Structured Word Inquiry: Using Images to Inspire Investigations

 As a huge fan of Ron Ritchart’s “Making Thinking Visible” critical thinking framework, I was wondering if there was a way to use images to inspire word investigations. I was thinking that I would just start by showing students this image without any context and use a “See/Think/Wonder” Thinking Routine.


Patrick A. Mackie, The Utah Monolith:

After completing the “See/Think/Wonder” routine, we would read and discuss the following article:

On a subsequent lesson, the word <monolith> could be selected as a word to investigate. If students are proficient at SWI, they may break into groups to investigate a word of their choosing or the teacher may decide to do a whole-class investigation.

It is helpful to do the investigation in advance but it is really important to meet the students where they are. For example, I had previously thought that mono- was a prefix but my latest thinking is that <mon> is a base with a connecting vowel <o>. Thanks to an extremely informative video in the Real Spelling Toolbox that featured word study on <mono>, I have revised my thinking. As long as students can show evidence for their thinking, it is important to respect the journey of the learner. I have been walking this path for almost three years and every day brings a new gleaning. Allow students their own gleanings and adopt a questioning stance as they conduct their own scientific word investigations.

Another interesting reason to do a matrix before instruction is that it can clarify my own understandings. For example, I became really stuck on the analysis of <monotheist> on the <mon> matrix. I analyzed <monotheist> as mon+o+the+ist as mon+o+the/+ist->*monothist because of the suffixing pattern for replacing the single, silent <e> until I realized that the final <e> in atheist is NOT silent so that the suffix is just added. To be sure, the word <monotheist> would most likely not be brainstormed by students when asked for words in the <mon> (meaning “single”) family but I felt like it was an important renewed understanding, nevertheless.

I think that connecting images from current events and word investigations holds a lot of promise and I look forward to doing some future investigations. Feel free to contact me at if you have any questions.


Patrick A. Mackie, The Utah Monolith:

Do a See/Think/Wonder Activity:

Read this “Time” article with your students:

Associated Press. (2020, November 30). Mysterious monolith disappears from Utah desert 10 days after it sparked international intrigue. Time. Retrieved from

Structured Word Inquiry (Bowers & Kirby, 2010)

Real Spelling Online Toolbox

Accessed 12/4/20

Who knew “transition”, “sedition” and “ambition” were related?

I started to write a blog post about all the ways that Structured Word Inquiry (Bowers & Kirby, 2010) supported orthographic mapping but I was reading the NYT and started to wonder about the word “transition.” So, using the 4 questions of SWI to guide my investigation, I discovered so many interesting things! First, that the <it> in <transition> is derived from the Latin ire meaning “to go.” <It> is the base of a whole family of words with a sense of “going”: transition (going from one place to another), exit (going out), obituary (what they write about you when you finally go), transitory (not going on forever), initiate, and initial (go first).

Two cousins of “transition” are “sedition” (going apart from the state) and “ambition.” “Ambition” historically meant “going around canvassing votes” and evolved to figuratively meant “seeking position or honor” (John Ayto).

Working through a Structured Word Inquiry allows for deep understandings motivated by the student’s own questions. I experienced a real sense of excitement when I realized that <it> was the base of <transition> because even though I had seen that word many times, it had not occurred to me to think about the structure of this word before. It was only when, as part of the SWI inquiry, that I was asked to explore the structure of “transition” using lexical word sums (transition->trans+it+ion), that the affixes (trans-; -ion) and bound base (<it>) became clear.

After exploring the etymology, I looked for related words using Etymology Online and the Word Searcher. Here were some words that I thought might be related, like <remit> but when I searched for the deepest root of the word, it was not the Latin ire. Only words that share the same base <it>, the same historical root and the same meaning are members of the immediate family. After some trial and error, I was able to connect this <it> base with the meaning of “go” to other words in the family that shared the same base and meaning. As part of the investigation, I realized that “sedition” and “ambition” were in the same family deriving from the historical Latin root but I could not say that they shared a base because I could not prove <amb> and <sed> were prefixes. I could find evidence of those word strings being prefixes historically (but not present-day) so I could not include them on my matrix. <Sedition> and <ambition> were more like cousins to <transition> than siblings but I did include them in my circle below to show that they are in the ire “go” extended family.

After constructing the matrix, I decided to sort the words by function: noun, adjectives, adverbs and verbs. I am investigating how stress impacts parts of speech and I wanted to explore the hypothesis that nouns, adjectives and adverbs receive the primary stress on the first syllable while verbs receive primary stress on the second syllable. It didn’t seem applicable in this investigation but these words would be interesting to share as examples of words that can be either nouns or verbs, depending on context.

The 4th question of Structured Word Inquiry is, “What are the graphemes that focus coherently here?” and I notice that all is as expected except that the <t> shifts pronunciation. It can be either /t/ or /ʃ/ when the <t> is in the medial position in words. In the final position in a word, the <t> is always <t>. It helps students to be aware of the realities of grapheme/phoneme correspondences-namely, that graphemes (letters or letter strings) can represent one or more phonemes. Knowing that <t> may represent a different phoneme when appearing in the medial position of words in this family than at the end is helpful information.

In the Structured Word Inquiry approach, students are encouraged to “spell it out” when they come to an unknown word instead of “sound it out.” This is because the first job of our (morphophonemic) spelling system is to represent meaning (as opposed to sound) and “spelling a word out” allows a student a chance to call their attention to the meaningful units in a word (bases, affixes). Not being able to immediately identify a word is a sign that the word has not been successfully mapped into long term memory so having the student “spell-it-out” instead of “sound-it out” reduces the potential embarrassment of the situation (Bowers, 2020, SWI Class.) Spelling out loud has a much lower cognitive load than “sounding it out” and I have noticed that it is a very successful strategy for instruction. If child spells out the word and doesn’t spontaneously identify the word, the teacher can ask, “Do you see a base that you know? An affix?” which a teacher can quickly point out while making a point to explore the word family and/or do a word sum for the target word at the earliest opportunity.

It is important to keep track of words a students do not identify immediately and correctly because that is a sign that the student has not mapped or bonded the graphemes, the phonemes and the meaning into long-term memory.

Investigating words using a Structured Word Inquiry approach depends on an inter-relationship of meaning, phonology and etymology. It shows students that spelling actually makes sense and allows for deep understanding of word meanings, structures and phonology. It is also a process that allows students to take ownership of their own learning. Like any inquiry process, it involves a lot of not knowing and resilience but results in such joy when understanding emerges. As Pete Bowers has frequently said, “Nothing motivates like understanding!” What word do you or your students want to understand or investigate today?

Why is “nation” pronounced differently than “national”?

One of the guiding principles of our English orthography system is that spelling remain consistent while pronunciations may shift across word families. So, when my friend, Q., asked, as we walked across the bridge formerly known as the Tappan Zee, “Why are “nation” and “national” pronounced differently?”, I knew that our spelling system is optimized for meaning not pronunciation. So, even when we have a question to explore in Structured Word Inquiry related to pronunciation (Question 4), it is always recommended to start any inquiry by discussing the meaning. (Question 1)

  1. What is the meaning?

“Nation” generally means a people within a border that are connected by a government while <national> refers to the status of belonging to a particular nation.

2. What are the etymological and morphological relatives?

After consulting Etymology Online and discovering that the historical root of <nation> and <national> is the Latin natal or nasci with the sense of “born.” “Nation” has the etymological sense of “that which has been born” or “breed” The notion of a “common ancestry” has been overtaken by the idea of <nation> as people being born within organized political boundaries. <National> emerged in the 17th century. (Dictionary of Word Origins by John Ayto)

Some morphological relatives (words that share a base and meaning) are nation, nationality, native, natural, naturalize and naturalization. All share the bound base <nate> and have a sense of “born.” If interested, a search can be made for other words in the <nate> family by using Word Searcher or Etymology Online. A word can be included if it shares the same base and meaning. Only words where students understand the meaning should be included.

3. What is the structure?

A lexical analytic word sum for <nation> and <national>:

nation -> nate/ + ion

national -> nate/ + ion+ al

A synthetic word sum for <nation> and <national>:

nate/ + ion -> nation

nate/+ ion + al -> national

The / after <nate> refers to the replacement of the <e> because of the vowel suffix.

Another way to look at structure is to create a lexical matrix which analyzes words into morphemes (bases, affixes (prefixes, suffixes). Here is a <nate> matrix from the RealSpelling Toolkit: (This is a very complete matrix. Remember, a matrix does not need to represent all possible words. For classroom use, I would only include words on a matrix that the students or I had volunteered during the class discussion, making certain that the meaning is understood.)

Now, for the fourth question:

4. How do the graphemes map to the phonemes?

In English, a phoneme (the smallest unit of sound that distinguishes one word from another) is represented by a grapheme (a written letter or letter string).


<n.a.t. + ion>

/ˈn..ʃ + ən /



<n.a.t. + ion + al>

/ˈn.æ.ʃ + ən + əl/

Now, the graphemes are represented by angle brackets <> while the phonemes are represented by the International Phonetic Alphabet within the slashes //.

We notice that the only pronunciation difference between “nation” and “national” is the initial vowel sound. In <nation> the // phoneme is represented by the grapheme <a> and in <national>, the /æ/ phoneme (short a) is represented by the grapheme <a>. We could represent this fact to children as follows (although this is not a complete list of the phonemes represented by <a>:

We know that the spelling will remain consistent throughout a word family and we see that consistency in the spelling of “nation” and “national.” This emphasizes the concept that the primary job of spelling is to represent meaning not pronunciation. We can see and hear that while the spelling remains consistent between these two words, the vowel pronunciation shifts from long to short. Why? I am not really sure but perhaps it was for ease of pronunciation as it seems more of a tongue twister to say <national> with a short sound than a long sound.

It is also noted that both words retain the primary stress on the first syllable. It is interesting to know that nouns, verbs and adjectives generally have the stress on the first syllable while verbs have the primary stress on the second syllable. (Vowels are usually reduced and represented by the schwa sound in unstressed syllables.)

See (or hear) for yourself with some words in the <nate> family:

So, at this point in the inquiry, I am not absolutely sure why <nation> and <national> are pronounced differently but I have made a reasonable hypothesis (ease of vocalization) and that will stand as I await further information.

What do you think? Do you have another hypothesis about why <nation> and <national> are pronounced differently? Thank you so much to Q. for this interesting question. I can’t wait to go for another bridge walk and talk about words!

SWI Inquiry: Spelling “Hear” vs “Heard”

My friend, Q., had a question, “Why is “heard” spelled that way?” Hmm. I don’t know why! What an interesting question, Q. !!! What we do know is that “heard” is the past of “hear” and we know that it is not spelled hear + ed-> *heared.

What is the reasoning behind the spelling of <heard>?

First, let’s look at both words:



We note that although <hear> appears in both words, there is a stem vowel shift as we say the words aloud: “hear”/”heard” although they both contain the letter strings <hear>; the words are pronounced differently.

Might this signal that these words come from Old English? I know from my class with Rebecca Loveless that there are Old English-derived strong (showing an action or state of being or feeling) verbs that mark the past tense with a stem vowel shift, such as “eat/ate”; “run/ran”; “sing/sang” and “”grow/grew.”

I notice that, unlike the words above, “hear” and “heard” have a vowel shift but not a letter string or grapheme shift. Hmmm. There is another example of am Old English-derived strong verb that has a vowel sound shift without a letter string or grapheme shift:

Present tense: “read”/Past tense: “read”

So, I ponder what, if anything does the “hear/heard” and “read/read” pairings have in common? They are all homophones: read/reed; read/red; hear/here; heard/herd. So, it would have been important for the scribes long ago to follow the homophone principle: “where possible, words that are pronounced the same, but are unrelated, will have different spellings to mark the differences in meaning.”

So, it would have been important to have graphemes in <heard> that would signal the meaningful connection to <hear> while differentiating <heard> from <herd>.

Now, with those thoughts in mind, I will investigate <heard> and <hear> using the 4 questions of Structured Word Inquiry (Bowers & Kirby, 2010) to determine if any other new understandings come to light:

So, after considering all my current understandings, my answer to Q.’s question, “Why is “heard” spelled that way?”, my hypothesis is that <heard>, as an Old English-derived strong verb, is spelled that way in order to show the meaningful connection to <hear> while differentiating <heard> from <herd>. I will continue to refine my hypothesis as I learn. Thank you, Q., for a very thought-provoking question!

Update: I wasn’t feeling confident about my analysis about why <heard> is spelled that way so I continued to do research. I came across Dr. Peter Bowers’ wonderful exploration of <say> and <said> here. His analysis taught me something new: <-d> is a non-productive suffix ( a suffix that is not currently being used to generate new words). In his etymological investigation, <saith> was historically replaced by <say> and he is aware of the OE suffix -th (-th, according to Gina Cooke, is a verbal inflection that has been replaced by the <-s> in says. It has an <-eth> variant, like in <prayeth>) so

saith-> sai + th

If <sai> is an earlier version of <say> and <-d> is an earlier, now non-productive, suffix then

sai + d-> said

Dr. Bowers concludes, “our word sum analysis gives us evidence that <say> and <said> do NOT share a base, but the etymological evidence shows that they derived from the same Old English root secgan. “

That makes me update my hypothesis that <heard> is spelled that way as a result of <hear> + the now unproductive -d suffix hence

hear + d->heard

Based on my updated hypothesis and my current understanding, our word sum analysis gives us evidence that <hear> and <heard> do NOT share a base, but the etymological evidence shows that they derived from the same Old English root heran.

Thank you, Dr. Bowers, for your illuminative analysis of <said> and for all your scholarship related to spelling and Structured Word Inquiry.

Note from Real Spellers Toolkit (Theme 2D):

  • We know that the homophones < hear > and < here > have different meanings because they are spelled differently. The one that is to do with listening shares a spelling with < ear > because that’s what we listen with.

The other word is the one to do with place: < here >. It shares a spelling with the other words that are clearly about place: < there > and < where >, even though the pronunciation is different.


Peter Bowers, Structured Word Inquiry

Rebecca Loveless

Etymology Online

The Noun Project

Neil Ramsden’s Mini-Matrix Maker

Real Spellers Forum and Toolbox

Gina Cooke and Lex Exchange

Oxford Dictionary

Embedding “Structured Word Inquiry” Within a Reading Lesson

After a text has been presented and discussed, ask students to select a word that interests them from a text.  Guide students through the following 4 questions using the Structured Word Inquiry (Bowers & Kirby, 2010) Framework.  (Note: This does not have to be explored during one sitting.)

  1. Meaning?
  2. Relatives? (Etymology)
  3. Structure?  (Morphology)
  4. Pronunciation? (Grapheme/Phoneme Correspondence)

Sample Text:

“Sorry, But It Was Beautiful” by Andrew Vecchione (6th Grade Student of Kenneth Koch)

This is just to say

Sorry I took your money

And burned it.

But it looked like the world falling apart 

When it crackled and burned.

So I think it was worth it.

After all

You can’t see the world fall apart every day.

See the following SWI Thinksheet for a guided exploration of the student-selected word, “crackle“:

SWI ThinkSheet:

One of the notable understandings for this investigation is that the suffix <-le>” is represented by the grapheme<l>marker <e> and the phonemes /ə/ and /l/.

After the exploration, ask students for their lingering questions or provide extension activities for further study. In this case, an exploration of the -le suffix in the Oxford Dictionary showed that one meaning of this suffix is “a repetitive action.” As one collects words with the -le suffix, it was noted that many of those words refer to repetitive sounds. This resulted in a sort of the words with an -le suffix (or ending, as we determined some of the words had <le> as part of the base, not as part of the suffix) into two groups: -le words that referred to repetitive sounds and those that did not.

As a result of this investigation, there was heightened understanding of the -le suffix and future investigations to explore the -le suffix in English as it seems to have evolved (Latin -el?) or be connected to the -el suffix.

From Etymology Online, entry for -el:

-el (3)

derivational suffix, also -le, used mostly with verbs but originally also with nouns, “often denoting diminutive, repetitive, or intensive actions or events” [The Middle English Compendium], from Old English. Compare brastlian alongside berstan (see burst); nestlian (see nestle) alongside nistan). It is likely also in wrestletrampledragglestruggletwinkle, also noddle “to make frequent nods” (1733). New formations in Middle English might be native formations (jostle from joust) with this or borrowings from Dutch.

I am wondering, for example, if a word sum for <prattle> (from the Middle Dutch praten, “to chatter“) which would be expected to be prattle-> prat+le-> *pratle except that is not correct. The “Etymology Online” site notes in the entry for -el that in Modern English the -el suffix is represented by -el. Note that prat(t)+el->*prattel would explain the doubling but is not the correct spelling. At this point I will just consider <prattle> a base based on my current understanding but I will research the -el/-le connection. (As I feel the pronunciation of the -le suffix, I realize that the first phoneme I hear is the schwa, which is a vowel. Is there any doubling rule that takes into account a suffix that starts with a vowel sound even though the suffix does not start with a vowel?) (Update: After examining the etymology of the words below here, I became aware of a class of verbs called frequentatives with in English marks its verbs with an -le or -er ending and refers to words that show some sort of small or intense repeated action. Read here and here and here. My latest understanding says that for a verb to be frequentative, it must be derived from another word with the addition of the suffix. That doesn’t explain why a word like <prattle> which is most likely frequentative and arising from <prate> does not follow the rules for doubling. More to ponder and explore!)(Update #2: I just saw this on Gina Cooke’s site:

“Tell me again about the “six syllable rules.” Do you mean like how you have children “count back 3” for words like table, ruffle, and the like? So instead of showing children the FACT that the ‘le’ is often a suffix — spark+le, hand+le, circ+le (compare circ+us) — but not always. Sometimes it’s a vestigial suffix, something I’ve been known to call a ‘footprint’ with my students. The ‘le’ in bumble and gamble and spindle can no longer be analyzed, but we can still see how they were historically built from boom + le and game + le and spin + le.

What’s really interesting about an ‘le’ suffix is that it functions as a vowel suffix, because that ‘l’ is syllabic: mid + le, side + le, lade + le (compare laden or lading), set + le. Mind blowing, isn’t it? And 2nd graders can totally get that. It’s adults that struggle with it.”

Based on my new understanding of <le> as a vestigial suffix that functions as a vowel suffix, we can use that information to understand how many of the words were historically built.

Embedding Structured Word Inquiry within a reading lesson offers many opportunities for generative learning. Using the basic framework of Structured Word Inquiry is appropriate for whole class instruction with follow-up activities that can be designed to address individual needs. After a SWI lesson, some students may use a matrix to write word sums while another student may illustrate the meaning of words in the word family (crackle, firecracker, cracked) while still another may do further research on words with -le in the suffix or the base.

Most of all, Structured Word Inquiry allows for a joyous exploration of understanding how our spelling system actually works. An educator does not have to have perfect understanding of the spelling system to embark on the journey, just an openness to learning as you (and your students) go along. Let your questions guide your learning within a true inquiry experience. I encourage you to conduct your own inquiries as part of your reading lessons. Feel free to email me at if you have any questions.

Update 12/5/20: After a careful study of the Real Spelling Toolbox (Kit 4; Theme J), my latest understanding is that there is NOT an <le> suffix, although it may seem to act as a suffix in some words like <crackle>. There was a historic -el suffix and we see the vestiges in word but in present-day English, there is too much variability for it to be considered a suffix. Real Spelling describes the present day final syllabic <le> as an “element-forming ‘particle’ that has combined with a base or root at some point in history to form a new and separate Modern English base. One of its commoner functions was that of a ‘frequentative verb-forming particle.’ It formed words denoting continuous or repeated action. Sometimes we can recognize the stem (wrestle, sparkle, trample, dazzle). Often, though, with these ‘frequentative verbs, the stem has no relatives that we now know of (giggle, fiddle, ogle, dawdle, bubble).” I am so very grateful for the scholarship of Real Spelling because the Real Spelling description continues to describe the historical connections that impact our present-day words. At this point, it makes sense to me to not consider <le> a suffix so that means I should redo my matrix. I am also reconsidering if the phoneme for <le> should be represented as /l/ or /(ə)l/ or what Michel of Real Spelling calls the “dark” <l>. This will be a question for my “Fun with Phonetics” instructor, Patti Bravo-Bottino!

My updated matrix:

For further information:

Dr. Peter Bowers:

Mary Beth Stevens:

Rebecca Loveless:

Gina Cooke: LEX

Noted from Real Spellers Toolbox:

-Screenshot from Real Spellers Toolbox (Theme 1D, “Final <e> that is not single” <coffee>)

Updated Theory of Reading

I used to think that Reading and Spelling was largely based on visual memory. Yes, you needed to know the graphemes and phonemes but the goal was to synthesize that information quickly to aid quick visual word identification or spelling.

Now I think (after my research on the theories of Dr. Peter Bowers and Dr. Linnea and the work of Dr. David Kilpatrick) over these past three years that Reading/Spelling is a morphophonemic process. When students encounter a word, they analyze the spellings and map the graphemes to the phonemes and to the meaning and retrieve this representation from memory. It can take from 1-20+ exposures for this representation to make it into long-term memory but, once there, the word is immediately accessed.

Thus, the 4 questions of Structured Word Inquiry (Bowers, 2010) can be analyzed through the lens of orthographic mapping (Linnea Ehri), which is the process by which sight words (any word, not limited to our previous understanding of irregular words), are immediately accessed.)

The 4 Questions of Structured Word Inquiry:

  1. What is the meaning? (Meaning)
  2. What are the relatives? (Etymological information adds a layer of meaning as it can explain why a word is spelled as it is while morphological information helps students to focus on morphemes (base, affix), the meaningful units in a word.)
  3. What is the structure? (analyze the spelling using word sums and/or a word matrix)
  4. Pronunciation? (map the graphemes to the phonemes, articulate the sounds, note pronunciation shifts across the word family, articulation activities, spelling-it-out, and writing-it-out to consolidate understandings. Chart new phoneme and grapheme understandings as references.)

I might add a fifth question, “Can you use this word in a meaningful way?” as a way to monitor and assess developing word understandings. I would also be interested to map exposures, as possible, to begin to understand how many exposures is optimal for each of my students, as well as continuous informal assessments of comprehension, decoding and encoding(spelling) (with grapheme/phoneme analysis).

Here is a recent SWI thinking sheet that I created for <judge>:

As I was working on this, I had some questions about <judge> and <judicial> which was resolved with the help of some friends:

I have seen Structured Word Inquiry have such a positive impact on the word learning and spelling of diverse learners. It has been exciting to see the research on orthographic mapping confirm this. Looking forward to more explorations on the science and the magic of word learning!

Explaining the “Why?” of Spelling

I just listened to a Reading League podcast with Dr. David Kilpatrick. Dr. Kilpatrick has helped translate Dr. Linnea Ehri’s excellent research on orthographic mapping, which is the process of automatic word recognition, into classroom practice. He has written two great books, Essentials of Assessing, Preventing, and Overcoming Reading Difficulties, and Equipped for Reading Success.

As I listened to Dr. Kilpatrick’s explanation on how he would explain the spelling of <yacht> to a student, he used the words “tricky parts” to refer to the <ch> in the word.

As part of my learning journey to uncover how our spelling system really works through Structured Word Inquiry (Bowers, 2010), I once again reflected on how closely SWI connects to the research on orthographic mapping. Orthographic mapping research has demonstrated that for words to be recognized automatically, the graphemes (letter or letter strings that represent phonemes) must be bonded to the phonemes (unit of sound) and the meaning in memory. It may take anywhere from 1 to over 20 exposures for the word to move from having to be decoded to being instantly recognized.

Structured Word Inquiry looks at a word and asks four questions:

  1. What is the meaning?
  2. What are the relatives?
  3. What is the structure?
  4. What are the sounds that matter?

The following is an inquiry into the word <yacht>:

An investigation of the word shows that the <ch> in the word <yacht> is an etymological marker that connects the spelling to its Germanic origins. It has been my experience that being able to share the real reason behind a spelling adds another layer of meaning that can be further explored during question 4 through grapheme/phoneme matching.

This would just be the initial inquiry into the word. Exposures can be continued through word-sums, spelling-out-loud, writing-out-loud and using the word in short student-created texts.

Map Tricky Phonemes to Graphemes with a Listening Activity!

Screen Shot 2020-10-27 at 12.00.15 PM

Help students at the “Partial Alphabetic” phase (Ehri, 1995) to map phonemes to graphemes by asking them to listen for words that represent a particular phoneme like /iː/ as in <eat> ( first give them time to feel the phoneme in their mouth and look in a mirror as they articulate the sound) as you reread a picture book aloud. Words with the targeted phoneme can be listed during the rereading and then sorted afterwards by grapheme (spelling pattern) and grapheme position. Afterwards, engage the students in the “So what, Now what” Thinking Routine. Some possible outcomes:

“So what?”: “The phoneme /iː/ can be represented by a variety of graphemes: <ie><ee><ea><i> and <y>. Are there other graphemes that represent /iː/?”

Now what?: “In my reading and spelling, I can use what I learned today to decode and spell. I will add graphemes and words to the chart as I find them. The words on the chart can be used as a word bank for my writing.”

Research shows that readers at the “Partial Alphabetic” stage have not yet “glued” the phonemes to the graphemes which, along with understanding the meaning of the word, is crucial to cementing these understanding into long-term memory so words can be automatically identified.

Teachers, through informal and formal assessments of decoding and spelling, can determine what phoneme/grapheme relationships are not yet stored in long-term memory. If using a Structured Word Inquiry approach, teachers can introduce, in addition to student-led inquiry, words that contain graphemes that are not yet fully bonded to the phonemes in memory. After word inquiry, including phoneme/grapheme correspondence, a teacher could create a chart, like the one above, to explore grapheme-phoneme correspondences.

The Science of Reading Instruction: Is My Word Reading Instruction Aligned with the Latest Research?

 By Mona Voelkel, NBCT

After reading, “The Science of Learning to Read Words: A Case for Systematic Phonics Instruction” by Linnea Ehri (Reading Research Quarterly, Volume 55(1); 30 August 20): , I wanted to compare my Structured Word Inquiry instructional approach with current research. (Underpinning all my reading instruction is that decoding and encoding (spelling) are related processes and instruction should support that.)

A deep understanding of The Big 5 Pillars of Reading Instruction  (Phonemic Awareness, Morphophonemic Instruction (usually called phonics but as a Structured Word Inquiry (Bowers, 2010) practitioner, I now reject phonics as the correct term for the study of our orthographic system), Fluency, Comprehension and Vocabulary) ground my instructional practice.

Each pillar needs to be carefully assessed and addressed in any classroom or reading intervention program. 

What has been so exciting over the past few years is how our knowledge of how children learn to read has been enhanced by developments in neuroscience and reading research.  What stands out to me, however, is the tireless research of Linnea Ehri.  She rejected Ken Goodman’s theory of reading as a psych-linguistic guessing game” many years ago and has spent her career researching topics in reading, especially how students learn words.  I think her most recent hypothesis of how students learn words has really bad news for devotees of the cueing of context clues, good news for disciples of the analytic and syllabic approaches to learning words and excellent news for converts to “Structured Word Inquiry” (Bowers, 2010).  The best news of all, though, is for the children who will benefit when her latest research is applied to classroom instruction.  After reading her article, I wanted to analyze how my instructional practice aligns with her latest research.  

While Ehri agrees that words CAN be read by using context clues or through the process of decoding letters into sounds, what contributes MOST to reading and comprehension is “reading words automatically from memory by sight.”


That means that there are no sub-classification of words called sight words.  All words are sight words when read by the accomplished reader.

Key Foundational Skills Needed to Move from Pre-alphabetic to Partial and Full Alphabetic:

What are the key foundational skills?

  • Letter Knowledge (names, shapes and sounds)
  • Phonemic Segmentation
  • Grapheme/Phoneme Knowledge (Including seeing spellings of words)
  • Onset-Rimes and Syllabic Knowledge (It is important to blend syllable knowledge within SWI as needed, especially when it comes to understanding the concept of stressed and unstressed syllables and syllabic understandings such as having each syllable contain a vowel.
  • Morphemic Knowledge

I would add to this list

  • Oral Language (Students must “float in a sea of talk”(Britton, 1970) as much as is possible to be able to read, write and spell.)

Spellings, letter strings, graphemes, however you refer to the building blocks of words, “attach to pronunciations and meanings in memory and enhance memory for vocabulary words.  When a word is seen, these connections are activated in memory to read the word.”

Ehri makes the point that spelling and word identification draw from the same knowledge and skills and that “spelling instruction benefits beginner word reading and that learning to read and spell are two sides of the same coin.”

In SWI, foundational approaches like “spelling-out-loud” and “writing-out loud” and phoneme/grapheme tapping help anchor grapheme/phoneme correspondence. For more about this important SWI practice:

This is a clarion call to make sure that practitioners teach the meaning of words, word identification and spelling simultaneously with grapheme/phoneme correspondence. 

I teach word identification(decoding) and spelling using “Structured Word Inquiry” (Bowers, 2010).  It is an inquiry framework where the teacher guides the students through 4 questions:

  1.  What is the meaning of the word?
  2. What are the relatives? (Morphological and Etymological)
  3. How is this word built?  (Analyzing and synthesizing elements using lexical word sums)
  4. What grapheme function coherently here?

So far, SWI would seem to set the stage for a word learning framework by grounding students in the meaning and spelling of the word while grounding students in the grapheme/phoneme correspondences.

See Dr. Pete Bowers demonstrating the “Structured Word Inquiry Approach” that he pioneered here:


Ehri talks about “unitization” as when all the identities of a word (meaning, spelling, pronunciation) are “immediately accessed from written to spoken words when seen.”  We know from prior research about orthographic mapping that some students require just 1-4 exposures to be able to achieve this automatic word identification but other students may need over 20 exposures. ( I would make quick spelling assessments during each intervention session to continuosly monitor developing orthographic understandings. It is also helpful to get a sense of how many exposures each student needs to achieve orthographic mapping or unitization.)

Teaching in Isolation vs Context

It is also important to note the importance of teaching words in isolation and in context.  Students who were taught words only in isolation had better spelling but struggled with meaning tasks while students who were taught words in context had better understanding of meaning but struggled with spelling, according to Ehri’s research.  

In SWI, the target word is presented in context and then the word is analyzed in isolation for meaning, structure and grapheme/phoneme correspondences.  I have asked students after the inquiry to use the word in a meaningful context to “show what they know” (see here) and Ehri’s discussion highlights the importance to me of bookending the isolated word study with meaningful context.

Should You Teach Syllabication?

Structured Word Inquiry teaches how our orthographic system is really organized and besides the shock of -tion not being a suffix (it’s -ion), the next shock was when I found out that English is not a syllable-timed language, it is a stress-timed language.  That means that our instructional focus should be on teaching meaningful units (morphemes: bases, affixes).  (Interestingly, Ehri references a study in the Portugese language (which is a syllable-timed language) that found that teaching grapheme/phoneme correspondences better prepared readers to read and spell than teaching syllables.) Understanding of syllables and syllabication is a key skill, however, and can be addressed as needed during the 4th question of Structured Word Inquiry. Students need to understand stressed and unstressed syllables and how that impacts pronunciation and parts of speech. It may be also helpful to let students know that each syllable in English contains a vowel and to demonstrate the emergence of the schwa sound in word pairs like athlete-athletic and nation-national.

Stretch But Do Not Break

As part of SWI, students learn to tap out the grapheme/phoneme correspondences on their arm and then slide their hand down their arm as they say the whole word.  It was interesting and supportive of this practice, along with a reminder for grapheme/phoneme analysis work, that Ehri found that students who were instructed to “stretch out but not break up” the speech stream of a word had better reading and spelling, including reading medial letters.  I have noted students in the past, before using SWI, that had great difficulty decoding medial letters and I wonder if those students may have benefitted more from a stretching out rather than a breaking up sound by sound speech stream.


SWI students become aware of many bases and affixes through the inquiry process and become better able to analyze unknown words through “spelling-out-loud” and applying their morphophonemic knowledge.  I have seen dyslexic students become so confident as they can trust their ability to deepen their understanding of any word and their understandings become generative and lead to even more confidence.  Given my informal experiences, I was very happy to see that Ehri’s research showed that when compared to a control group, students who were morphophonemically trained showed the greatest transference and gains for reading words and nonwords.  

What are the Stages for Word Reading Development? (From Linnea Ehri)

Pre-alphabetic Phase

-relies on visual/context clues but NOT letter-sound cues.

Example:  “Reading” the McDonald’s sign

Partial-alphabetic Phase

-uses knowledge of letter names and sounds to write and read but can’t decode unfamiliar words.

Example:  Can write and read <dog> but also gets “stuck” on many words.

Full Alphabetic Phase

-acquired decoding skill and can fully analyze and form grapheme/phoneme connections within words to read and spell from memory.

Example:  Student can read and spell many words but struggles with content vocabulary.

Consolidated Alphabetic Phase

Accumulated fully analyzed spellings of many words in lexical memory and has acquired knowledge of larger spelling patterns and morphemes.  Students can write multi-syllabic words from memory.

Example: Students are comfortable writing and reading content vocabulary and other words.

Ehri notes that students should receive both structured phonics- and meaning- based instruction tailored to student needs and developmental phase.  Teachers need to assess students in order to determine appropriate instruction.

Practices to Improve Word Reading Instruction:

  1.  Ehri cites research that spelling is improved by creating “spelling connections” for words.  For example, in order to spell <chocolate>, the research advocates having students pronounce the word by emphasizing the <o>”  choc-O-late.  

 As a SWI practitioner, I would create a very different “spelling connection” based on etymology by framing an inquiry into, “Why is there an <o> in chocolate”? To find out, visit Etymology Online to discover that the spelling may have been influenced by the Mayan word for hot which was <chocol>. The <o> in <chocolate> could be an etymological marker that connects the word <chocolate> to the Mayan word, <chocol>.

  1.  If you are using analogy-based spelling instruction like Words Their Way or other programs, Ehri’s research shows that word reading outcomes improved by adding grapheme/phoneme instruction.  It is a key part of SWI instruction to explicitly map graphemes to phonemes and phonemes to graphemes.
  1.  Students learn letter best when taught using embedded pictorial mnemonics in the  letters as opposed to teaching the letter with a separate picture.  In the distant past, I had used “Secret Stories” and other programs (Ehri mentions “Letterland”) that used letters embedded with pictorial mnemonics to highlight the grapheme/phoneme connection but I realize that I have not been using this tool for SWI.  I was very excited but as I thought about this I realized that there are 44 phonemes that are represented by letter strings of one or more letters and that phonemes may be represented by multiple graphemes.  It didn’t seem enough to just represent the consonant and short vowels graphemes with embedded pictorial phoneme mnemonics, perhaps I should try to represent the range of grapheme/phoneme possibilities.  I need to think more about this but here is a quick attempt at some embedded pictorial mnemonics:

          In SWI, I share the IPA with my students to represent phonemes.  I am wondering if 

          I should create foundational cards with the IPA symbols and the embedded pictorial  


  1.  When teaching for word reading development, teach students to articulate along with letters in order to strengthen the connection between grapheme and phonemes.  Structured Word Inquiry practitioners like Peter Bowers and Rebecca Loveless have taught me to, instead of asking students, “What do you hear?”, ask students “What do you feel?” when articulating the phonemes.  Ehri says that, “Teaching students to segment using articulation along with letters helps to strengthen the connection between graphemes and phonemes.  Articulation is more central to the representation of phonemes in the brain than acoustic cues are.”  Ehri would agree with Rebecca Loveless about the importance of teaching beginning readers to monitor mouth positions and sounds during phoneme segmentation instruction.  This is a reminder to add a mirror and visual representation of the sounds in the mouth to my word teaching toolkit.  

My Next Steps:

  1.  I want to deepen my knowledge of articulation by taking a course on phonetics. (I just signed up for a course in phonetics starting 11/28)
  2. Add a mirror and visuals that show how sounds are articulated in the mouth to my toolkit. (Patti Bottino-Bravo, who is teaching the “Fun with Phonetics class that I am taking recommends these resources.
  3. Ponder if embedded pictorial representations would be helpful for digraphs. (In email conversation with Dr. Ehri, she mentioned that the embedded representations were for learning letters so no, I don’t think at this point that the embedded pictorial representations are necessary for digraphs but it could be used as an intervention for someone having difficulty mapping digraphs.)
  4. Ponder if I want to create reference cards that would contain the IPA symbol with phoneme/grapheme representation and the mouth position. (I do not need to create these as they are available here.
  5. Continue to deepen my orthographic understanding through study of the Real Spelling Toolkit and SWI classes with Pete Bowers, Rebecca Loveless and others. (11/20 Update: I am taking another wonderful class with Dr. Peter Bowers and have recently started attending his free Monday evening Zoom drop-in classes.)
  6. Continue to keep up with current educational research.