choice.

If we question speech-to-print, the presumed alternative is often traditional phonics or Orton-Gillingham instruction.

Teach the grapheme.

Teach its sound.

Teach a syllable type.

Teach a division rule.

Practice a controlled word list.

Read a controlled text containing only the patterns that have already been taught.

But those are not the only two possibilities.

We do not have to choose between elaborate print-to-speech phonics and streamlined speech-to-print phonics.

We can ask a different question entirely:

What if the word, not the phoneme, is the object of study?

That changes the architecture of the instruction.

Start with the word

In Applied SWI, I am not trying to eliminate phonology.

I am changing where phonology belongs in the investigation.

We begin with a word.

What does the word mean?

How is it built?

What other words are related to it?

What can those relationships tell us about its spelling?

Then we investigate what the graphemes are doing within that morphological structure.

That sequence is key.

Take heal.

We can begin with the word itself.

What does heal mean?

What words are related to it?

heal

heals

healed

healing

healer

health

healthy

Now <ea> is no longer simply one item from a list of possible graphemes for /iː/.

It is part of an orthographic relationship across a family of words.

And something particularly useful happens when we reach health.

The pronunciation changes.

The spelling relationship remains visible.

That is not an unfortunate irregularity that needs to be explained away.

That is useful information.

The spelling helps us see a meaningful relationship that speech alone does not preserve transparently.

If we begin and end with phoneme-grapheme mapping, that relationship sits outside the central lesson.

If we begin with the word and investigate its meaning, structure, and relatives, the relationship is available from the beginning.

We have given the learner evidence that constrains the spelling.

Phonology has not disappeared

This distinction is often misunderstood.

When I say phonology should not organize the investigation, people sometimes hear that I am saying phonology does not matter.

Of course it matters.

We still investigate phonemes.

We still investigate graphemes.

We still investigate how graphemes represent phonemes in words.

But we investigate those relationships inside morphological structure, rather than assuming that phoneme-grapheme correspondence can explain the spelling system by itself.

The central question changes.

Instead of asking only:

What letters can spell this sound?

we can ask:

Why is this word spelled this way?

Sometimes phonology provides much of the answer.

Sometimes morphology provides critical evidence.

Sometimes the history of the word matters.

Often several sources of information interact.

That is not a complication we need to hide from children.

That is the writing system.

Reading a word aloud is not the same thing as understanding it

There is another limitation in the way we talk about phonics interventions that I think deserves considerably more attention.

A child can become much better at reading words aloud without making a comparable gain in comprehension.

Those are not interchangeable outcomes.

If a learner already has strong oral language, a broad vocabulary, good background knowledge, and relatively intact language comprehension, improved access to written words may unlock a great deal.

Once the child identifies the written word, there is already a rich linguistic representation waiting for it.

That may be why some children appear to make remarkable progress once a more efficient approach finally gives them access to print. Their oral language was already strong. Their vocabulary was already there. Their knowledge of the world was already there. Their comprehension system was already functioning.

They needed better access to the written words.

But that is not every struggling reader.

For a child with weak vocabulary, successfully producing the spoken form of a written word does not create the meaning of that word.

For a learner with significant receptive-language difficulties, accurate oral reading does not guarantee understanding.

For a learner who struggles with complex syntax, reading every word in a sentence correctly does not resolve the syntactic problem.

For a learner who has difficulty making inferences, integrating information across sentences, understanding nonliteral language, maintaining referents, or constructing a coherent mental representation of a text, more efficient phoneme-grapheme mapping does not automatically solve those problems either.

This is why I become concerned when improvements in oral word reading are treated as evidence that we have solved “reading.”

We have addressed one part of reading.

Sometimes it is a very important part.

But sometimes the child who can now read the passage aloud still cannot tell us what it means.

That child needs more than a better route between speech and print.

Language cannot remain something we get to later

If vocabulary, morphology, syntax, meaning, and comprehension are central to literacy, they cannot remain the things we address after children have mastered the code.

They cannot be the enrichment layer.

They cannot be the pieces we add after phonics has done its work.

For some children, those linguistic systems are precisely what make the orthography learnable.

Word families give memory support.

Morphological structure gives letters a reason to be where they are.

Meaning narrows possibilities.

Known words support new words.

Sentence context strengthens lexical representations.

Repeated encounters with related words create networks rather than isolated mappings.

For a learner with limited working memory or fragile phonological processing, a meaningful word family may provide more support than another round of phoneme segmentation and grapheme selection.

That is not avoiding phonology.

It is giving phonology something to attach to.

Children can handle a richer system

One of the more encouraging developments in newer approaches to literacy instruction is the recognition that learning does not have to proceed through perfectly controlled little steps.

Children can encounter variability.

Learning can be cumulative.

Previously encountered information can return in new contexts.

Practice can be interleaved.

Reading and spelling can be integrated.

Learners do not necessarily need to “master” one pattern before being allowed to encounter another.

Good.

Now keep going.

If children can handle variation among phoneme-grapheme correspondences, why would we protect them from morphological complexity?

If learning can be interleaved, why would we interleave only sound-spelling relationships?

If children benefit from real words, why would we strip those words of the very relationships that help explain their spelling?

If learning is recursive, why would morphology, etymology, syntax, vocabulary, and meaning have to wait their turn in a linear scope and sequence?

Children already possess language.

They know words.

They know meanings.

They recognize relationships.

They notice patterns.

They make predictions.

They revise their hypotheses when they encounter new evidence.

Those are not obstacles to literacy instruction.

Those are resources.

We can build from them.

Changing direction is not changing the model

Speech-to-print may represent a substantial improvement over traditional phonics and Orton-Gillingham instruction because it can remove a great deal of unnecessary instructional complexity.

That is crucial.

It may reduce extraneous cognitive load.

It may integrate reading and spelling more naturally.

It may permit greater variability.

It may encourage cumulative and interleaved learning.

For some learners, those changes may make a substantial difference.

But reducing unnecessary instructional load is not the same thing as meeting the language, memory, executive-function, and comprehension needs of every learner.

And if the instructional architecture still begins with phonemes and organizes spelling primarily around the question of which graphemes can represent them, then the underlying model remains phonological.

The arrow has changed direction.

The organizing assumption has not.

Speech-to-print changes the direction of travel. Structured Word Inquiry changes the object of study.

One asks us to become more efficient at moving between phonemes and graphemes.

The other asks us to investigate words.

And words give us meaning, structure, relatives, history, pronunciation, syntax, and somewhere much larger to go.

Maybe phonics is too small for the job

English orthography represents pronunciation.

It also represents meaning, structure, relationships, and history.

If that is our description of the writing system, then our instruction should give learners access to that same system.

Not phonology first and everything else later.

Not morphology after the “basic code.”

Not etymology as an interesting story we tell when a spelling seems strange.

Not vocabulary and comprehension after a child can read the words aloud.

Not a longer list of possible graphemes followed by guess and check.

All of it belongs to the word.

Changing print-to-speech into speech-to-print may improve phonics instruction considerably.

But it remains a change within phonics.

So perhaps the most useful question raised by the current discussion is not whether phonics should travel from print to speech or from speech to print.

Perhaps the question is whether phonics is large enough to serve as the map.

Structured Word Inquiry Is Not a Brand

Recently, Structured Word Inquiry has been discussed in terms of brands, camps, gangs, and even cult-like allegiance. I think that characterization deserves a response because it gets the direction of travel almost exactly backward.

Structured Word Inquiry is not another entry in the marketplace of literacy programs, sitting alongside Orton-Gillingham, Wilson, speech-to-print, or linguistic phonics and asking practitioners to choose a team and defend it. If anything, it is a call to leave that kind of program allegiance behind and study the writing system itself.

And when the language of “gangs” is invoked in a discussion of SWI, I object to it. Not because SWI or the people who use it should be protected from criticism. Quite the opposite. Challenge any claim an SWI practitioner makes. Ask for the evidence. Test the analysis. Find a counterexample. Revise the hypothesis if it does not survive examination. That is exactly what inquiry requires.

But “gang” is not an argument.

Nor does declining to identify with a named method somehow place anyone outside professional communities, intellectual influences, or schools of thought. We all learn from other people. Our practices develop because we encounter ideas, investigate them, reject some, retain others, and change what we do.

No individual, program, or professional community has a monopoly on linguistic truth. English orthography does not belong to SWI, phonics, speech-to-print, or any particular author, trainer, or practitioner. Claims about the writing system stand or fall on the evidence.

That is precisely the point.

Structured Word Inquiry is not a commercial reading program. It is not a proprietary curriculum with a prescribed sequence, required materials, scripted routines, or lessons practitioners are expected to implement with fidelity. There are no universal SWI worksheets or official SWI lesson plans. There is no sequence dictating which feature of the writing system a learner may encounter this week and which one must wait until Lesson 46.

Structured Word Inquiry is a method of using four questions to investigate English orthography:

1. What does the word mean?

2. How is the word built?

3. What other words are related to it in structure and meaning?

4. What are the letters and graphemes doing?

Those questions provide structure for an investigation. They do not prescribe the investigation itself.

And the thing we are studying is not SWI.

We are studying English orthography.

We needed something to call the work

Whenever people begin working in a particular way, it helps to have language for referring to what they are doing. Structured word inquiry gives us a convenient way to name the practice of using structured questions and linguistic evidence to investigate written words.

Having a name does not make something a brand.

I sometimes hear people say they “believe in the principles of SWI.” I understand what they mean, but those principles are more accurately described as principles of English orthography.

SWI did not invent morphemes, morphological families, phonology, etymology, graphemes, or the historical processes that shaped English spelling. It did not create the fact that written words preserve relationships of structure and meaning even when pronunciation varies across members of a family.

Those are properties of the writing system.

If nobody ever used the term Structured Word Inquiry again, <sign> would remain related to <signal> and <signature>. <Heal> would remain related to <health>. Morphological structure would continue to constrain grapheme analysis. Pronunciations would continue to change while spellings preserved relationships that are not always recoverable from the pronunciation of an isolated word.

The facts belong to English.

SWI gives us a disciplined way to investigate them.

English orthography is the scope

This is one of the most important distinctions between Structured Word Inquiry and a literacy program.

Programs necessarily define their scope. Someone decides what will be taught, in what order, through which activities, with which materials, and according to what criteria for progression.

SWI does not impose a scope on English orthography.

English orthography is the scope.

If something is relevant to understanding how an English word is written, it is available for investigation: meaning, morphology, etymology, phonology, graphemes, stress, historical changes in pronunciation, orthographic conventions, lexical relationships, and the interactions among all of them.

This is why reducing SWI to “morphology instruction” misses the point. Morphology is something we investigate. So are etymology and phonology. Matrices and word sums are tools we may use to make structural relationships visible.

None of those things, by itself, is SWI.

The method is the inquiry. The scope is the writing system.

The same is true of the four questions. They are not four boxes to be checked in sequence. An investigation is recursive. Evidence discovered through one question often changes what we think about another. A proposed relative may make us reconsider structure. Structure constrains grapheme analysis. Etymological evidence may reveal that words that look similar do not belong to the same family, or that a relationship obscured in present-day English becomes clear once we investigate the history.

Sometimes our first hypothesis is wrong.

Good.

Revising a hypothesis in response to evidence is not a failure of inquiry. It is inquiry.

One of the most useful things I can say to a client is, “I don’t know. How could we find out?”

I would much rather investigate an uncertainty than cover it with a rule that collapses the moment we look beyond the word list on which it was taught.

The science leaves room for art

The absence of a universal SWI script is not a weakness. It is one of the features I value most.

People using the four questions to study English orthography bring different kinds of professional knowledge to that work. A classroom teacher, tutor, linguist, reading specialist, and speech-language pathologist will not necessarily conduct the same investigation in the same way. Nor should they.

The linguistic evidence constrains what we can claim about a word. It does not dictate every instructional decision we make while helping another person understand it.

That is where the art enters the science.

I call my own application Applied SWI because the way I use Structured Word Inquiry is inseparable from the fact that I am a speech-language pathologist and clinician. I diagnose and treat language and literacy disorders, language-based learning disabilities, and speech impairments. That clinical background shapes what I notice, what I prioritize, and where I choose to enter an investigation.

When I work with a client, I am paying attention to much more than whether a word was read or spelled correctly. I am listening to speech production and thinking about phonological representations. I am considering vocabulary, semantics, syntax, language processing, working memory, comprehension, and the relationships among spoken and written language.

I am noticing what remains stable and what falls apart under cognitive load. I am paying attention to what causes a client to disengage and what allows that same client to persist.

Those observations influence what I do next.

That is my application of SWI. It is not a version everyone else is expected to reproduce.

The freedom is key

One of the things I love most about studying English orthography through the four questions is the flexibility. The freedom.

I can begin anywhere.

An investigation might begin with something a client wrote, a word in a book, a misreading, an unfamiliar word, an interesting spelling, a speech production, or a question that appears spontaneously in the middle of something else. It might begin with meaning, structure, a morphological relative, or a grapheme whose presence has suddenly become interesting.

Where I begin depends on the client.

It depends on what that person already knows, what interests them, how their language system is functioning, what places too much demand on memory, what has happened during previous literacy instruction, what causes them to shut down, and what keeps them thinking.

I make those decisions for every client I work with.

And when an investigation unexpectedly opens a productive path, I can follow it. I do not have to abandon a useful question because somebody else’s scope and sequence has decided the concept belongs several months later.

That freedom is not randomness. Responsive teaching is not random teaching.

The structure comes from the four questions. The constraints come from English orthography and the linguistic evidence we use to investigate it. The clinical and instructional decisions are mine.

SWI gives me a principled framework without requiring me to surrender professional judgment.

Ten years ago, SWI introduced me to English orthography

For me, none of this is merely theoretical.

Ten years ago, SWI introduced me to something that had been almost entirely absent from my professional literacy training: English orthography itself.

I had been trained to teach reading and spelling. I had used programs with brand names. I had manuals, materials, rules, procedures, prescribed sequences, and explanations for what to do when a learner did not acquire something the first time.

What I had not been taught to do was study the English writing system.

SWI changed that.

I began looking at written words as meaningful linguistic structures that could be investigated rather than primarily as strings of letters whose pronunciations had to be decoded or strings of phonemes that had to be encoded.

I could ask why a word was spelled the way it was and actually investigate the question.

Meaning, morphology, etymology, phonology, historical relationships, and orthographic conventions stopped being separate pieces of information that might occasionally be added to a literacy lesson. They became interacting sources of evidence for understanding a single writing system.

And thank goodness SWI introduced me to that system, because the programs with brand names I had been using were not working for many of the children who came to me.

I did not need another brand.

I needed to understand the thing I was trying to teach.

And when practitioners understand orthography, they may no longer need someone selling them a program at all.

That shift fundamentally changed my practice.

SWI is not simply another position in the phonics debate

This distinction becomes particularly important when SWI gets pulled into arguments about traditional phonics, speech-to-print, and so-called linguistic phonics.

Much of that debate concerns the direction of phoneme-grapheme instruction. Should instruction move from letters to sounds or from sounds to possible spellings? Which is easier? Which reduces instructional complexity? Which is more efficient?

Those may be useful questions, but they leave a more fundamental assumption largely untouched: that phoneme-grapheme mapping should organize our understanding of English orthography in the first place.

Structured Word Inquiry allows us to investigate that assumption rather than simply inherit it.

Knowing that a phoneme can be represented by several graphemes tells us something important about English. It tells us what is possible. It does not necessarily explain why a particular word has the spelling it has.

To answer that question, we may need to know how the word is structured, what it means, which words belong to its family, where it came from, how its pronunciation has changed, and which orthographic conventions are operating within it.

Phonology is indispensable to that investigation.

It is simply not synonymous with the writing system, nor must every investigation begin there.

That is a very different proposition from replacing print-to-speech phonics with speech-to-print phonics.

What SWI actually asks of us

SWI is not a commercial curriculum, a collection of morphology lessons, an etymology program, a matrix, a word sum, a set of worksheets, a scripted routine, or a prescribed developmental sequence.

Nor does it demand that every practitioner teach in the same way.

People certainly form professional communities around SWI. They attend courses, read particular authors, use shared terminology, develop resources, and learn from practitioners whose work they respect. The same is true of virtually every substantial body of professional work. The existence of a community does not transform the underlying method into a gang, a cult, or a proprietary brand.

What SWI asks is much less comfortable than allegiance to a program.

It asks us to investigate.

It asks us to understand the writing system well enough to evaluate our own explanations. It asks us to support claims with evidence and to give them up when the evidence does not support them. It asks practitioners to think rather than simply implement.

Structured Word Inquiry is the name we give that method of inquiry.

English orthography is what we study.

The name is convenient. The science does not belong to the name. And no one gets to claim ownership of the linguistic facts simply by declining to use the label.

English orthography provides the scope. The four questions provide the method of inquiry. The practitioner brings the art of application.

That is why I do not think of myself as someone who abandoned one literacy brand and adopted another.

Ten years ago, I stopped relying on programs to tell me how English worked and began studying English itself.

If anything, Structured Word Inquiry was my way out of the gangs.

I was rereading a 2003 paper by Amanda Walley, Jamie Metsala, and Victoria Garlock called Spoken Vocabulary Growth: Its Role in the Development of Phoneme Awareness and Early Reading Ability, and something clicked for me in a way it never had before.

The paper describes the Lexical Restructuring Model. The basic idea is that as children’s spoken vocabularies grow, the words in their lexicons have to become increasingly well differentiated from one another. As more words are learned, there are more phonologically similar neighbors competing with one another.

At first I was thinking about this in the usual way. A child learns cat. Then cap. Then cab. Then can. The lexicon keeps growing. Similar words have to become more precisely differentiated, and phonological awareness develops along with that growth.

And then I thought:

Wait a minute.

The kids I work with often don’t learn those words.

That is exactly the problem.

But my most severely struggling learners weren’t learning more words.

They were stuck.

They were still working on sit, set, pin, pen, and other supposedly simple CVC words years after those words had first been introduced. They had segmented them, blended them, sorted them, mapped them, practiced the short vowels, and read them in lists.

And the words still weren’t stable.

So their lexicons weren’t expanding in the way this model describes.

And if lexical growth is one of the things that helps phonemic awareness develop, then while these children were stuck practicing the same CVC words over and over, their phonemic awareness wasn’t getting the benefit of that lexical growth either.

They were waiting for phonemic awareness to improve so they could move on to more words.

But learning more words may have been part of what they needed in order for phonemic awareness to improve.

They were stuck in the CVC trap.

What if the “easy” words aren’t easy?

We have defined easy beginning words largely according to the needs of a phonics scope and sequence.

Three letters.

CVC.

One short vowel in the middle.

Therefore, simple.

But three letters tell us nothing about how easy a word is for a particular child to perceive, distinguish, represent, retrieve, read, or spell.

This is especially interesting when we look at vowels.

Some of my most severely struggling learners have enormous difficulty with /ɪ/ and /ɛ/. Sit and set. Pin and pen. Those vowels are close enough perceptually that differentiating them can remain difficult even after years of instruction.

That isn’t just something I happen to see clinically. Research on children with language difficulties has found problems processing phonetically similar vowel contrasts, and the broader literature on phonological awareness also shows that the phonological characteristics of the words themselves affect children’s performance.

So imagine what we may be doing when we decide that the proper starting point for a struggling learner is a tightly controlled collection of CVC words distinguished by exactly the vowel contrasts that learner finds hardest to perceive.

Then we call those words simple because they are short.

I don’t think those are the same thing anymore.

Let me tell you about play

One of my most severely struggling learners has a complex neurodevelopmental profile involving two syndromes that affect language and literacy development, along with language disorder and autism. I’ll call him Leo here.

I have worked with him for four years.

Leo’s first anchor word was play.

Anchor word is my own term. I use it for a word the learner can really get hold of, a known word that gives us somewhere to begin and somewhere to return as we investigate the writing system.

I would love to tell you that I chose play because ten and a half years ago I had already worked out everything I am writing about here.

I didn’t.

I chose play because it is practically the quintessential Structured Word Inquiry word.

Children know play. Children play.

And play has wonderful polysemy. We can play outside, play a game, play music, play a role, play a video, or watch a play. There are all kinds of ways to use it.

Morphologically, it gives us plenty to work with too:

<play + s → plays>
<play + ed → played>
<play + ing → playing>
<play + er → player>
<play + ful → playful>

Then there are compounds such as playground and playmate.

It gives us meaning, structure, relatives, suffixes, compounds, grammatical changes, sentences, and repeated encounters with the same base.

That is why I chose play.

And Leo learned it.

He also learned the prefixes and suffixes quickly and easily. In fact, he learned them so well that he started trying them out with other words we studied.

“Can we add <-ful>?”

“Can we add <-ness>?”

He especially liked it when we could add one of those less frequent derivational suffixes instead of the inflectional suffixes he saw all the time. There was a little excitement in discovering that yes, this one works here too.

That was important.

He was not just memorizing play, played, playing, and player as separate words. He was learning something about how words are built, and then he was carrying that knowledge into new investigations.

Play became a word he could depend on, but it also became a structure he could build from.

But now, after reading the Walley, Metsala, and Garlock paper again, I see something about play that I did not see when I chose it.

It may also be a remarkably good phonological anchor for a child like Leo.

Play is phonologically simple too

The word play has three phonemes: /p/, /l/, /eɪ/.

Its spelling gives us three graphemes: <p>, <l>, and <ay>.

For Leo, /eɪ/ is a very clear vowel. It is tense, sustained, and easy for him to perceive and feel. It comes at the end of the word, with nothing following it.

And once he had play, he had other places to go.

He reads say.

He reads way.

He reads stay.

He reads away.

He reads today.

Now here is the part that really got me.

After four years with me, Leo reads many words. He reads plenty of words that are longer and morphologically more complex than the CVC words he was originally supposed to master first. We have investigated those words in families. He has encountered them repeatedly in reading. They have become part of his lexicon.

But just the other day, he couldn’t read mud.

He couldn’t read sit.

He couldn’t read set.

He can read away, but not reliably read sit.

How is sit the simpler word?

It is shorter.

That is not the same thing.

Leo doesn’t blend

There is another important part of this.

Leo does not blend.

He either knows the word or he doesn’t.

When he knows the word, he reads it.

When he gets to a word that he does not know, sometimes his entire speech system seems to start searching. The best way I can describe it is like a car backfiring. His articulators start moving in different directions, producing attempted starts, but the movements are not necessarily corresponding in any orderly way to the first grapheme, then the second grapheme, then the third.

There is no neat:

/m/ … /ʌ/ … /d/ … mud.

That process doesn’t get him to the word.

Even after all of our work, he can still have difficulty with <sh> and <ch> in initial position. He may want to treat the first letter by itself rather than recognizing the two-letter grapheme.

So when someone tells me this child has a blending problem, I have to ask what that actually explains.

If he hasn’t correctly identified the grapheme in the first place, what exactly are we asking him to blend?

If the phonological representation of the word itself isn’t stable, why would producing a series of isolated approximations of phonemes necessarily lead him to the whole word?

For Leo, it doesn’t.

What has worked has been almost the reverse.

We establish a meaningful word.

We investigate its structure.

We build its family.

We encounter it repeatedly.

Its orthographic representation becomes stable.

And inside that increasingly stable word representation, we investigate what the graphemes are doing.

His grapheme-phoneme knowledge has grown enormously this way.

And <pl> is not a “blend”

This also made me think differently about the beginning of play.

In reading instruction, <pl> is usually called a consonant “blend.”

I don’t call it that.

In linguistics, blend already has a meaning. A lexical blend or portmanteau combines material from words, as in motel from motor and hotel.

In play, we have a consonant cluster.

<p> and <l> are two graphemes representing two phonemes.

And for a struggling learner, I don’t want to blur the distinction between them.

I want to accentuate it.

There is a reason for that. Children are known to omit the second consonant in initial consonant clusters. Rebecca Treiman documented children spelling play as pay. Bruck and Treiman likewise found that both young typical readers and dyslexic readers had difficulty analyzing consonants in word-initial clusters.

Think about pay for play.

That child does not need <p> and <l> to become more blended.

The child needs to notice the <l>.

I want more differentiation, not less.

<p> is here. <l> is here. They are two separate graphemes. They represent two separate phonemes. We can see them separately, hear them separately, and pay attention to what the articulators are doing for each.

Then, of course, when we say play naturally, we do not speak /p/ and /l/ as two little detached noises and glue them back together.

We just say play.

That brings me to Mark Seidenberg.

Speech is not beads on a string

Seidenberg has been making a point that should have major implications for how we talk about phonemic awareness.

Phonemes are abstractions.

Spoken words do not literally contain a row of neatly separated sound particles waiting to be pulled apart and pushed back together.

His description is memorable: speech isn’t “beads on a string.”

Yet a tremendous amount of beginning reading instruction is built around a model that behaves as though it is.

Produce this isolated sound.

Now this one.

Now this one.

Blend them together.

But the isolated sounds produced during that procedure are not little chunks that were originally sitting separately inside the spoken word.

In his 2026 piece On Blending, Seidenberg makes the point even more explicitly. He calls blending an artificial procedure that arises because children are first taught to decode serially, grapheme by grapheme, generating isolated phoneme-like sounds that then have to be converted somehow into the naturally pronounced word.

He uses the example of consonant clusters too. A child who can talk already knows how to produce sequences such as /st/ in stop, step, and stand. The child does not need to learn how to combine two independently produced sounds in order to speak that cluster.

This makes me think immediately about Leo.

When he knows play, he says play.

When he doesn’t know mud, producing isolated pieces of mud does not reliably get him there.

Maybe the answer isn’t more blending practice.

Maybe the word isn’t stable yet.

What makes an anchor word work?

I have used other words such as snow, plow, green, and grass as anchor words with struggling learners.

Again, I originally chose words like these because they were meaningful and gave us useful things to investigate.

Now I am looking at their phonology too.

Play gives us /eɪ/.

Snow gives us /oʊ/.

Plow gives us /aʊ/.

Green gives us <ee> representing /i/.

And grass gives us /æ/, which in my clinical experience is often more perceptually accessible for severely struggling learners than /ɪ/, /ɛ/, or /ʌ/.

I am not about to replace one rigid scope and sequence with another and announce that all children must now start with play, snow, plow, green, and grass.

That would defeat the whole point.

The question is not:

What is the easiest word according to the program?

The question is:

What can this learner perceive, know, and stabilize?

A useful anchor word may give us meaning the learner already understands, a pronunciation that is perceptually accessible, morphological structure that gives us somewhere to go, and a spelling we can investigate repeatedly.

That is a much richer definition of simple.

We don’t even know the graphemes until we know the morphemes

There is another reason I don’t accept the sequence in which we are supposed to teach grapheme-phoneme correspondences first and morphology later.

We don’t necessarily know what the graphemes are until we know the morphemes.

Graphemes do not cross morpheme boundaries.

Consider <been>.

The word sum is:

<be + en → been>

The first <e> belongs to the base <be>. The second <e> belongs to the suffix <en>.

There is no <ee> digraph in <been>.

Or look at <mishap>:

<mis + hap → mishap>

The <s> and <h> happen to stand beside one another in the finished word, but they do not form an <sh> grapheme. The <s> belongs to <mis> and the <h> belongs to <hap>.

So morphology is not some advanced layer we sprinkle on after children have mastered phonics.

Morphological analysis helps us determine what the graphemes are in the first place.

For the past ten and a half years, this is how I have worked clinically.

I am a speech-language pathologist who identifies, diagnoses, and treats language disorders and language- and literacy-based learning disorders. Many of the learners who reach me have already spent years in phoneme-first programs and made little or no meaningful progress.

I begin with morphology from the beginning.

I don’t wait for phonemic awareness mastery.

I don’t wait for short vowels.

I don’t require a child to pass through CVC words before giving them access to richer words.

We begin with a word the learner knows.

What does it mean?

How is it built?

What other words are related to it?

What is each letter doing in this word?

And over and over again, I have watched grapheme-phoneme knowledge improve dramatically within that work.

There is research that makes that clinical experience even more interesting.

Teach morphology and phonological awareness improves

In 2010, Amanda Goodwin and Soyeon Ahn published a meta-analysis of morphological interventions for children with literacy difficulties. They synthesized 79 effects from 17 independent studies.

Morphological instruction was associated with improvement in morphological awareness, vocabulary, reading comprehension, spelling—and phonological awareness.

The effect for phonological awareness was d = 0.49.

Morphological instruction.

Phonological-awareness gains.

Goodwin and Ahn’s findings were particularly strong for children with speech and language difficulties.

Then Julie Wolter and Laura Green published a 2013 case study involving morphological-awareness intervention with an eight-year-old child with language and literacy difficulties. The intervention was designed around morphology, but improvement showed up in phonological awareness too.

The authors specifically noted that the phonemic-awareness improvement was interesting because phonemic awareness had not been directly targeted.

That sounds very familiar to me.

When I teach morphology, phonology has never gone anywhere.

We say words.

We compare them.

We notice contrasts.

We examine different pronunciations of graphemes and morphemes.

We pay attention to what changes and what stays the same.

We look at spelling and pronunciation together.

The learner is doing phonological work all the time.

It just has somewhere to live.

Now go back to Walley, Metsala, and Garlock

This is where all of these pieces came together for me.

Walley, Metsala, and Garlock propose that vocabulary growth helps create increasingly fine-grained lexical representations and that this restructuring contributes to phoneme awareness.

Goodwin and Ahn find that morphological intervention improves phonological awareness.

Wolter and Green show phonological improvement occurring during morphological intervention even though phonemic awareness was not the direct target.

Treiman shows us that children can lose the second consonant in a cluster, even writing pay for play.

Seidenberg reminds us that phonemes are abstractions, that speech is not beads on a string, and that the instructional procedure we call blending is artificial rather than something inherent in spoken language.

And then I have Leo.

He reads play.

He reads stay.

He reads away.

He reads today.

He reads many longer, morphologically complex words that we have studied and encountered over and over.

But he can still get to sit and have no idea what to do with it.

He can still fail to retrieve set.

He can still fail to read mud.

Those words did not become foundational just because somebody put them at the foundation of a scope and sequence.

What if keeping them on CVC words is part of what keeps them stuck?

This is the question I cannot stop thinking about now.

What happens when a learner struggles with the short-vowel words at the beginning of the sequence and we respond by refusing to move beyond them?

More short-vowel practice.

More CVC lists.

More segmenting.

More blending.

More mastery checks.

And shudder retaining children who don’t perform up to the “standards.”

The learner stays in the same tiny lexical neighborhood, working on the same poorly differentiated representations, because we have decided those representations must become stable before we allow the learner access to anything more complex.

But according to the Lexical Restructuring Model, lexical growth itself helps drive phonological differentiation.

What if the learner needs more words?

Not more random words and not harder words for the sake of being harder.

Rich words.

Known words.

Interesting words.

Words with pronunciations the learner can perceive clearly.

Words with morphological families.

Words that connect to other words.

Words we can use in sentences and encounter again in reading.

Words that give the developing language system more information with which to organize itself.

Maybe some children remain stuck on CVC words because we keep waiting for their phonological system to develop before allowing their lexicon to expand, when expanding and organizing the lexicon may be one of the things that helps the phonological system develop.

That is a very different way of looking at intervention.

Simple for whom?

The longer I do this work, the less interested I am in whether a word looks simple to an non-dyslexic adult.

Three letters do not make a word simple.

One syllable does not make a word simple.

A so-called regular spelling does not make a word simple.

A word is useful when the learner can get hold of it.

Play turned out to be that word for Leo.

It had meaning he already knew.

It had rich polysemy.

It had a large morphological family.

It had three phonemes he could perceive.

It had a clear final vowel.

It gave us two adjacent consonants that we could explicitly differentiate rather than blur together as a “blend.”

It gave us a stable spelling to return to.

And most importantly, it gave us somewhere to go.

I chose play because it was morphologically rich.

I chose it because every child understands what it means to play.

I had no idea when I chose it that I might also have stumbled onto a word that was phonologically more accessible for this learner than the supposedly simple CVC words he had been expected to master first.

Now I wonder how often that happens.

Maybe we have the sequence backward

For decades, the dominant instructional response to struggling readers has been to strip language down.

Take away the sentence.

Take away the meaning.

Take away the morphology.

Take away the whole word.

Get down to the phoneme. And, God forbid, nonsense syllables.

Teach those pieces first and build upward.

But for some of the children I work with, that route has failed spectacularly.

What if we gave them more language instead?

Give them the word.

Give them its meaning.

Give them its structure.

Give them its relatives.

Use it in sentences.

Encounter it in text.

Give them a spelling they can investigate and return to.

Then help them notice the phonology inside an increasingly stable representation of that word.

We do not have to remove meaning to teach sound.

We do not have to remove morphology to teach graphemes.

We do not have to master isolated phonemes before children are allowed to learn interesting words.

And we certainly don’t have to keep a fifth-, sixth-, or seventh-grade learner trapped in sit, set, pin, and pen because those words appear earlier in somebody’s scope and sequence.

Maybe phonemic awareness does not always have to come first.

Maybe for some of our most severely struggling learners, knowledge of words is one of the ways we get there.

And maybe our first question should not be:

Which phoneme does this child need to master next?

Maybe it should be:

Which word can this child get hold of?

Start there.

References

Bruck, M., & Treiman, R. (1990). Phonological awareness and spelling in normal children and dyslexics: The case of initial consonant clusters. Journal of Experimental Child Psychology, 50(1), 156–178.

De Cara, B., & Goswami, U. (2003). Phonological neighbourhood density: Effects in a rhyme awareness task in five-year-old children. Journal of Child Language, 30(3), 695–710.

Goodwin, A. P., & Ahn, S. (2010). A meta-analysis of morphological interventions: Effects on literacy achievement of children with literacy difficulties. Annals of Dyslexia, 60(2), 183–208.

Seidenberg, M. S. (2024). More on teaching phonemes. Reading Matters.

Seidenberg, M. S. (2026). On blending.

Treiman, R. (1991). Children’s spelling errors on syllable-initial consonant clusters. Journal of Educational Psychology, 83(3), 346–360.

Walley, A. C., Metsala, J. L., & Garlock, V. M. (2003). Spoken vocabulary growth: Its role in the development of phoneme awareness and early reading ability. Reading and Writing, 16, 5–20.

Wolter, J. A., & Green, L. (2013). Morphological awareness intervention in school-age children with language and literacy deficits: A case study. Topics in Language Disorders, 33(1), 27–41.

Structured Literacy Interventions: Teaching Students with Reading Difficulties, Grades K–6, edited by Louise Spear-Swerling, is an interesting book to read at this particular point in the science of reading conversation.

Published in 2022, it presents Structured Literacy as much more than phonemic awareness and phonics. Across ten chapters, the contributors address spelling, morphology, multisyllabic word reading, fluency, vocabulary, oral language comprehension, reading comprehension, written expression, and multicomponent intervention. The book also takes seriously the fact that struggling readers have different profiles and therefore need different kinds of intervention.

There is a lot here that I agree with.

The book repeatedly moves toward a more integrated understanding of literacy.

But it never fully reorganizes instruction around that understanding.

Even as Structured Literacy expands to include morphology, semantics, syntax, comprehension, and written expression, its word-level foundation remains the same: beginning instruction is organized around synthetic phonics at the grapheme-phoneme level.

That creates the tension that runs through this book.

Its strongest chapters describe written language as an interconnected system of meaning, structure, pronunciation, vocabulary, grammar, and use. Its foundational instructional model still asks learners to enter that system through phonology first.

By the end of the book, I was left with a different question than the one I started with:

What happens when Structured Literacy itself has grown beyond the phonics-first framework it was built upon?

Structured Literacy is now an enormous umbrella

The book makes clear that Structured Literacy is not one commercial program and is not intended only for students with dyslexia. It describes Structured Literacy as an umbrella covering phonemic awareness, phonics, orthography, morphology, syntax, semantics, reading comprehension, spelling, written expression, and other areas of language and literacy.

Under this definition, a Structured Literacy intervention might include direct vocabulary instruction, work with complex sentences, narrative language, inferencing, paragraph organization, morphological families, fluency, spelling, or phonemic awareness. What is taught depends partly upon the learner’s profile.

That flexibility is one of the book’s strengths. Children who struggle with reading do not all have the same needs. A fluent word reader with poor language comprehension does not need the same intervention as a learner with severe word-recognition difficulties. An older student who can read one-syllable words does not necessarily need to return to basic phonemic-awareness drills.

The book repeatedly emphasizes assessment and individualized intervention. It distinguishes among specific word-recognition difficulties, specific reading-comprehension difficulties, and mixed reading difficulties. It also acknowledges that an older learner may need work with spelling, automaticity, or fluency without needing basic phonics or phonemic awareness.

All of that is reasonable. But when the book summarizes the defining features of Structured Literacy, one instructional commitment remains fixed: initial phonics and spelling instruction use a synthetic phonics approach beginning at the grapheme-phoneme level.

So the framework has become broad enough to include the whole language system, but its starting point has not changed. Meaning, morphology, syntax, and history are included, but they still do not organize the beginning of written-word instruction. Phonology does.

Adding more components does not change the architecture

A program can include morphology, vocabulary, syntax, and semantics while still leaving phonics in charge of the instructional sequence. Those components may appear in separate lessons or be added after basic decoding has been established, while written words continue to be treated primarily as sequences of sound-symbol correspondences.

Their inclusion alone does not change how the system is being taught. The more useful question is not whether morphology, meaning, syntax, and vocabulary appear somewhere in the program. The question is how they are related.

Are they taught as separate components that are assembled after basic decoding has been established? Or does instruction help the learner understand from the beginning that written words coordinate meaning, structure, and pronunciation?

The book often describes language as an integrated system. Its instructional sequence, however, continues to divide that system into levels. Learners begin with phonemes and grapheme-phoneme correspondences in simple words. They proceed through increasingly complicated phonics patterns and controlled text. Morphology becomes prominent when words become longer. Vocabulary, syntax, comprehension, and writing may be addressed in other lesson components or later chapters.

That sequence assumes that a written word can first be understood phonologically and then enriched with meaning and structure. But morphology is not an optional layer added to a phonological spelling system. English spelling represents meaningful relationships from the beginning.

The plural suffix in <dogs> is the same suffix found in <cats>, even though its pronunciation is different. The base <sign> remains visible in <signal> and <signature>, even though its pronunciation changes. The spelling of a word cannot always be accurately explained by starting with its pronunciation and treating whatever does not fit as irregular.

The writing system has not separated phonology, morphology, syntax, and semantics into instructional stages. We are the ones doing that.

The book questions some of the practices inherited from phonics programs

One reason this book is so useful is that it does not simply defend every practice associated with Structured Literacy or Orton-Gillingham.

Chapter 3 is even more direct when it discusses syllable division.

That is a significant concession because syllable types and syllable-division rules are central features of many programs promoted as highly structured and evidence based. Students may spend years identifying vowel patterns, labeling syllable types, dividing written words, and reciting procedures before attempting a pronunciation.

Yet this Structured Literacy text acknowledges that the rules may be inconsistent, unnecessarily complicated, poorly supported, and especially difficult for learners with limited working memory.

The recommended alternative is greater flexibility. Students may identify familiar patterns, try possible pronunciations, use their oral vocabulary, locate meaningful elements, and adjust an initial pronunciation until they recognize a real word.

When the traditional rules fail, the learner has to draw upon more of the language system.

The morphology chapter knows more than the framework allows it to use

Chapter 3 contains some of the strongest arguments for morphological instruction in the entire book.

They also state that morphemes provide information about meaning, spelling, and pronunciation at the same time. Because long words frequently contain fewer morphemes than syllables or grapheme-phoneme units, morphological analysis can reduce the number of pieces a learner must hold in memory.

Here, morphology is doing much more than helping a student divide a long word into larger pronounceable chunks. A morpheme coordinates information about meaning, spelling, and pronunciation at the same time.

A learner investigating <unpresentable> can identify <un->, <present>, and <-able>. Those elements do not simply shorten the word into manageable pieces. They explain its structure, connect it to relatives such as <present>, <presentation>, <represent>, and <presentable>, contribute to its meaning, and provide a context for examining its pronunciation and graphemes.

The chapter recognizes this efficiency. It recognizes that morphology reduces memory demands. It recognizes the value of word families. It recognizes that morphemes connect spelling, pronunciation, and meaning.

And then the instructional recommendations pull back from that understanding.

For the purpose of decoding long words, the chapter says that explicit instruction in the meanings of affixes and bases is not one of its recommended strategies. Meaning may be added when vocabulary is also being addressed, but it is not treated as necessary to the decoding procedure itself.

The letters <un> occur at the beginning of both <unhappy> and <uncle>. They are a prefix in one and not the other.

The letters <er> occur at the end of both <teacher> and <father>. They are a suffix in one and not the other.

Position and pronunciation do not establish morphological structure. Meaning and morphological relatives provide the evidence.

Yet the recommended routines often involve memorizing lists of “beginnings” and “endings,” circling possible affixes, pronouncing them, and calling what remains the root. The book includes lists that treat <ion>, <tion>, <ation>, and <ition> as separate suffixes to be memorized and pronounced.

This is morphology filtered through a phonics framework.

The learner is taught to recognize recurring letter strings and use them as pronunciation units. The meaningful structure of the word remains secondary.

That is very different from investigating a word to establish its meaning, identifying its written morphemes, examining its relatives, and then determining what the graphemes are doing within those morphemes.

We cannot accurately analyze the graphemes until we know which written morphemes we are looking at.

The spelling chapter describes a richer system than its lessons teach

The spelling chapter also moves between two different understandings of English orthography.

At times, it provides an excellent description of the system. It acknowledges that written English represents morphological relationships as well as grapheme-phoneme relationships. It explains that morphology is necessary to understand spellings such as <dogs>, <healthy>, and <psychology>, and that morphological knowledge contributes to vocabulary, reading, and spelling.

It recognizes that related words preserve meaningful spelling connections across changes in pronunciation. It refers to etymology as a useful source of evidence. It rejects the idea that high-frequency words are stored through visual memorization alone.

Then the instructional activities return to familiar categories:

• magic <e>

• VCe words

• syllable types

• the “floss rule”

• dropping silent <e>

• changing <y> to <i>

• regular and irregular spellings

These procedures may help learners produce some conventional spellings, but they do not necessarily help learners construct an accurate understanding of the system. And for many struggling learners they don’t help much at all.

Consider a word such as <making>.

A rule-based explanation tells the learner to “drop the silent <e> before adding <-ing>.”

A morphological analysis begins with the word sum:

make + ing → making

The word sum identifies the base and suffix and allows us to investigate what happens to the final, non-syllabic <e> during suffixing. The learner can then compare <making> with <maker>, <makes>, <made>, and <remake>.

One explanation gives the learner a procedure.

The other shows the structure and relationships that explain the spelling. When a word is studied in its family, everything comes together.

Suffixing conventions make much more sense once we notice the subtle patterns in the writing system. The <e>is replaced by a vowel suffix and returns when it is needed. The <y> and <i> change places while continuing to write the same phoneme.

That contradiction runs throughout the book. It frequently describes English as a meaningful, morphophonemic writing system, but continues to teach it through phonics categories supplemented with morphology.

Controlled text can conceal the very weaknesses intervention should reveal

The discussion of decodable text contains another revealing contradiction.

The book strongly supports decodable text for learners in the early stages of word reading. It presents controlled text as a way to practice taught phonics patterns and discourage guessing.

Later, however, the book presents the case of Samantha.

Samantha appears to comprehend the decodable texts used during intervention. Her interventionist therefore does not recognize comprehension as a concern. Broader assessment reveals weaknesses in oral comprehension and vocabulary in addition to poor word reading.

The explanation provided in the book is straightforward: the decodable text does not place heavy demands on comprehension, so Samantha’s language weaknesses are less visible than they would be in grade-level material.

Controlled text controls more than phonics. It also controls the language available to the learner and the language abilities visible to the teacher.

That does not mean that every use of decodable text causes comprehension difficulties. It means we cannot use performance in linguistically restricted text as evidence that comprehension is developing normally.

It also raises a broader instructional concern. If a learner spends a substantial amount of intervention time reading language designed primarily around a phonics sequence, what opportunities are being lost?

The book attempts to solve this problem by adding read-alouds, oral vocabulary instruction, and separate comprehension activities. That is better than ignoring language altogether, but it preserves the division: rich language is heard, while the learner reads controlled language.

The learner’s decoding system and language system are being developed in parallel rather than through the same meaningful written material.

The strongest chapters move beyond phonics-first instruction

The later chapters contain some of the book’s most clinically useful work.

The oral-language chapter addresses sentence-level and discourse-level processing, narrative structure, inferencing, and figurative language. It recognizes that serious language and reading difficulties can go unnoticed in children who decode accurately and appear to read well.

It also notes that sentence-level confusion is often neglected, even in programs that include morphology and vocabulary.

That is describing Structured Word Inquiry to a,“T.”

The writing chapter emphasizes topical vocabulary, oral rehearsal, sentence formulation, cohesion, and paragraph structure. It criticizes disconnected language exercises that force students to shift among unrelated topics and instead recommends using the same meaningful vocabulary across listening, speaking, reading, and writing.

A learner does not first become a decoder and later become a language user. The learner brings oral language, concepts, experiences, vocabulary, grammar, speech, and personal history to every written word encountered.

Some learners bring significant weaknesses in those areas. That makes it even more important to work with them together rather than postponing meaning while a phonics sequence is completed.

When I begin with a meaningful written word, I can work at the learner’s level across all of these areas.

We can discuss what the word means and how it is being used. We can examine its morphological structure. We can find relatives. We can investigate the jobs of its graphemes. We can compare pronunciations. We can build sentences, discuss grammar, read a related text, and use the word in writing.

The components do not have to wait in separate lines.

Explicit and systematic do not mean phonics first

One of the most persistent confusions in literacy education is the assumption that rejecting a phonics-first sequence means rejecting explicit or systematic instruction.

It does not.

Instruction can be highly explicit without reducing written words to phoneme sequences, and it can be systematic without following a predetermined order of CVC words, blends, digraphs, silent <e>, vowel teams, syllable types, and affixes.

A system is not the same thing as a sequence.

English orthography is structured. Its spellings are constrained by morphology, etymology, phonology, and conventions for how written morphemes are represented. A systematic approach should help learners investigate that system.

The instructional order can begin with a meaningful word and move through a consistent set of questions:

What does the word mean here?

What is its structure?

What related words provide evidence?

What are the graphemes doing within the morphemes?

That is explicit. It is systematic. It is cumulative. It can be carefully scaffolded. It can include direct teaching and corrective feedback.

What it does not do is assume that phonology must organize everything else.

Structured Literacy has reached the limits of adding components

The best parts of this book do not support a narrow phonics program.

Morphology has been added, but phonics still determines the instructional entry point.

Meaning has been added, but it may remain separate from decoding.

Syntax and comprehension have been added, but often in different lesson components.

Writing has been added, but word instruction may still begin with isolated sound-symbol correspondences.

At some point, adding more components stops solving the problem because the architecture itself needs to change.

The next step is not to place morphology beside phonics, vocabulary beside morphology, and comprehension beside vocabulary. The next step is to recognize that written language coordinates all of these.

A more useful question for instruction is this:

That question changes the instructional order.

We first establish which meaningful word we are investigating and how it is being used. Then we determine its morphological structure. Once the written morphemes are identified, we can accurately investigate the jobs of the graphemes within them. Word history and pronunciation provide additional evidence when needed.

Phonology remains essential evidence, but it is no longer the organizing principle of the writing system.

I’m glad this book was posted because its strongest chapters show how far the field has already moved. Structured Literacy has become broader, more flexible, more language based, and more attentive to the actual needs of learners.

Now its foundation needs to catch up.

Parents looking for help with dyslexia are often told to find an Orton-Gillingham tutor, choose an Orton-Gillingham curriculum, or make sure the school uses an Orton-Gillingham-based intervention.

Some training providers, private schools, and tutoring businesses call Orton-Gillingham the “gold standard” for dyslexia instruction. In the sources I found, that description came from organizations providing, promoting, or selling Orton-Gillingham services. It was not a designation awarded by an independent scientific organization.

Calling an intervention a gold standard carries a specific implication. It indicates that the intervention has been clearly defined, studied against credible alternatives, and shown to be the best available option.

Orton-Gillingham has not met those conditions. The research has not established its superiority over other interventions, and Orton-Gillingham does not identify one uniform intervention in the first place.

The Orton-Gillingham Academy describes it as an approach rather than a method, program, or system. The Academy says that trained practitioners design lessons and materials around the student’s current level, strengths, and difficulties.

An approach provides principles that guide instruction. It does not require everyone using its name to teach the same material in the same order, use the same lesson format, require the same amount of practice, or complete the same practitioner training.

One instructor may develop each lesson in response to the student. Another may follow a commercial program with fixed books, levels, procedures, controlled texts, and requirements for advancement. One child may receive several hours of instruction each week from a practitioner who completed extensive coursework and supervised teaching. Another may work with someone who learned the procedures for the current program level through a set of videos.

Both services may be called Orton-Gillingham.

When a parent says that a child is receiving Orton-Gillingham instruction, we still do not know what happens during the lesson. The child may be tapping sounds, manipulating tiles, tracing letters, dividing written words into syllable types, memorizing spelling rules, reading word lists, working with nonsense words, completing dictation, studying morphology, reading controlled passages, or writing original sentences.

We do not know where the child began, which skills must be mastered before the child advances, how long the sessions last, how often instruction occurs, how much freedom the provider has to alter the curriculum, or how much supervised preparation the provider completed.

The name identifies a broad instructional tradition. It does not describe a particular child’s intervention.

What the formal standards require

Although Orton-Gillingham is not one program, the Orton-Gillingham Academy has established principles for its version of the approach.

The Academy describes OG instruction as diagnostic and prescriptive, individualized, language-based, alphabetic and phonetic, simultaneous and multisensory, direct and explicit, structured, sequential, cumulative, synthetic, analytic, cognitive, and emotionally sound.

Its fuller description shows how the instruction is organized. It begins with phonemes and the alphabetic principle, then proceeds toward more complex word and text structures. Synthetic instruction moves from parts to wholes, such as blending sounds, syllables, and morphemes to read words. Analytic instruction moves from wholes to parts, such as segmenting words into sounds, syllables, or morphemes for spelling.

The sequence begins with phonemes and sound-letter relationships. Morphology and larger written structures are included, but they enter an instructional framework organized from smaller parts toward more complex forms. The programs are organized around sounds and sounding-out.

The International Multisensory Structured Language Education Council, another influential organization in this instructional tradition, provides more detailed accreditation standards. Its required content includes phonological and phonemic awareness, sound-symbol associations taught and mastered in both directions, blending, segmenting, six syllable types, directly taught syllable-division rules, morphology, syntax, and semantics. Its instructional standards include simultaneous visual, auditory, kinesthetic, and tactile teaching; systematic review; direct instruction; and mastery to automaticity.

These are not occasional activities selected by a few tutors. Blending, segmenting, written syllable categories, syllable-division procedures, and formal multisensory routines are built into the standards used to accredit multisensory structured language programs.

Morphology is required too. It would be inaccurate to say that Orton-Gillingham excludes bases, roots, prefixes, suffixes, or word relationships. The concern is how morphology is positioned within the sequence and how it is taught.

In one widely used commercial system, students begin with phonemic-awareness lessons that include CV, VC, CVC, VCC, and CCV nonsense forms. Closed syllables appear at Level 3, syllable division at Level 4, prefixes and suffixes at Level 5, and Greek forms and Latin roots at Level 10.

Morphology is present, but students reach much of it only after working through a sequence governed by phonemic awareness, sound-symbol correspondences, syllable categories, vowel patterns, and spelling rules.

This reflects a theory about which information children need first and which knowledge may be withheld until later. This is a concern for struggling learners who can’t get past the lower levels. The learners who can’t blend or segment or sound words out even though they have tried for several weeks, months, or years.

What does “individualized” mean?

In the Academy’s practitioner model, individualized instruction means that the instructor continually studies the learner’s spoken, written, verbal, and nonverbal responses. That information is used to plan the next lesson, address difficulties, and build on progress.

Commercial programs may use individualized in a more restricted way.

A publisher may determine the levels, lesson order, terminology, rules, word lists, activities, controlled texts, and mastery assessments. The tutor can slow the pace, repeat a lesson, add practice, provide more prompts, or delay advancement. However, the student continues to move through the publisher’s sequence. So, the amount of practice has been adjusted but the curriculum has not been changed.

Official program materials show how different commercial OG-based systems can be. One describes a complete lesson as a ten-part, 90-minute session and recommends at least two complete lessons each week. It also requires students to reach specified mastery levels before proceeding to the next step.

Another commercial system tells new tutors that they do not need to learn the entire program before beginning. They can learn and teach one level at a time, watch the training for that level, and begin tutoring. Its public materials state that most students, regardless of age, must begin with Level 1. The tutor training contains five to six hours of instruction per level, and the tutor manual provides complete lesson plans and optional full scripting.

Those systems differ in dosage, provider preparation, lesson structure, materials, scripting, and expectations. Describing both as Orton-Gillingham conceals most of the information a parent would need to compare them.

When the sequence controls the teaching

I recently read a tutor’s question about an incoming second-grade student who was nearing the end of the second book in a commercial OG-based program. The child did not yet have short <e> and <i> “solid.”

The tutor had spent several sessions working only on those two vowels. She had removed digraphs, used vowel tapping, introduced pictures and special names for the sounds, and brought in a mirror so the child could watch the mouth movements. She was concerned about keeping the student in the same lesson for too long and asked whether she could move into the next book. She noted that she has seen many students still struggle with those phonemes for up to a year.

The responses offered more ways to practice the contrast. People suggested a separate phonemic-awareness program, a short-vowel game, physical and visual prompts, slowed pronunciation, and advance warnings before dictating a word containing <e> or <i>.

The tutor had reached the limit of the program’s decision-making structure. She could continue practicing the required contrast, add more prompts, or move forward without the expected level of mastery.

Other questions could have been asked. Does the confusion occur in reading, spelling, or both? Does it occur in every phonological environment? How are the vowels pronounced in the child’s dialect? Does the student distinguish the written words when their meanings are clear? Would work with actual words and their relatives provide more useful information than additional isolated-vowel practice?

The program had already defined the problem as insufficient mastery of <e> and <i>. The teaching therefore remained focused on producing the performance required for the next lesson.

This is one way a fixed sequence can be called individualized. The tutor varies the number of sessions, the prompts, and the supporting activities while the publisher continues to determine what the child must accomplish before moving forward.

When a required practice becomes an obstacle

Parents are often told that certain activities are necessary because they are included in an established program. That answer is inadequate when the activity is not helping the child.

What if a child does not benefit from nonsense words?

Some students may complete nonsense-word tasks without difficulty. Other students find them confusing, frustrating, or disconnected from everything they know about language. A child may perform poorly because the form has no meaning, no familiar context, no morphological relatives, and no established lexical representation.

If nonsense words repeatedly produce errors without improving the child’s reading and spelling of actual words, can the instructor stop using them? Or does the program require continued practice because nonsense-word performance is built into its sequence or assessment system?

The same questions apply to written syllable types.

What if labeling written letter patterns as closed, open, vowel-consonant-e, vowel team, vowel-r, or consonant-le does not help the child read or spell? What if dividing words according to written syllable procedures increases confusion? What if the child attends so closely to the category and division routine that the word’s morphology becomes harder to see?

Can the instructor replace that procedure, or would doing so be considered a failure to follow the program?

A practice should not be protected from scrutiny because it appears in a recognized curriculum. The instructor should be able to explain what the activity is intended to teach, what evidence supports its use, how improvement will be measured, and what will happen if it creates an obstacle for the child.

Without that flexibility, the student is expected to adapt to the program while the program remains unchanged.

Research on one program cannot validate all of them

The variation among OG-based programs creates a basic research problem.

A study of one commercial program provides information about that program under the conditions used in the study. The results depend on who delivered the instruction, the preparation those providers received, the frequency and duration of sessions, the students who participated, the comparison instruction, and the outcomes the researchers measured.

A program delivered by trained therapists four or five days a week is not equivalent to a scripted program used twice a week by a parent or tutor. Shared terminology or historical influences do not make them the same intervention. Furthermore, research on one program cannot be transferred to every other program that describes itself as Orton-Gillingham-based.

The broader research does not establish an advantage for OG either. A 2021 meta-analysis examined 24 studies of Orton-Gillingham interventions. For foundational reading skills, the average effect size was 0.22 in favor of OG. For vocabulary and comprehension, the average effect size was 0.14 in favor of OG. In both cases, however, the differences between OG and the comparison conditions were not statistically significant. The reported p-values were .40 and .59.

The positive effect sizes mean that the numerical averages leaned toward OG. They do not mean that the researchers established that OG produced better outcomes. The evidence was too uncertain to rule out the possibility that the apparent advantage resulted from variation among the studies rather than a dependable benefit of OG.

This distinction is important. The meta-analysis did not show that children receiving OG learned nothing. It showed that the available studies did not demonstrate a reliable advantage for OG over the instruction received by the comparison groups.

The authors concluded that larger, more rigorous studies were needed before the effects of Orton-Gillingham interventions could be determined with confidence.

That is not evidence for calling OG the gold standard. At most, the pooled estimates pointed slightly in its direction while remaining too uncertain to establish that it outperformed the alternatives.Questions parents should ask

Parents need more information than whether a provider uses Orton-Gillingham.

They need to know which program or version of the approach is being offered and what their child will do during instruction. Where will the child begin? Does every student begin in the same book or level? Which skills must be demonstrated before advancing? What happens when a student repeatedly fails to reach the program’s mastery criterion?

Parents should ask how much time is spent on oral phonemic-awareness exercises, isolated sound-symbol practice, nonsense words, word lists, spelling rules, written syllable types, and syllable division. They should also ask when instruction begins to include morphological structure, vocabulary, grammar, comprehension, authentic text, and original writing.

They should ask what happens when a required activity does not help their child.

What if my child does not benefit from nonsense words?

What if written syllable types or syllable-division procedures create confusion?

What if tapping, tracing, or manipulating tiles becomes a distraction rather than a support?

Can the instructor remove an activity that is not helping, or must it remain because it is required by the program?

How will the instructor distinguish genuine learning from accuracy produced by advance warnings, visual cues, exaggerated pronunciation, repeated prompting, or extensive rehearsal?

Parents also need precise information about dosage. How long is each session? How many sessions are recommended each week? How long does the complete sequence generally take? What progress should be visible after several months, and how will that progress be measured outside the program’s own lessons and mastery checks?

The provider’s training should be described clearly. Did it include substantial coursework and supervised work with students, or was it limited to a workshop, implementation session, or a set of videos for the current level? Is the provider qualified to evaluate the child’s broader language and literacy profile, or primarily trained to administer one curriculum?

A child’s completion of lessons, books, or levels does not by itself show that the child can use the information independently. Parents should see evidence that learning is transferring into unfamiliar reading, spelling, and writing without the program’s prompts and materials.

Orton-Gillingham does not identify one treatment

Orton-Gillingham refers to an instructional approach with a long history and a recognizable collection of principles. It has also produced commercial programs that differ in lesson length, intensity, provider preparation, sequencing, scripting, materials, mastery requirements, use of nonsense words, treatment of written syllables, and timing of morphological instruction.

Some providers construct lessons around the individual student. Others administer a fixed commercial sequence. Some complete lengthy supervised training. Others learn one program level at a time. Some students receive several hours of instruction each week. Others receive much less.

These differences affect what students are taught and how their progress is interpreted. They also determine whether an activity can be removed when it does not help, or whether the child must continue because the activity belongs to the program.

A parent who has been told to find “an Orton-Gillingham program” has not yet received a specific recommendation. The curriculum, practitioner qualifications, dosage, lesson content, progression requirements, adaptability, and supporting evidence still need to be identified.

Before anyone calls Orton-Gillingham the gold standard, they need to specify which version they mean and provide evidence for that version. Orton-Gillingham is not one program.

A question comes up repeatedly among Orton-Gillingham tutors:

How do you decide that a student has mastered a skill and is ready to move on?

The answers vary. Some programs reportedly require 80% accuracy, while others require 90% or 95%. Some instructors want to see the same performance across three sessions. Others consider the child’s reading, spelling, drills, and passage work together.

The discussion usually turns into a debate over which number is correct. I am more concerned about what happens to the child while the program waits for that number.

What is the child being required to complete before moving forward? It might be a syllable type, a grapheme-phoneme correspondence, several graphemes associated with a phoneme, a spelling rule, or a collection of words chosen to demonstrate a pattern. Whatever the unit, the child practices it repeatedly until the required score is reached.

The child reads it, spells it, marks it, reviews it, and practices it again. When the score remains too low, the child receives more instruction on the same restricted material. Meanwhile, bases, affixes, meaningful word families, longer words, and other parts of the writing system remain several levels away.

The mastery criterion does more than record performance. It controls which parts of written English the child is permitted to study.

80% of what?

Before deciding whether mastery should be 80%, 90%, or 95%, we need to know what the percentage represents.

Did the child read eight of ten practiced words correctly? Did the child spell eight dictated words? Was the child reading a controlled passage built around the lesson’s target? Were the words directly taught, or had the child never seen them before? Was the child spelling isolated words or writing an original sentence? Was the score collected immediately after instruction or several weeks later?

These tasks place different demands on the learner and provide different information.

A child may read a word correctly while looking at its completed spelling and still be unable to produce that spelling from memory. A child may spell a word correctly in dictation and omit part of it while composing a sentence. A child may perform well on words practiced throughout the lesson and then fail to recognize the same relationship in an unfamiliar word.

Combining these performances into one mastery score hides the differences among them.

What does a high score show?

Suppose a lesson focuses on one spelling feature. The instructor introduces the target, practices it during drills, presents a list of words selected to contain it, dictates more words containing the same feature, and finishes with a controlled passage that includes additional examples.

The child scores 95%.

The score shows that the child performed accurately on material organized around the feature that had just been taught and practiced. The child knew which category was under discussion, and the instructor had narrowed the range of possible responses. Other spellings and structures may have been deliberately excluded from the lesson.

The score cannot tell us whether the child will recognize when that information is relevant in another word. It cannot tell us whether the child can distinguish it from other possible spellings, retrieve it during independent writing, or remember it after several more program levels have been introduced.

A child can become very good at completing the task placed directly in front of them without developing knowledge that remains available elsewhere.

What did 98% accuracy measure?

Roxanne Hudson and her colleagues studied 56 second-grade struggling readers receiving a small-group decoding intervention. One group practiced each page until the students reached 98% accuracy. The other group had to meet both the 98% accuracy requirement and a reading-rate requirement.

The students who met the accuracy requirement learned to read the practiced material accurately. When the researchers added a rate requirement, the students also became faster and more automatic in decoding. The additional requirement did not produce greater improvement in comprehension or in fluency while reading grade-level text.

The study separated decoding accuracy, decoding automaticity, comprehension, and grade-level text reading because they were not the same outcome. Reaching 98% accuracy on the practiced decoding material showed that the students had reached 98% accuracy on that material. It did not demonstrate broad mastery of reading.

OG-based programs often take a score from a narrowly controlled task and use it to make a much larger instructional decision. A child who has not reached the cutoff may be required to remain at that point in the sequence, even though the score says nothing about whether the child would benefit from learning more about morphology, word families, meaning, or the structure of other words.

Reading and spelling do not progress together

Reading a printed word and generating its spelling are different tasks.

During reading, the completed word is visible. The child can use the letters on the page, familiarity with the word, and the surrounding sentence. During spelling, the child must retrieve the word’s structure and choose among several possible graphemes without seeing the completed form.

Spelling during composition adds still more demands. The child is generating ideas, organizing language, remembering punctuation, and producing the written form at the same time.

A child may make substantial progress in reading while continuing to struggle with spelling. Holding the child at the same reading level until spelling catches up does not resolve that difference. It restricts reading instruction because a more demanding task is developing at a different rate.

The instructor needs to examine what the child can do in reading and what the child can do in spelling. An average of the two produces a clean number but a poor description of the learner.

Performance on taught words is not transfer

Literacy researchers routinely separate words that were directly taught from unfamiliar words that contain related structures. They may also assess whether children can use taught spellings in sentences and whether those spellings remain available after instruction.

They separate these outcomes because success on one does not establish success on the others.

A 2024 systematic review and meta-analysis of morphological instruction found the largest effects on words that children had directly studied. Evidence of transfer to untrained words was more limited and differed for reading and spelling.

The finding is not an argument against teaching specific words. Children learn a great deal by studying words carefully. It does show why we cannot take a high score on a practiced list and assume that the child has acquired a general skill that will automatically transfer to other words.

A score of 95% on the current lesson’s list describes performance on that list. The child’s ability to use the information elsewhere still has to be observed.

Passing the level does not make the learning durable

Tutors frequently describe children who completed a concept many lessons earlier and continue to misspell words associated with it despite cumulative review.

The program recorded mastery, but the spelling did not remain available.

The child may have performed well because the target had been practiced intensively that week. The list may have contained predictable examples. The child may have remembered the specific words long enough to pass the lesson without developing an understanding that could be applied more broadly. The information may have been available during reading but not during spelling, or during dictation but not during composition.

The original score may have been accurate. The child did achieve the required percentage on that task at that time. The program treated that temporary performance as evidence that the skill had been secured.

A sequence of passed lessons does not necessarily become a connected and retrievable understanding of written language.

Three sessions is still only three sessions

Checking performance across three sessions is more informative than relying on one score collected immediately after instruction. The child has at least returned to the material more than once.

However, three successful performances do not show that a spelling has become permanently available. Many children need to encounter words and their relatives repeatedly over long periods. A spelling may need to be revisited ten times, twenty times, or whenever it reappears in reading and writing.

There is no need to replace three sessions with another fixed number. The problem begins when review is treated as a reason to stop introducing anything else.

Instruction can continue while earlier words and structures are revisited. A child can study a new family and return to a previously confusing spelling. A later word may clarify something that did not become clear during the earlier lesson. New learning does not always depend on complete command of the previous unit.

When mastery becomes a gate

A child who has not reached the required criterion for a syllable type, spelling pattern, grapheme, correspondence, or word set may spend additional weeks practicing the same category. This is described as careful sequencing or prevention of gaps.

It can produce a different kind of gap.

The child spends months practicing pieces of the writing system without being shown how those pieces connect. The child may be required to complete short-vowel levels before studying bases and affixes, or to master syllable types before examining how words are related.

Written English is not organized as a succession of isolated skills. Words have morphological structure, phonological structure, and histories. Their spellings represent meaning and relationships as well as pronunciation.

A child does not need to master every short-vowel category before studying a base and its family. A child does not need to complete every syllable type before examining how a suffix changes a word. A child does not need to achieve 95% accuracy with a digraph before learning why a particular word is spelled as it is.

Those relationships may help the child understand and remember the spelling. When the program saves them for later, it may withhold the information the child needs now.

More drilling may repeat the original problem

When a child fails to reach the required score, the usual response is more practice with the same material. The child receives more review cards, more word lists, more marking procedures, more syllable division, and more reminders to apply the program’s strategy.

If the information is not becoming durable, repeating the same restricted presentation may accomplish little.

An actual word offers more than a spelling pattern. It has a meaning, a structure, and relatives. Its pronunciation may change across the family while its spelling remains stable. Its history may explain a letter that cannot be accounted for by examining the present pronunciation alone.

A child who is struggling with an isolated spelling may need to see the base that remains consistent across several related words. The child may need to compare two words that have been confused, talk about how their meanings differ, and examine the structures that account for their spellings. The child may need to use the word in a sentence and encounter it again in meaningful reading.

These are not enrichment activities to be added after the basic sequence has been completed. They are part of learning how the writing system works.

The child may need more information rather than more repetitions of less information.

Strategies can consume working memory

OG-based lessons may require children to follow several procedures before reading a word. They may identify a syllable type, mark vowels, locate consonant patterns, divide the word, recall a rule, assign a pronunciation, blend the parts, and decide whether the result is a familiar word.

For some children, the procedure becomes another task layered on top of reading. The child is trying to remember what to mark, where to divide, which category applies, and which rule comes next.

A strategy is only useful when it makes the task easier. When the procedure uses more attention than the word itself, the child has fewer resources available for meaning, structure, and recognition.

We can ask what the word means, how it is built, which words are related to it, and what the letters are doing within its morphemes. The child can investigate the word itself rather than trying to process it through a memorized series of steps.

Children do not need to earn access to morphology

In many phonics-based sequences, morphology is introduced after children have completed levels devoted to correspondences, syllable types, and spelling rules. This arrangement treats morphology as advanced content.

Words have morphological structure from the beginning.

A young child can understand that <help> is related to <helped>, <helping>, and <helpful>. The child can see the same base in each word and notice what changes when a suffix is added. The words can be simple, the family can be small, and the explanation can be adjusted to the child’s language and attention.

The child does not need to complete a phonics sequence before being shown that words have structure and relatives. We cannot accurately investigate the jobs of graphemes until we have identified the morphemes in which those graphemes occur.

Word families are not material to be saved until the child has earned access to them. They are part of beginning literacy instruction.

What happens when the gate is removed

Every time I begin working with a child who has spent months or years in an Orton-Gillingham-based program, I see some version of the same thing.

The child has practiced rules, syllable types, sound-symbol correspondences, marking procedures, and strategies. The child may have drilled the same categories repeatedly while waiting to meet a program’s mastery criterion, but no one has invited the child to investigate how words work.

Then we place words beside other members of their families.

We compare words the child has confused. We discuss what they mean, how they are built, which elements they share, how their spellings are related, and what changes in their pronunciation. We study spellings as representations of meaningful words rather than as isolated phonics patterns to be memorized.

The child lights up.

Again and again, children tell me that no one has shown them words this way. They tell me they dreaded returning to their previous tutoring sessions, but they happily come to our sessions and sometimes have said would happily spend all day investigating words with me! Talk about melting my heart, yes I have time we can do a few more.

A child who supposedly needed more drilling and more time at the same program level becomes curious, attentive, and eager to continue once the words are placed in meaningful relationships.

Perhaps the child had not failed to practice enough. Perhaps the child had been practicing information stripped of the connections that would make it understandable and memorable.

When a child’s engagement and understanding change as soon as we stop drilling isolated categories and begin studying meaningful words in families, I do not conclude that the child needed ten more lessons on the same syllable type.

I conclude that the child needed access to the writing system.

Children can investigate meaning, structure, relatives, and phonology from the beginning. They do not need to earn that access by first reaching 80%, 90%, or 95% on a restricted set of patterns.

When children finally see that words make sense, their response tells me what had been missing. That is all the evidence I need to stop keeping the rest of the writing system behind a mastery gate.

Move forward and keep returning

None of this means that we introduce a word or concept once and assume the child has learned it. Children need repeated encounters. Some spellings will need to be revisited many times and in several contexts.

Review does not have to prevent forward movement.

We can introduce new words while returning to earlier structures. We can encounter the same base in another relative, compare a previous spelling with a new one, and revisit a word when the child misreads it in a book or misspells it in a sentence.

Later learning may strengthen earlier learning. A child may understand a spelling more fully after encountering a related word. A new suffix may reveal the structure of a word studied several weeks earlier. A comparison between families may clarify a grapheme that never became secure through drills.

Sometimes the next investigation helps the child understand what came before.

The real question

The debate over whether mastery should be set at 80%, 90%, or 95% overlooks the larger instructional problem.

Why must a child reach a particular percentage on one restricted category before being allowed to investigate the rest of written English? Why should difficulty with a syllable type prevent the child from studying meaningful bases and affixes? Why should inconsistent spelling of a practiced pattern lead automatically to another round of the same drilling?

A percentage can describe how a child performed on a particular task. It cannot prove that the learning will last, that the child can use the information in unfamiliar words, or that another lesson on the same category is the best response.

It should not determine whether a child is ready to learn how the writing system works.

Children do not need to complete one fragment of English before they are allowed to see another. They need opportunities to study words, their structures, their relatives, their pronunciations, and their histories from the beginning.

We can continue returning to what remains difficult without keeping the rest of the writing system behind a mastery gate.

References

Colenbrander, D., von Hagen, A., Kohnen, S., Wegener, S., Ko, K., Beyersmann, E., Behzadnia, A., Parrila, R., and Castles, A. (2024). The effects of morphological instruction on literacy outcomes for children in English-speaking countries: A systematic review and meta-analysis. Educational Psychology Review, 36, 119.

Hudson, R. F., Isakson, C., Richman, T., Lane, H. B., and Arriaza-Allen, S. (2011). An examination of a small-group decoding intervention for struggling readers: Comparing accuracy and automaticity criteria. Learning Disabilities Research & Practice, 26(1), 15–27.

Somewhere along the way, adults decided that children who did not learn through the interventions they were given needed a name.

They called them treatment resisters.

Read that phrase again. The treatment did not work for the child, and the child was named as the person resisting it.

The child did not select the theory behind the intervention. The child did not choose the materials, establish the sequence, set the pace, determine how much repetition would be required, or decide what adults would do when progress did not occur. Adults made every one of those decisions.

Then the child did not learn, and the adults called the child the resister.

How can anyone look at that arrangement and not see the problem?

If the material and methods we selected did not teach the child, the first question should be what we need to change. Perhaps our understanding of the difficulty is incomplete. Perhaps the instruction depends on abilities the child does not have. Perhaps we are beginning with the most difficult and abstract part of the writing system. Perhaps the child needs a different route.

Instead, we created a label that places the action inside the child.

The word can be analyzed as <resist> + <er>. The <er> suffix names the person performing the action. A teacher teaches. A writer writes. A runner runs.

A resister resists.

The child is no longer described as a learner whom the intervention failed to reach. The child becomes someone who resisted the intervention.

That one grammatical move protects everything except the child. The treatment remains the treatment. The materials remain appropriate. The methods remain evidence-based. The theory remains intact. The adults retain their expertise.

The failure is transferred to the learner.

That is a blame-the-victim game.

The child becomes the problem

The phrase treatment resister did not remain confined to an unfortunate professional conversation. It appears in the formal reading-intervention literature. In 2000, Joseph Torgesen published a paper titled Individual Differences in Response to Early Interventions in Reading: The Lingering Problem of Treatment Resisters.

The term may have been intended as a technical way to classify research participants, but technical language is not exempt from scrutiny. Words reveal how a field has framed a problem, and that framing determines which questions researchers and practitioners ask next.

When the intervention is treated as the fixed point, the question becomes:

What is wrong with the children who did not respond?

What cognitive deficit explains their failure?

How are these children different from the responders?

How much more intensive must the treatment become?

Those questions may produce information about differences among learners, but they leave another question largely untouched:

What was missing from the instruction?

Of course children differ. They differ in language, memory, attention, sensory processing, motor skills, prior knowledge, educational history, emotional regulation, and countless other ways. Those differences should help us build more accessible instruction. They should not be used to shield an instructional approach from examination.

Yet when children do not progress, the usual response is to increase the dose.

More phonological awareness.

More segmenting and blending.

More grapheme-phoneme correspondence practice.

More controlled text.

More repetition.

More minutes.

More fidelity to the program.

The child has already shown that the route is not working, so we send the child down the same route more often and with greater intensity.

When the child still does not learn, the lack of progress becomes evidence of the severity of the child’s impairment. The intervention is rarely placed on trial. The child always is.

Then more labels appear.

The child is avoidant.

The child is noncompliant.

The child is unmotivated.

The child refuses to work.

The child is resistant to intervention.

The child is inattentive.

The child is oppositional.

The child is defiant.

With every new label, we move farther away from examining what the instruction is asking the child to do.

What exactly is the child resisting?

Before describing a child as resistant, we should examine what the child has repeatedly experienced.

Imagine being unable to segment and blend spoken words reliably and being asked to use segmentation and blending to identify unfamiliar written words.

A word is placed in front of you.

The adult tells you to sound it out.

You cannot.

The adult repeats the prompt.

You make an attempt, but the pronunciation is wrong.

The adult asks what else one of the letters could say.

You cannot hold several possible phonemes in mind while maintaining the rest of the word.

The adult stretches the sounds.

You still cannot blend them into a recognizable spoken word.

The adult points to the letters and waits.

Perhaps other children are waiting too.

Then the same thing happens with the next word and the next one.

The child is not receiving new information. The child is being repeatedly asked to perform an ability the child does not possess.

At some point, an unfamiliar word stops feeling like an opportunity to learn. It becomes a test the child expects to fail.

Some children freeze. They stare at the page and stop speaking. Some immediately say, “I don’t know,” before looking closely at the word. Some put their heads down, cover their faces, or pull their hoods over their eyes.

Other children become silly. They joke, change the subject, drop materials, break pencils, or suddenly need to go to the bathroom. Some leave their seats, refuse to enter the room, push the book away, tear the paper, yell, kick, or hit.

Adults sometimes call these unwanted behaviors.

From the child’s perspective, the behavior may be the only reliable way to end a demand that has become intolerable.

The child may not be resisting reading.

The child may be resisting another experience of public failure.

Children want to be successful

Children do not arrive at school hoping to fail. They do not enjoy watching classmates read easily while they struggle over every word. They do not choose the embarrassment of being corrected again and again in front of teachers, tutors, classmates, or parents.

Children want to be successful.

When a child stops looking at the page, rushes through the work, changes the subject, gets out of the chair, or refuses to try, adults often decide that the child is not paying attention. Attention becomes the explanation for why the instruction is not working.

I hear it constantly.

“He could do it if he would just pay attention.”

“She knows this. She needs to focus.”

“He is rushing.”

“She is not even looking at the word.”

There is always a reason for behavior. Not paying attention is a description of what an adult sees. It is not an explanation.

Sometimes the child is not attending because the child has no idea what to do.

Sometimes the child looks away because looking at the word has never provided a usable way into it. Sometimes the child rushes because slowing down only extends an experience that has repeatedly ended in failure. Sometimes the child refuses to try because experience has taught the child that trying will lead to another wrong answer and another demand to try again.

That is not a child who does not want to learn.

That is a child who has learned what to expect.

A child may be able to sustain attention while drawing, building something, talking about an area of interest, listening to a story, playing a game, or investigating a meaningful family of words. Then the same kind of phonics task the child has failed hundreds of times is placed on the table, and the child is suddenly described as inattentive.

The change in behavior should make us curious about the task, not just suspicious of the child.

What is the instruction asking the child to do?

Does the child possess the abilities required to do it?

Does the child understand what is being asked?

Is there a meaningful way into the work?

Does the child expect help, or another round of guessing followed by correction?

What changes when support is provided immediately?

What happens when the materials or instructional approach change?

What happens when the child is allowed to investigate the word rather than being asked to produce it through a method that has never worked?

Then someone says ADD

ADHD is sometimes a valid diagnosis. It can affect a child’s regulation of attention, working memory, inhibition, organization, and persistence. It can also occur alongside dyslexia, language disorders, anxiety, depression, and other conditions.

But ADHD cannot become the automatic explanation whenever a child disengages during difficult academic work.

A proper ADHD diagnosis is not based on one teacher noticing that a child does not attend during reading instruction. Diagnosis involves information from different settings and consideration of other conditions that may produce similar symptoms or occur alongside ADHD, including anxiety, depression, sleep problems, learning disabilities, and language-related difficulties.

Too often, adults see inattention and stop looking.

The child’s attention becomes the explanation for why the intervention failed. Parents may be encouraged to seek an evaluation or medication so the child can sit longer, try harder, and attend more closely to the same instruction that has not taught the child to read.

I have watched this happen. Instead of asking why an intelligent child can attend to countless other activities but becomes distressed, distracted, or unavailable during literacy instruction, the attention is treated as the entire problem.

Medication may be appropriate and extremely beneficial for a child who has ADHD. This is not an argument against diagnosing or treating ADHD.

It is an argument against using medication as a way to make a child tolerate ineffective instruction.

Medication is not literacy instruction.

It cannot explain how a word is built.

It cannot connect a spelling to meaning.

It cannot reveal a family of related words.

It cannot make a phonology-first approach accessible to a child who has been unable to segment or blend despite years of practice.

A child can have ADHD and still be receiving instruction that does not work. A child can have ADHD, dyslexia, and a language disorder. A child can have ADHD and a history of educational trauma.

A child can also be treated as though attention is the primary problem when the child is responding predictably to years of confusion, pressure, and failure.

Before deciding that a child cannot learn because the child will not attend, we need to make sure we have provided something the child can understand, participate in, and reasonably expect to succeed with.

Educational trauma can accumulate

I use the term educational trauma to describe the effects of repeated experiences of helplessness, humiliation, fear, failure, and loss of control within an educational setting.

It does not always begin with one dramatic event.

It can accumulate slowly.

The child is repeatedly asked to perform in front of other people.

The child repeatedly watches classmates do easily what requires enormous effort for them.

The child is repeatedly removed for intervention that does not solve the problem.

The child is repeatedly told to try harder, slow down, focus, or use a strategy that has never worked.

The child is repeatedly given less interesting and less meaningful material than peers are reading.

The child is repeatedly praised for effort while privately concluding that effort makes no difference.

Adults may experience each event as small. The child experiences the accumulation.

Some children begin to panic when they see a word list or a decodable book. Some cry during homework. Some develop stomachaches before school. Some become perfectionistic because making a mistake has become unbearable.

Some withdraw and become quiet. Some conclude that they are stupid. Some become anxious or depressed.

Others become angry.

Research cannot tell us that poor instruction caused every emotional or behavioral difficulty experienced by a struggling reader. It does tell us that these difficulties do not occur in separate compartments. A 2023 systematic review and meta-analysis covering 96 studies and more than 83,000 participants found that people with reading, mathematics, or unspecified learning difficulties experienced more internalizing difficulties, including anxiety, depression, somatic complaints, and social withdrawal, than comparison groups.

A 2024 meta-analysis also found a significant association between lower reading proficiency and externalizing behavior among children ages 5 through 12. The authors were careful to describe a correlation rather than claim a single direction of cause, but the relationship should make educators cautious about treating reading and behavior as entirely unrelated problems.

We tend to treat the quiet child as emotionally distressed and the loud child as badly behaved, even when both may be responding to the same history of failure.

Behavior does not appear without context. It has a history.

We need to be careful with oppositional behavior

Oppositional defiant disorder is another valid diagnosis. I am not claiming that every child diagnosed with ODD was simply taught badly or that every instance of refusal or aggression is caused by school.

I am saying that we need to examine trauma, communication, disability, academic demands, and the environments in which behavior occurs before deciding that defiance is an inherent feature of the child.

A child whose nervous system has learned to anticipate embarrassment or helplessness may fight, flee, freeze, or shut down when the triggering demand appears. The response may look oppositional to the adult issuing the instruction. It may feel protective to the child.

Research has found associations between trauma exposure and ODD. One clinical study found that traumatic victimization contributed uniquely to the prediction of an ODD diagnosis, and later research has continued to examine how adverse experiences and biological factors may interact in the development of oppositional behavior. None of this means that trauma explains every case. It means trauma history should not be ignored.

Before asking how to make the child comply, we should ask what happens when the task is presented.

Does the behavior occur across environments, or primarily during reading and writing?

Does it begin when particular materials appear?

What does the child believe is about to happen?

What abilities does the task require?

Does the child possess those abilities?

What happens when the adult supplies immediate help?

What happens when the demand to sound out every unfamiliar word is removed?

Does the child remain regulated during meaningful conversation, shared reading, or a guided word investigation?

No one should be permitted to hit or harm another person. Adults are responsible for maintaining safety.

But keeping people safe does not require us to stop investigating why the behavior occurs.

Calling behavior noncompliance does not explain it.

Calling the child oppositional does not tell us what to teach differently.

Support is not cheating

I do not make children struggle through every unfamiliar word to prove that they can identify it independently.

While we are reading, I give them the support they need.

That is not teaching children to guess from context. The word is directly in front of the child. I am helping the child connect a known meaning and pronunciation with the visible spelling.

The child remains connected to the story, the information, and the language. Reading does not have to stop so that every unfamiliar word can become another test.

I make a note of the word, mentally or on paper, because the hesitation gives me useful information. We may return to it later and discuss its meaning and structure. We may compare it with one or two relatives, write a word sum, or carry out a fuller investigation with a matrix. Then we can examine the phonology within a meaningful family.

Not every word requires a full investigation. The amount of work depends on the learner, the word, and the purpose of the reading.

What becomes harmful is repeatedly withholding needed support, forcing the child to perform a skill the child does not have, and treating every hesitation as an opportunity to test the same method that has already failed.

Support allows the child to remain in the learning experience. It also gives the adult information about what needs to be taught next.

The child is not resisting learning

Children should encounter difficulty. They should have opportunities to think, revise, persist, and work through uncertainty.

But productive difficulty is not the same as repeated helplessness.

The child is not being asked to pull an answer out of nowhere.

The teacher is not standing back and waiting for the child to fail.

The teacher also is not performing an elaborate linguistic investigation while the child watches passively.

The teacher knows the writing system well enough to guide the child into it.

That changes the relationship.

The child is no longer the subject on whom a treatment is performed. The child is a participant in an investigation.

The teacher is no longer protecting a program. The teacher is responding to a learner.

A child who has not learned through an intervention is not a treatment resister.

The child is a learner the intervention did not reach.

That distinction determines what happens next.

If the child is the resister, we try to change the child. We increase compliance, add rewards and consequences, create behavior plans, demand more attention, pursue medication, and intensify the same treatment.

If the intervention failed to reach the learner, we change our instruction.

Adults created the materials. Adults selected the methods. Adults determined the expected response. Adults hold the power to change what happens next.

The child did not resist our treatment.

Our treatment failed to provide a path the child could use.

The real resistance begins when adults are shown that the path is not working and refuse to consider another one.

For decades, one question sat at the center of arguments about reading instruction:

What should children do when they come to a word they do not recognize?

Should they look at the picture? Use the first letter? Think about what would make sense in the sentence?

Or should they look at the letters, assign sounds to them, blend those sounds together, and pronounce the word?

The phonics reform movement was right to reject picture cues, three-cueing, predictable texts, and instruction that taught children to look everywhere except at the written word.

Balanced literacy taught children to guess a word from its context. Phonics teaches children to generate a possible pronunciation from the letters and then check whether that pronunciation sounds like a word they know. When it does not, they are expected to try other phonemes until one of the attempted pronunciations produces a recognizable word.

The technical name for this is set for variability.

I call it guess and check.

Guessing did not disappear

Set for variability is described as the ability to resolve the difference between the pronunciation produced through decoding and the actual pronunciation of a known word. The child generates an approximation, recognizes that it does not sound quite right, and adjusts one or more phonemes until the pronunciation matches a word already stored in oral language.

In one frequently cited example, a child might decode wasp as though it rhymes with clasp. The child must recognize that the attempted pronunciation is close to a word already known and then adjust the vowel.

This is described as phonological flexibility.

But let’s be honest about what the child is being asked to do.

The child has to guess a pronunciation and check it.

If it does not work, the child has to guess again.

To do this successfully, the child must hold the letters, the first attempted pronunciation, and several possible alternative phonemes in working memory. The child must remember what has already been tried, suppress the first response, change one portion of the pronunciation without losing the rest of the sequence, blend the new combination, and compare it with known spoken words.

This is a complicated task even for children who can already segment and blend, have strong working memory, possess large oral vocabularies, and can flexibly manipulate phonemes.

Now consider the children who cannot segment or blend reliably in the first place.

We are asking them to use the abilities that are most difficult for them as the route into an unfamiliar written word.

The child has to know the word already

Guess and check also depends on vocabulary.

The child has to know the word well enough to recognize it when one of the attempted pronunciations comes close. A child cannot determine that a pronunciation “sounds right” when the word is not part of the child’s language.

No amount of phonological flexibility can recover a word the child has never heard, does not understand, or cannot retrieve.

This means the strategy works best for children who already possess much of the knowledge required to succeed. They know the word. They can retrieve it. They can hold possible pronunciations in mind. They can manipulate the phonemes. They have enough experience with English spelling to narrow the possibilities.

When the procedure works, we credit the strategy.

But the child may be succeeding because of all the language, memory, and phonological abilities the strategy assumes.

Trying more possible pronunciations is not the same thing as understanding the word.

Guess and check may work sometimes. Guessing works sometimes in mathematics too. A student can try a possible answer, check it, and eventually arrive at the correct one.

We would not make that the foundation of mathematics instruction.

It should not be the foundation of reading and spelling instruction either.

We have been asking the wrong question

The question “What should a child do when they encounter an unfamiliar word?” begins too late.

Before asking children to produce a pronunciation, we need to ask what kind of system they are looking at.

These questions do not provide children with another strategy for getting past one word. They help children understand how the writing system works.

The goal should not be to replace contextual guessing with phonological guessing.

The goal should be to remove the need for guessing.

Phonology is the most abstract place to begin

Phonology is important. But phonology does not operate independently of morphology, meaning, or the history of the word.

It is also the most abstract part of the writing system.

Speech disappears as soon as it is produced. Phonemes do not arrive as separate beads on a string. Speech sounds overlap. They influence one another. They change with stress, dialect, surrounding sounds, and a word’s morphological structure.

People speak with phonemes throughout their lives without consciously isolating them.

A phoneme is an abstraction from speech.

Yet literacy instruction commonly takes this least visible and least stable source of information and makes it the entry point to the entire writing system. Children are expected to segment words into abstract phonemes, connect those phonemes to graphemes, blend them back together, and then revise the pronunciation when the result does not match a word they know.

For many struggling learners, we have chosen the hardest possible starting place.

Then, when they cannot do it, we give them more of it.

Meaning gives us a more accessible place to begin. Children know and use meaningful words before they learn to read and spell them. Written words remain visible. We can examine them, compare them, write them, take them apart, rebuild them, and return to them.

Written morphemes allow children to see structure.

Related words provide evidence.

Phonology can then be investigated within a meaningful structure that the child can see and understand.

The spelling preserves more than one pronunciation

Consider these historically related words:

finite

infinite

infinity

final

They share a history connected with an end, boundary, or limit. Their spellings preserve that relationship, but the pronunciation of the first <i> is not stable.

We hear one pronunciation in finite and final and another in infinite and infinity.

If spelling were designed to record the surface pronunciation of each word as directly as possible, we might expect the spelling to change whenever the pronunciation changes.

It does not.

The spelling preserves the relationship among the words.

Now consider:

wild

wilderness

The pronunciation of the <i> changes, but the spelling of wild remains visible in wilderness.

We see a similar relationship in:

malign and malignant

benign and benignant

The pronunciation shifts across the related words. The spelling continues to identify the relationship.

These spellings are not failures of a phonetic code.

They are evidence that English spelling is doing more than representing the pronunciation of one word at one moment.

When we begin with the pronunciation of an isolated word, these spellings appear irregular.

When we investigate the family, the organization becomes visible.

Morphology cannot be added as an afterthought

Morphology is now appearing everywhere.

Publishers are adding prefixes, suffixes, bases, matrices, and word sums to phonics programs. Professional organizations are beginning to acknowledge that English orthography is morphophonemic. Teachers are hearing more about morphological and etymological word families and meaningful structure.

Some of this represents genuine progress.

Some of it amounts to adding the appearance of morphology without changing the underlying phonics-first model.

A program does not change its understanding of the writing system merely because it adds a morphology lesson after several years of phonics.

If children are first taught that words are strings of sounds represented by letters, morphology will continue to be treated as an additional layer placed on top of the supposed phonological foundation.

Meaning comes later.

Structure comes later.

Etymology comes later, if it comes at all.

The original architecture remains intact.

English is not a phonological code with morphology attached to it. Written words are meaningful structures from the beginning. Morphology, etymology, and phonology operate together in the words themselves.

We do not need to preserve the old foundation and add more components.

We need a more accurate starting point.

Children need guided inquiry

Changing our understanding of the writing system also changes instruction.

The answer is not for an adult to stand in front of children and deliver linguistic explanations for them to memorize.

It is also not to leave children alone to reinvent the English writing system without support.

Children need guided inquiry.

The teacher needs enough knowledge of the writing system to choose productive words, pose useful questions, supply information, notice misconceptions, and direct attention toward relationships the child may not yet see.

The child participates in the thinking.

A teacher might place finite, infinite, infinity, and final together and begin with meaning.

What do these words have in common?

Where have you heard them?

What do finite and infinite mean?

What changes when we add <in->?

Which letters remain consistent across the words?

What happens to the pronunciation?

What information might the spelling be preserving?

The teacher does not have to announce the conclusion before the investigation begins. The children examine the evidence, form ideas, test those ideas, and revise them.

They are not merely learning an answer.

They are learning how to find one.

Children develop critical thinking by thinking. They learn to make claims based on evidence because we give them opportunities to examine evidence. They learn that an idea can be revised because they are allowed to test ideas. They learn how language works because they are actively investigating language.

Guided inquiry does not mean that the teacher withholds information. There are facts children cannot reasonably discover alone. The teacher carries the responsibility for knowing when to ask, when to tell, when to demonstrate, and when to let the child continue thinking.

The teacher knows more than the child, but the teacher does not have to do all the thinking for the child.

Giving a child a cue is not the same as teaching guessing

I want to be clear about something.

I give children cues while they are reading.

I give them whatever help they need to remain successful, stay engaged with the text, and avoid the embarrassment of repeatedly producing wrong answers in front of me.

Sometimes I simply tell them the word.

Sometimes I provide a phonemic cue.

Sometimes I give a semantic cue. I may turn the sentence into a fill-in-the-blank question that helps the child retrieve the word from meaning while looking directly at its spelling on the page.

That is okay.

It is not a crime to help a child say a word.

The child can connect the known meaning and pronunciation with the visible spelling without being required to work out every part of the phonology at that moment. We do not have to stop the reading, increase the pressure, and turn every unfamiliar word into an impromptu decoding test.

That is not the same thing as teaching children to identify unfamiliar words by guessing from context.

The difference is the purpose.

Three-cueing treats semantic or syntactic guessing as a word-identification strategy. The child is expected to use the picture or sentence to predict what the written word might be, often without fully attending to the spelling.

That is not what I am doing.

The word is right there on the page. The child is looking at it. I am helping the child connect that spelling with a word already present in oral language so that we can continue reading successfully.

Giving a child a cue does not mean accepting a substitution that fails to account for the spelling. The child’s attention still needs to remain on the written word.

The cue is support in the moment. It is not the end of the instruction.

I also make a mental or written note that we need to return to that word.

What happens next depends on the child and the word.

We might briefly discuss its meaning and structure.

We might write a word sum.

We might compare it with one or two relatives.

We might investigate a full family using word sums and a matrix.

Then we can examine the phonology within that meaningful structure.

Not every word requires a complete investigation during the reading. Stopping to perform a full lesson every time a child hesitates would destroy comprehension and make reading exhausting. The teacher has to decide what support is needed now and what instruction should happen later.

The reading itself is not always the time to find out whether the child can identify a word without help.

Sometimes we are reading to understand a story, learn information, develop vocabulary, or experience book language together. I am not going to sacrifice all of that so I can test the child on every unfamiliar spelling.

Helping a child succeed is not cheating.

Supplying a word is not lowering expectations.

A semantic cue is not inherently harmful.

A phonemic cue is not inherently harmful.

What becomes harmful is repeatedly withholding needed support, demanding that the child guess, and treating every hesitation as an opportunity to test a skill that has not yet developed.

The cue helps us continue.

The child’s response tells me what we may need to study.

Then the instructional work happens through guided inquiry, where we can give the word the time and attention it deserves.

An unfamiliar word should not become a daily quiz

There is a human cost to guess and check.

Imagine repeatedly putting an unfamiliar word in front of a child who cannot segment, blend, or hold multiple possible pronunciations in mind.

Then we say:

“Try it.”

“Sound it out.”

“What else could that letter say?”

“Does that sound right?”

“Try another sound.”

The child does not have the knowledge or the cognitive resources needed to answer, but the questioning continues as though greater effort will produce information that has not been built.

Every unfamiliar word becomes a quiz.

If the child is not ready for the quiz, that is not a failure in the child. It tells us that our instructional work is not finished.

Repeated premature testing creates anxiety. Some children freeze. Some immediately say, “I don’t know,” even when they know something. Some put their heads down, cover their faces, or stop speaking. Others joke, argue, leave their seats, or try to escape the activity.

The response is then treated as the problem.

We should not create an instructional situation that repeatedly produces panic or shutdown and then blame children for trying to protect themselves from it.

Learning can be difficult without making children helpless.

During guided inquiry, the child has somewhere to go. The child can consider meaning, compare words, examine letters, identify structure, ask questions, and use evidence. The teacher can provide support without turning the lesson into either a lecture or a guessing game.

The child is not being asked to perform knowledge that has not yet been developed.

We finish more of the instructional work first.

Literacy instruction has to stop guessing

The phonics reform movement rejected guessing from pictures and context.

That correction was necessary.

But contextual guessing was replaced with phonological guess and check. The location of the guessing changed. The underlying demand remained: produce a word you do not yet know how to identify and keep trying until something works.

We need to begin with a different question:

What is the writing system we are asking children to understand?

If we present English as a code in which letters primarily represent sounds, we will continue producing irregular words, exception words, flexible pronunciations, heart words, syllable types, and increasingly complicated procedures for guessing and checking.

If we present English as an organized writing system that represents meaning and pronunciation through the interaction of morphology, etymology, and phonology, many of those supposed irregularities become evidence.

For years, those of us using Structured Word Inquiry argued that written syllable types were not the organizing structure of English spelling. We argued that morphology gave students access to something written syllable division could not: meaning, structure, word families, and an explanation for why spellings remain stable even when pronunciation changes.

Those arguments were not always polite.

People defended written syllable types at length. Some made remarkably condescending remarks to those of us who were using morphology instead. We were told that students needed to identify open syllables, closed syllables, vowel-consonant-e syllables, r-controlled syllables, consonant-le syllables, and vowel-team syllables before they could manage longer words.

Some insisted that morphology should come later.

Others said they would teach both.

Now, increasingly, I see many of those same people moving away from written syllable types and toward morphology.

Some now speak as though morphology had always been central to their work.

The list is growing.

My first response is: good. I’m genuinely glad to see this change.

Welcome aboard. I am glad you finally got the memo.

I am genuinely glad people are changing course. I am simply not going to pretend the last eight years didn’t happen.

People working in Structured Word Inquiry have been explaining this for well over a decade. I have been actively talking about it with others for approximately eight years. The point was never to win an argument. The point was to fight for the children traditional phonics instruction did not reach.

Those are the students I work with.

They are often students with dyslexia, receptive and expressive language disorders, difficulty with sequential processing, reduced working-memory capacity, or long histories of intervention that did not stick. Many have already spent years identifying syllable types, marking vowels, drawing division lines, reciting rules, and attempting to pronounce isolated chunks that bear little resemblance to the words they already know.

Written syllable types have been deeply embedded in the phonics world for a very long time. In some circles, they still are. But they are slowly losing their status as unquestioned instructional truth.

That shift deserves attention.

What written syllable types ask students to do

Consider what happens when a student encounters a longer word through a written syllable-type procedure.

The student may be asked to:

• locate the vowels

• identify the consonant pattern

• decide where to divide the word

• label each written chunk by syllable type

• remember the pronunciation associated with each type

• pronounce the chunks separately

• hold those pronunciations in working memory

• blend them back together

• compare the result with a known spoken word

• decide whether the result sounds right

That is not a simple reading task.

It is a demanding sequence of rule retrieval, inhibition, auditory memory, self-monitoring, and metalinguistic decision-making.

For students with language and working-memory difficulties, the procedure itself can become the obstacle.

The student has to suppress the way the word is naturally pronounced, follow an artificial sequence, retain each part, and then reconstruct the whole. Cognitive resources are spent managing the procedure instead of accessing meaning.

This becomes especially problematic because English speech does not preserve clear, fully pronounced vowels in every syllable. Unstressed vowels are frequently reduced. Syllable boundaries are often difficult to locate precisely, and spoken syllables do not always align cleanly with written structure.

The issue is not that syllables do not exist. The issue is that spoken syllables and written word structure are not the same thing. In English, syllables can often be counted, but their boundaries are not always clear or stable. Written syllable division can also cut across meaningful word structure, separating bases from affixes and obscuring the relationships that explain the spelling.

A familiar example

Take the word family.

A written syllable procedure may present it as:

fam · i · ly

The student may then be expected to pronounce each written part distinctly.

But many English speakers ordinarily pronounce family with two spoken beats, closer to fam-ly. The medial vowel may be heavily reduced or disappear in casual speech.

The student already knows the word.

The problem is not a lack of oral vocabulary.

The problem is that the instructional procedure asks the student to pronounce the spelling in a way that does not match ordinary speech.

A teacher may then say, “Now blend it.”

The student tries again, this time holding three artificial chunks in working memory while searching for a familiar word.

From the outside, this may look like a weakness in blending.

It may actually be the predictable result of cognitive overload.

What morphology changes

Morphology begins somewhere else.

Instead of asking, “Where can we divide this word into pronounceable written chunks?” we ask:

What does the word mean?

How is it built?

What other words are related?

The spelling of family becomes more understandable when it is connected to familiar and familiarity. The written form preserves a relationship across a family of words, even when pronunciation shifts.

That is a more useful explanation than telling a student to pronounce a vowel clearly in a position where English speakers routinely reduce it.

Morphological instruction also preserves the meaningful structure of words.

Consider intensive.

A syllable-based division may produce:

in · ten · sive

But the meaningful structure is:

in + tense + ive

The syllable chunks themselves do not explain the word.

The morphemes do.

One procedure directs the student toward temporary sound units. The other directs the student toward meaning, structure, and related words.

The cost of teaching both

For a time, one common response to morphology was: “I teach both.”

That sounds balanced.

For many students, it simply means more procedures.

The student is asked to identify written syllable types while also identifying prefixes, bases, and suffixes. When the two systems divide the word differently, the learner has to decide which structure matters for which task.

For students with strong language, memory, and cognitive flexibility, that may be manageable.

For many of the students I serve, it is a nightmare.

They do not need two competing maps of the same word. They need one coherent account of how the word is built.

When instruction already strains receptive language, expressive language, working memory, and sequential processing, adding another analytical system does not make the instruction more comprehensive.

It makes it heavier.

What helps instead

What helps is not more rehearsal of a complicated procedure.

What helps is reducing the number of arbitrary decisions a student has to make.

Instruction can begin with a known word and its meaning. It can identify meaningful structure, connect the word to relatives, examine affixes, preserve natural pronunciation, discuss stress and vowel reduction honestly, and study grapheme-phoneme correspondences within the meaningful structure of the word.

Structured Word Inquiry explains how English works.

Applied SWI™ is the instructional framework I have developed to translate those principles into meaningful intervention for students with complex language and literacy needs.

Moving away from written syllable types is only one example of what that looks like in practice. Applied SWI™ organizes intervention around meaningful word structure rather than written syllable categories. It also creates room for the broader language work that students need: discussion, collocations, sentence writing, structured reinforcement, and repeated opportunities to use words meaningfully.

The point is not to hand a student a matrix and expect understanding to appear.

The point is to take the linguistic science seriously enough to know how to make it usable.

A teacher moment

A student stalls over a longer word.

The teacher could ask the student to locate the vowels, mark the consonants, divide the word, name the syllable types, pronounce each chunk, and blend the result.

Instead, she asks, “Do you know this word when you hear it?”

The student does.

They talk about what it means. They identify the base. They notice a related word. They examine the affix and then look at how the graphemes represent the pronunciation in that word.

The student has not avoided phonology.

The student has encountered phonology in its actual linguistic setting: inside a meaningful morpheme, inside a real word, and inside a family of related words.

The burden shifts from memorizing a procedure to understanding a system.

The last thing they are holding onto

Written syllable types are gradually being abandoned.

Morphology is becoming more visible.

That is progress.

The final assumption many people still cling to is that instruction must begin with isolated grapheme-phoneme correspondences and only later move toward meaning and structure.

That may be the next idea to fall.

I wonder how long it will be before more people stop beginning with isolated correspondences and start with the word as a meaningful linguistic unit.

I also wonder how many will then say they were doing that all along.

I do not regret speaking up, even when those conversations became uncomfortable. If the result is that more children receive instruction grounded in meaningful word structure, those years of debate were worth it.

The history matters because educational ideas do not change harmlessly. Children spend years receiving instruction while the field slowly catches up.

For the children who succeed under almost any approach, these arguments may seem theoretical.

For the children I serve, they are not theoretical at all.

They determine whether intervention adds another procedure to memorize or finally gives the student a way to understand how English works.

Ideas change.

I have watched that happen over the last eight years.

My hope is that the next shift doesn’t take another eight.

Children shouldn’t have to wait that long.

Applied SWI™ is a trademark identifying my instructional framework for translating the linguistic principles of Structured Word Inquiry into meaningful intervention for students with complex language and literacy needs.

Orthographic mapping has become one of the most frequently discussed ideas in literacy, and there is value in helping teachers understand how written words become familiar enough to recognize quickly. Linnea Ehri’s work has been central to that discussion. Her account describes how spellings become connected in memory with pronunciations and meanings. More recently, morphology has also been given a larger place in that explanation.

The trouble begins when this kind of psychological model is treated as though it were also a description of written English. It is not. A model of what a learner connects, stores, or retrieves does not tell us how the written word itself is constructed. Those are different questions, even though literacy education has often blurred them.

Many educators have seen diagrams linking orthography, phonology, and semantics. Some newer versions include morphology. The addition of morphology may make the model more complete as an account of the information involved in recognizing and remembering words, but it does not turn the diagram into a model of the writing system. The labels still represent kinds of knowledge or mental representation. The connections still describe what may become associated in the learner. They do not show the linguistic organization of the written word.

A model of written English would need to show how letters form graphemes, how graphemes operate within written morphemes, and how written morphemes combine within written words. It would also need to distinguish those written units from the corresponding units of speech. A triangle connecting orthography, phonology, meaning, and even morphology does none of that.

I thought they were showing me the writing system

I was confused by these models for a long time.

When I saw words such as phonemes, graphemes, and morphemes in diagrams about reading and spelling, I assumed the diagrams were showing the structure of written language. Why wouldn’t I? We were talking about learning to read and spell. I thought orthography meant the actual writing system, graphemes meant the written units inside it, and morphemes meant the meaningful written structures we could see and investigate in words.

It took me a surprisingly long time to understand that this was not necessarily what the models were describing. In many cases, the labels referred to mental representations: what the learner knows, stores, activates, or connects while recognizing a word. A grapheme in the diagram was not necessarily being presented as a unit inside a written morpheme. A morpheme was not necessarily the written structure visible in a word family. These were proposed representations and processes within the mind of the reader.

Once I understood that, I had another question. Why had I been given so many models of what might be happening inside the learner and so few models of the writing system the learner was expected to understand?

We were teaching reading and spelling. The written language was sitting directly in front of us. Yet the dominant conversation kept directing our attention inward, toward storage, mapping, processing, and retrieval. I could not understand why the structure of written English itself was not the starting point.

I did not clearly see oral and written language modeled as separate but corresponding systems until I began studying Structured Word Inquiry and encountered Peter Bowers’s writing. His parallel nested model showed me something that now seems obvious: speech and writing have related but different units. Phonemes are not graphemes. Spoken morphemes are not written morphemes. Spoken words are not written words.

That was when I finally understood why the earlier models had been so difficult for me to interpret. I had been looking at psychological models and assuming they were linguistic models of written English.

They were never intended to do the same job.

What Ehri’s account is actually explaining

Ehri’s orthographic-mapping account is concerned with how a written word becomes bonded in memory to a word the learner already knows. The learner knows the pronunciation and meaning. Through attention to the spelling and its relationship to speech, the written form becomes connected with that existing knowledge.

More recently, Ehri has acknowledged a larger role for morphology in the process. That is an important development because morphology was pushed aside for years or treated as something to introduce only after children had mastered the supposedly basic code.

Still, the model remains concerned with learning and memory. It tells us about the connections a reader forms. It does not display the internal organization of the written word itself.

We do not see letters forming graphemes, graphemes operating within written morphemes, or written morphemes combining to form words. We also do not see a clear distinction between a morpheme in speech and a morpheme in writing.

Adding morphology to the diagram does not change the kind of model it is.

Including morphology is not the same as modeling written morphology

This distinction has become especially important now that morphology is appearing in literacy models, courses, books, and professional development. Its growing prominence is welcome. For years, morphology was delayed, minimized, or treated as advanced vocabulary work. But the presence of the word morphology does not tell us how morphology is being understood.

In an account of orthographic mapping, morphology may be treated as one type of information involved in storing and recognizing a word. A learner may connect a spelling with a pronunciation, a meaning, and an awareness of meaningful parts. That is still an account of the learner’s knowledge and memory.

Written morphology is something more specific. Written morphemes are units of the writing system. They contain graphemes and combine to form written words. Their boundaries affect how the letters in a word can be analyzed. A model may acknowledge morphological knowledge without showing any of this structure.

That is why including morphology in a diagram does not establish that the diagram represents written English. It only establishes that morphology has been included among the forms of information believed to contribute to word learning.

A triangle with morphology in it is still not the writing system.

Spoken and written morphemes are not identical

The word morpheme is often used without specifying whether the discussion concerns speech or writing. That makes the explanation sound simpler, but it conceals an important distinction.

A spoken morpheme belongs to oral language. A written morpheme belongs to written language. The two correspond, but they are not the same object and do not always have the same form.

Consider the plural suffix. In cats, the written suffix <s> is pronounced /s/. In dogs, the same written suffix is pronounced /z/. In wishes, the plural is written <es>, and the spoken form includes a vowel. The grammatical meaning remains plural, while the spoken and written forms vary.

Word families make the distinction even clearer. The written base <heal> is visible in both heal and health:

<heal>

<heal + th → health>

The pronunciation changes, but the written relationship remains visible. The spelling preserves the connection between the words even though speech does not preserve a single pronunciation of the base.

When spelling is approached mainly as a representation of pronunciation, examples such as heal and health appear troublesome. When written words are examined through their morphological structure, the relationship is straightforward. The writing system is preserving meaning and family connection across a change in pronunciation.

A model of memory may tell us that readers connect these forms. It does not necessarily show why the written relationship exists in the first place.

What Peter Bowers’s model makes visible

Peter Bowers’s parallel nested model addresses a different question from Ehri’s orthographic-mapping account. It shows the organization of oral and written language and keeps their units distinct.

On the oral side, phonemes operate within spoken morphemes, and spoken morphemes combine within spoken words. On the written side, graphemes operate within written morphemes, and written morphemes combine within written words. Letters and phones can also be shown at the level below graphemes and phonemes.

The model does not deny the relationship between speech and writing. It makes the correspondence visible without treating one system as though it were simply a version of the other. Phonemes correspond with graphemes, spoken morphemes correspond with written morphemes, and spoken words correspond with written words. Correspondence does not make the units identical.

That may sound elementary, but literacy instruction frequently collapses the distinction. Children are taught to begin with a spoken word, divide it into phonemes, and assign spellings to those sounds. Morphology and meaning are often introduced only after the word has already been analyzed through speech.

Bowers’s model shows why that sequence can be inadequate. Written words already have morphological structure before the learner begins sounding them out. Graphemes operate inside written morphemes, and they cannot cross morpheme boundaries. The structure of the word therefore has a direct bearing on the graphemic analysis.

Consider reacted:

<re + act + ed>

The word is not simply a continuous series of letters waiting to be divided according to pronunciation. It contains three written morphemes. Those morphemes are present from the beginning and help determine how the word should be investigated.

A morphology-inclusive account of orthographic mapping. Original explanatory schematic based on Linnea Ehri’s account of how spellings become connected with pronunciations and meanings in memory. Not a reproduction of a published figure.

Figure 2. Peter Bowers’s parallel nested model of oral and written language. Reproduced from Bowers and Bowers (2018).

Placed together, the difference is much easier to see. One model is concerned with what the learner stores and connects. The other shows the structure of the language systems the learner is trying to understand.

Both questions are worth asking. They simply require different models.

The writing system had been described, but most teachers were not shown it

None of this means that the structure of English orthography was unknown before Peter Bowers. Linguists, orthographers, spelling researchers, and some educators had examined the organization of written English for decades. Richard Venezky’s The Structure of English Orthography, published in 1970, is an important example. There was also substantial work on spelling development, morphology, word relationships, and the ways English spelling represents more than pronunciation.

Some of that knowledge entered education through developmental spelling, word study, and morphological instruction. It did not, however, become the central picture most teachers were given when they were taught how reading works.

The models that dominated literacy training were usually models of the learner. They described phonological processing, lexical storage, recognition, memory, and the connections formed among spellings, pronunciations, and meanings. Teachers could receive extensive explanations of what was believed to be happening in the mind without receiving an equally clear account of how written words themselves were structured.

Peter Bowers did not invent graphemes, written morphemes, or the organization of English spelling. What his work did for me, and has done for many educators, was bring that linguistic structure into literacy instruction in a form that could actually be seen and used. The parallel nested model places oral and written language beside one another. Word sums, matrices, and morphological families then give teachers and learners ways to investigate that structure directly.

The writing system had been described for years. I simply had not been shown it.

Why the distinction matters in classrooms

This is not simply an argument about diagrams. The model a teacher receives affects what that teacher notices in a word.

When orthography is treated as a visual form connected with pronunciation and meaning, its internal organization can disappear. Graphemes are then taught as though they can be identified across an uninterrupted string of letters. Morphology becomes something to add later, often as vocabulary enrichment or advanced word study.

When the written word is treated as a structured linguistic object, the investigation begins differently. Meaning and morphology help identify the written units. Graphemes are considered within morphemes. Changes in pronunciation across a family are no longer automatic evidence of irregularity. The spelling can be examined for the information it preserves.

English writing represents phonology, but it also represents morphology, meaning, and historical relationships. A psychological model of word learning may include all of those forms of information. It still does not show how they are organized within the writing system.

Orthographic mapping can help explain how a written word becomes familiar to a learner. It cannot, by itself, explain what the written word is made of.

That is the distinction teachers need to see. A model of the learner is not a model of written English, even when morphology has been added to the diagram.

That phrase turns the conversation away from whether teachers are doing enough phonics or too much phonics, and toward a more useful question: Is the instruction actually aligned to the writing system children are trying to learn?

Much of reading instruction is aligned to only one layer of the system. It treats spelling as if its main job is to represent pronunciation. Once that assumption is in place, English has no choice but to look messy. Words like does, sign, been, two, and your get labeled irregular because they do not behave well inside a sound-first model.

But English spelling is not only representing pronunciation. It is representing pronunciation and meaning. Once we understand that, many of the spellings that look irregular from a sound-first view become understandable when we examine the word’s structure, relatives, and history.

This is where the usual language of “deep” and “shallow” orthographies becomes a problem. Those terms describe the consistency of spelling-pronunciation correspondences. They do not describe the whole writing system. They measure the part of orthography that has to do with pronunciation while leaving out the part that has to do with meaning.

English spelling is not a failed shallow orthography. English spelling is a system that preserves meaning relationships across related words, even when pronunciation shifts. That is the language people need to hear because it directly challenges one of the most damaging assumptions in reading instruction.

Children are often taught that English spelling is full of exceptions, rule-breakers, and words that simply have to be memorized. When students are told again and again that the words they struggle to read and spell are irregular, we are not giving them a system. We are giving them a reason to stop looking for one.

Pete’s chapter helps us see the problem more precisely. If orthography is evaluated only by spelling-pronunciation consistency, then English will appear unusually difficult. But that analysis leaves out spelling-meaning consistency, which is central to how English spelling works.

The word sign is a good example. If we look only at the pronunciation of sign, the <g> appears to be a problem. But when we place sign with signal, signature, design, and designate, the spelling is no longer random. The <g> is preserving the family relationship. It helps show that these words are connected in meaning, even though the pronunciation shifts across the family.

The same is true with do, does, and doing. If does is treated only as a word with an unexpected pronunciation, it becomes another item on a list of irregular words. But if students understand that does belongs to the same family as do and doing, the spelling has an explanation. It is preserving the written morpheme even when pronunciation changes.

The word two also becomes far less strange when it is studied with twice, twin, and twenty. The <w> is not a random silent letter. It preserves a spelling-meaning connection to related words. It also distinguishes two from the homophones to and too. Once again, the problem is not the spelling. The problem is the instructional frame that treats pronunciation as the only relationship worth noticing.

Instruction-orthography alignment gives us better language for this problem. If instruction is aligned primarily to grapheme-phoneme correspondences, it is aligned to only one layer of English orthography. It may look sensible if the writing system is being treated as a pronunciation code, but English is not only a pronunciation code.

English has many morphemes whose pronunciations shift across related words while the spelling holds the meaning relationship steady. English instruction has to account for morphology from the beginning. Morphology cannot be treated as an enrichment activity, a vocabulary add-on, or something students get after they have mastered phonics. Morphology is part of the system children are trying to learn.

This does not mean phonology disappears. It means phonology is taught in context. The pronunciation of a grapheme is understood inside the structure of an actual word, not as an isolated sound-spelling fact floating around outside meaning.

In Structured Word Inquiry, we begin with meaning, then structure, then relatives, then the jobs of the letters. That sequence reflects the reality of the writing system. The letters’ jobs make more sense after we know what the word means, how it is built, and what other words belong to the same family.

Take the word question. Many students are taught to see tion as a chunk. Some are taught to divide words into syllables in ways that cut across the actual structure of the word. But question belongs with quest, request, inquest, and conquest. Once that family is visible, the spelling of the <t> is no longer arbitrary. The pronunciation shifts, but the spelling preserves the relationship.

When instruction teaches students to divide written words in ways that violate morphemic boundaries, it does more than give them an incomplete explanation. It builds the wrong mental model of how English spelling works. Students learn to look past the structure that would have helped them understand the word.

I see this clinically all the time. The students who struggle most are often the ones who have already had years of sound-first instruction. They have practiced phonics patterns. They have blended. They have marked syllables. They have memorized lists. They have been handed more and more isolated correspondences, but the system still has not become stable for them.

Those students do not need English reduced further into disconnected pieces. They need the system put back together. They need to see how spelling, meaning, and pronunciation work together. They need to understand that a spelling can stay consistent because it is preserving a meaning relationship, even when pronunciation changes.

This is why Structured Word Inquiry is not simply “adding morphology” to phonics. SWI begins from a different understanding of the writing system. It does not ask students to memorize the places where a sound-first model breaks down. It asks students to investigate how the writing system works.

A phonics-first model tends to frame morphology as something extra. First students learn the code, then later they learn meaningful word parts. But English spelling does not operate as a simple sound code with meaning added later. Meaning is built into the system from the start.

When we teach children that spelling is mainly about sounds, and then we give them hundreds of words that do not fit that explanation, we should not be surprised when many of them lose confidence. They are not failing to notice the system. They are often noticing that the system they were taught does not explain the words in front of them.

Pete’s article also raises a serious research problem.

Are they telling us something about the writing system itself? Are they telling us something about the instruction that was used? Are they telling us something about assessments that were also aligned to the wrong layer of the system?

This is a major confound. A science of reading that studies English while under-teaching morphology is not studying the full writing system. It is studying what happens when children are taught one part of the system and then assessed as if that part explains the whole.

The same issue applies to assessment.

We are measuring the part of the system we chose to teach and test.

This is one reason morphology has to move out of the margins. English spelling cannot be understood adequately through phonics alone because English spelling was never only a phonics system. It is a morphophonemic system, and the morphology is not optional.

The matrix and word sum are powerful because they make this visible. They show students how written morphemes combine to build words. They show how a spelling can remain stable across a family. They allow students to study pronunciation shifts without losing the meaning structure that holds the family together.

That is very different from handing students isolated “irregular” words to memorize. Isolated memorization builds narrow knowledge. It may help a student remember one word for one test on one day, but it does not necessarily build a generative understanding of the system. When students learn word families, word sums, matrices, and spelling-meaning correspondences, they are learning relationships that can transfer.

Transfer is the goal of instruction. Students do not need to memorize English one word at a time. They need access to the structure that helps them make sense of words they have not been directly taught.

This is especially important for struggling readers and spellers. These are the students most likely to be given more practice with isolated correspondences and more lists of exceptions. But if their core difficulty is building stable, high-quality lexical representations, then they need stronger binding between spelling, pronunciation, and meaning. Morphology is the only layer of the system that binds all three.

The old sequence needs to be challenged. The assumption that students must first master grapheme-phoneme correspondences before morphology can be useful does not match what many of us see in practice. It also does not match the direction of the findings Pete discusses. Younger and less able students can benefit from morphological instruction, and morphological instruction has shown to have positive effects on phonological outcomes.

If morphology helps bind spelling, meaning, and pronunciation, then delaying it may be especially harmful for the students who need that structure most. These are often the same students who are given the most repetitive sound-first instruction, even after that instruction has failed to give them stable access to written words.

This does not mean every child needs a graduate-level linguistics lesson. It means the adults responsible for instruction need a more accurate understanding of the system, and children need access to that system in developmentally appropriate ways. A child does not need all the terminology to investigate whether sign and signal are related. A child does not need a lecture on orthographic depth to see that two, twice, and twin share a meaningful spelling connection.

Children can investigate real words. They can build word sums. They can sort relatives. They can notice pronunciation shifts. They can learn that English spelling is ordered and meaningful.

That gives students a very different message from “this word is irregular.” It tells them there is a system worth investigating, and it gives them tools for doing that investigation.

The language Pete gives us moves us out of the tired argument about whether phonics is good or bad. That argument is too small. The real issue is whether the instruction matches the writing system.

If instruction is aligned to only one layer of English orthography, it will keep producing the same problems. It will keep calling words irregular when they are not. It will keep treating morphology as advanced when it is foundational. It will keep asking students to memorize what they could have understood.

English spelling is not broken. English instruction is often misaligned to the system it claims to teach.

If we want instruction to match English orthography, written morphology has to be there from the beginning.