Syllable Division Rules: Five Programs, Five Answers
You were taught the syllable division rules as though they were settled. Divide between the two consonants in rabbit. Divide before the single consonant in tiger. Label each syllable, read it, blend it back. If a student still cannot read the word, the problem is practice.
Here is the awkward part. The syllable division rules are not one system taught five ways. They are five different systems, published by five organisations, that disagree about how many rules there are, what they are called, which order they come in, and whether they exist at all.
And the only study that has ever tested the two most-taught patterns against the English lexicon found that one of them holds about seven times in ten, and the other about three.
This is not an argument for abandoning syllable work. It is an argument for knowing exactly what you are teaching, where it came from, and what the published record will and will not support when a parent or a school asks you why.
Every program in this comparison teaches something real. None of them teaches the same thing, and none of the accrediting bodies says which one is right.
The one lexicon-wide test, and what it actually found
Until 2020 nobody had checked whether the patterns hold. Devin Kearns did, in Does English Have Useful Syllable Division Patterns?, Reading Research Quarterly, volume 55, special issue 1, pages S145 to S160. The ERIC record is EJ1267452.
The corpus was 14,844 words drawn from texts for grades one through eight. Two patterns were coded. VCCV, where the rule predicts a short first vowel, as in rabbit. VCV, where the rule predicts a long first vowel, as in mason.
The headline figures from the ERIC abstract:
VCCV followed the expected pattern in 70.6 percent of applicable words.
VCV followed the expected pattern in 30.5 percent.
Restricted to two-syllable words, VCCV rose to 78.8 percent and VCV to 47.3 percent.
In words of three syllables or more, VCCV fell to 62.5 percent and VCV to 18.8 percent.
Comparing only short against long vowels, with schwa and the rest set aside, VCCV reached 94.3 percent and VCV 53.3 percent.
The University of Michigan DyslexiaHelp summary, written by Joanne Marttila Pierson and last modified in May 2025, adds two figures the abstract does not carry. V bar CV followed the rule for long u about 84 percent of the time and for long i about 18 percent. It also quotes Kearns directly.
"The data suggest that there is really no V|CV division pattern at all."
One caution on citing this, and it is the kind of thing worth checking before you put a number in a report. The ERIC abstract and the Michigan summary do not agree on V bar CV in longer words — 18.8 percent in the abstract, about 33 percent in the summary. Cite the abstract figure, or read the article itself. Do not cite both.
What the author has said since
Kearns has not stopped teaching syllable strategies. In a slide deck delivered to the Ohio Department of Education Literacy Academy on 4 June 2026, one slide reads, in full:
"Syllable division is a common approach in programs ... but the data about the value are not strong."
The same deck then teaches a procedure. That combination — the patterns are weak, the work is still worth doing — is the position almost every careful source in this area has arrived at, and it is not what the rule charts imply.
Kearns made the broader version of the argument in The Reading Teacher in 2024, under a title that gives the game away: following the rules in an unruly writing system. Lesley University's literacy centre reached a similar place in its own review of what the research says about syllables.
Wilson organises the rules by how many consonants stand between the vowels
The Wilson Reading System publishes its Student Notebook reference pages inside the freely downloadable WRS Materials Information Booklet. One page is headed The Rules of Syllable Division (3.1 to 6.4), and the organising principle is not the VCCV and VCV labels at all. It is the number of consonants.
Under One Consonant (vcv) the notebook gives three provisions. Use one consonant to close in the first syllable, at substep 3.1. If the word is a compound, split between the two words, at 3.1. And then, at substep 5.2, one consonant usually goes with the second syllable.
Read those two again. The short-vowel treatment arrives at 3.1 and the long-vowel treatment arrives at 5.2, which is two full steps and, for most students, many months later.
Under Two Consonants (vccv) there are five more. Divide between two consonants at 3.1. Keep digraphs together at 3.1. Compound words at 3.1. Then an exception at 5.2 — a blend may stay together in the second syllable, especially if there is a prefix in the first. Then, at 6.4, divide just before the consonant in a word ending in a final stable syllable.
Under Three Consonants (vcccv) there are four more again, including a blend that stays together in the second syllable at 3.2 and a final-stable-syllable provision that divides after the first consonant rather than before it.
What that means in practice
That is roughly a dozen provisions, introduced across substeps 3.1 to 6.4, organised by consonant count and sequenced so that the two competing treatments of the same letter pattern are separated by most of a program.
Two word counts from the same booklet are worth having. Across about 11,300 words, the phrase "six syllable" appears zero times and the phrase "syllable type" appears twice. "Closed syllable" appears 28 times. Wilson's own materials do not present this as a six-types system, whatever the posters in the staff room say.
The booklet also states the purpose plainly: the teaching of the syllable type is to aid in the decoding of unknown words. It is a decoding aid, not a taxonomy to be learned for its own sake.
Barton numbers them, and there are four
The Barton Reading and Spelling System publishes its full level and lesson list. Level 4 is titled Syllable Division and Vowel Teams, and the division rules appear in it as numbered lessons:
Lesson 2: Syllable Division Rule number 1
Lesson 3: Syllable Division Rule number 2
Lesson 7: Syllable Division Rule number 3
Lesson 8: Syllable Division Rule number 4
Four numbered rules, in one level, interleaved with spelling lessons. Level 4 Lesson 1 teaches Open Syllables, Lesson 6 teaches Spelling Schwa and Lesson 9 teaches Three-Syllable Words.
A fifth treatment arrives two levels later. Level 6, titled Six Reasons for Silent-E, carries Lesson 2: Syllable Division with Silent-E, and Lesson 12: Consonant-LE Syllables. Level 7 is Vowel-R Syllables. Level 3 is Closed Syllables.
So Barton's answer to "how many syllable division rules are there" is four, plus silent-e, plus consonant-le, distributed across four of the ten levels. It is a coherent sequence. It is not Wilson's sequence, and the rules do not map one to one.
Reading Horizons reduces the whole thing to two decoding skills
Reading Horizons publishes a parent letter describing its method, copyright September 2018. The method has three components: the 42 Sounds of the Alphabet, the Five Phonetic Skills, and the Two Decoding Skills.
Not six syllable types. Five phonetic skills. Not a dozen division provisions. Two decoding skills, quoted here as published:
"Decoding Skill 1: If there is only one guardian consonant following the vowel, that consonant will move on to the next syllable."
"Decoding Skill 2: When a vowel is followed by two guardian consonants, the consonants will split. The first consonant will stay in the first syllable, and the second consonant will move on to the next syllable."
The worked examples in the letter are beside and program for Skill 1, and subject for Skill 2.
The term nobody else uses
"Guardian consonant" is Reading Horizons vocabulary. It does not appear in Wilson's notebook pages or Barton's lesson list. If you inherit a student mid-program and the previous provider used it, you are not looking at a synonym for anything in your own materials — you are looking at a different unit of analysis.
The letter is also explicit about why the skills exist: the Two Decoding Skills allow students to break words of any length into syllables so they can apply the Five Phonetic Skills to each syllable. Division is in service of the phonetic pattern, and the phonetic patterns number five.
Really Great Reading teaches the types and declines the rule list
Really Great Reading's own document for students with dyslexia describes its approach as helping students digest multisyllabic words by breaking them into the six syllable types. So here, finally, is the six-type framing.
And in the same document, the publisher rejects the rule list outright:
"... common patterns rather than having students memorize a list of syllabication rules that are difficult to apply."
That is a publisher of a structured literacy program, in its own dyslexia positioning paper, saying the syllabication rules are difficult to apply and that it has chosen not to have students memorise them. The substructures it does name are syllable types, prefixes, suffixes, base words and root words, and spelling patterns.
If you have ever sat in a training where the six types and the four division rules were presented as one inseparable package, that document is worth reading. One of the best-known programs in the category treats them as separable, and separates them.
Take Flight's published materials name neither
Take Flight, written by the staff of the Luke Waites Center for Dyslexia and Learning Disorders at Scottish Rite for Children, is among the most researched curricula in this space. Its program page enumerates its content precisely: all 44 sounds of English, 96 letter-sound correspondence rules, 87 affixes, and spelling rules for base words and derivatives.
Count the things that list does not contain. It does not name syllable division. It does not name syllable types. The 2022 elementary flyer organises the program around the five components identified by the National Reading Panel and does not name them either.
This is an absence in published marketing and summary material, not proof of an absence in the curriculum, which is not public. State it that way and no further.
The materials are also restricted, which is why. The program page states Take Flight was designed for use by a Certified Academic Language Therapist, for groups of no more than six students, four days a week at sixty minutes or five days a week at forty-five, across two years and 230 days of direct instruction.
What you can say honestly is this. A curriculum with that much published specificity about affixes and correspondences chose not to foreground a division rule set, and two separate publisher documents decline to mention one.
The accrediting lineage prescribes no rule set either
It would be reasonable to assume the Orton-Gillingham accrediting bodies settle this. They do not.
Course content published by an accredited training provider lists the parts of a lesson plan, and the relevant line reads "Syllable Division according to student needs." According to student needs. Not according to rule one, two, three and four.
That is consistent with everything this site has found about the label. Our post on what "Orton-Gillingham based" actually means established that no organisation adjudicates the claim, and the same is true one level down: no organisation adjudicates which division rules an Orton-Gillingham lesson must contain.
And the professional body that governs educational therapists has nothing to say about it. The Association of Educational Therapists Code of Ethics, September 2025, runs to roughly 2,976 words. The word "syllable" appears zero times. So do "phonics", "decoding", "curriculum" and "method". The word "reading" appears once.
That is not a criticism of the Code, which governs conduct rather than instruction. But the AET Fact Sheet, which is where the role's instructional claims actually live, is silent too: across roughly 615 words of readable page text, "syllable", "phonics", "decoding", "curriculum", "method", "multisensory" and "Orton" each appear zero times.
So the obligation to choose a rule set, and to be able to explain the choice, sits entirely with you.
The federal guide that tells you to decode long words never names a division rule
This is the finding that surprised us most, and it took three shell commands to establish.
The IES practice guide Foundational Skills to Support Reading for Understanding in Kindergarten Through 3rd Grade is the federal government's own recommendation document for exactly this instruction. Extracted to text it runs to 57,591 words. Counted as whole words and as phrases:
"syllable division" — zero
"syllabication" — zero
"syllabification" — zero
"syllable type" — zero
"VCCV" — zero
"VCV" — zero
"closed syllable" — one
"open syllable" — one
"schwa" — six
"multisyllabic" — six
The single occurrences of "closed syllable" and "open syllable" both sit inside one table of spelling patterns, each with hyphenated examples — in-sect and stu-dent for closed, pro-gram and tor-na-do for open. The table shows where the breaks fall. It never gives a rule for finding them.
What the guide says instead
Where the guide does address long words, it recommends flexibility rather than a pattern. Its text tells teachers that when students sound out or blend a multisyllabic word, they should be encouraged to be flexible with their vowel sounds, and it discusses reducing the schwa.
That is the same advice the research-side sources give, arrived at independently. Set against the five published programs above, the federal guide is the only document in this comparison that recommends a procedure without naming a single division pattern.
What the three trials actually tested
Here is where the evidence gets thin, and where a practitioner has to be careful. The AIM Institute's deep dive on syllable types and division assembles the trials, and its reading of them is more cautious than most summaries you will see.
Three studies are named. All three concern struggling readers in the middle grades and above, and not one of them isolates a division rule set as the single active ingredient.
Diliberto and colleagues. Rule-based instruction improved real-word reading, nonsense-word reading and comprehension for struggling middle schoolers. The fluency gain was not statistically significant. AIM also flags that the control group's activities are not clearly described and the groups were not matched. Note that the in-text year and the reference-list year differ in that source, 2008 against 2009.
Bhattacharya and Ehri, 2004. Flexible syllable analysis improved nonword reading, unfamiliar word reading and spelling for struggling readers in grades six through ten, with the weakest readers gaining most.
Filderman and Toste. Students improved on multisyllabic word reading and scored worse than controls on a combined measure of fluency and comprehension. The in-text citation reads 2021 and the reference list reads 2022.
AIM's own conclusion is the one to carry into a parent meeting: the research on syllable division rules as a standalone component is limited, most of the evidence comes from broader programs, and no empirical comparison exists between rule-based, flexible and morphological approaches. None of the three is demonstrably superior to the others.
The time question
Two sources converge on dosage, which is more useful than it sounds. AIM suggests a moderate amount for older struggling readers, in the range of roughly three to fifteen hours, and warns that educators should be cautious about allocating large amounts of instructional time, giving twenty hours or more as the example of too much.
Timothy Shanahan, writing on syllabication, makes the same point from the other direction: in the studies he reviews, students received only two to nine hours of teaching. His reading of Bhattacharya and Ehri is the sharpest sentence in the literature on this question.
"Those regimes that taught rigid spelling rules for syllabication didn't improve reading."
The programs that worked, on his account, were the ones aimed at fostering conditionality and flexibility. He is explicit that he supports teaching syllabication, and equally explicit that teaching it as a rigid set of rules makes no sense because the orthography does not work that way. He also notes, fairly, that Kearns examined a single category of patterns.
You may not draw from any of this that syllable division does not work. The honest claim is narrower and more useful: the trials are small, short, concentrated in older struggling readers, confounded with whole programs, and more supportive of flexible procedures than of fixed rules.
A trial from a transparent orthography, and why it does not transfer
One more study is worth knowing about precisely so you do not over-read it.
Ennemoser and colleagues, in Frontiers in Psychology, randomised 75 poor readers in grade four to a sixteen-session syllable-based intervention or a control condition. The intervention was built on the 500 most frequent syllables those readers encounter. Results: a significant and strong effect on the fluency of recognising single words, and text-level reading comprehension was not significantly improved.
Two boundaries belong in the same breath as that finding. The study was conducted in German, which the authors describe as a transparent orthography, and the intervention taught frequent syllables rather than division rules. Neither the language nor the method matches an English rule chart, so this is adjacent evidence about syllables as a unit, not evidence for or against anything in the five programs above.
It does support one modest claim. The syllable is a real unit in reading, and working at that grain can move single-word fluency without moving comprehension — which is the same split the English trials found.
Worth knowing alongside it: Reading Universe's overview of multisyllabic words reports that multisyllabic words make up 48 percent of the words in first-grade texts and a larger share by third grade, attributing the figures to Hiebert and to Kearns and colleagues. The problem is not rare, whatever you decide to do about it.
Where the schwa breaks all of it
Every rule set in this comparison assumes the first vowel will be long or short. In running text, very often it is neither.
Kearns's own figures make the point arithmetically. VCCV holds 70.6 percent of the time across all applicable words and 94.3 percent when only short and long vowels are compared. The gap between those two numbers is the schwa and its relatives. Roughly a quarter of the cases are words where the question the rule answers is not the question the word asks.
The AIM deep dive makes the same objection qualitatively, pointing at the unstressed vowel in above and at words like patient and action where assigning a syllable-type label is genuinely contestable. Barton gives schwa its own lesson, at Level 4 Lesson 6. The IES guide uses the word six times and tells teachers to expect students to reduce it.
This is the most practically useful thing on this page. When a student divides correctly and still cannot read the word, the rule has usually not failed — the vowel has reduced. The next move is to flex the vowel, not to re-divide.
The flexible procedure, as the published sources state it
If you want a procedure that the research-side sources actually endorse, it exists and it is short. The Kearns deck for Ohio names the check ESHALOV — every syllable has at least one vowel — which Reading Universe attributes to O'Connor and colleagues, 2015. The same page names two further strategies from the literature: BEST, attributed to Benedict and colleagues, 2013, and peeling off, attributed to Lovett and colleagues, 2000.
The flexible division sequence in the Ohio deck runs as follows:
Mark the vowels.
Divide the word one way so as to make a word.
For single-letter vowels, make the syllable open or closed.
If that does not make a word, try the other way.
Check that every syllable has at least one vowel.
Note what is absent. There is no rule number, no letter-pattern name and no syllable-type label to assign. The procedure is a search with a stopping condition, and the stopping condition is that the result is a real word the student recognises.
That is also the approach that sits most comfortably beside morphology. Our post on morphology instruction programs covers the evidence for the third option in AIM's list, and the honest summary remains that nobody has compared the three head to head.
What you can and cannot tell a parent or a school
This is a scope question as much as an instructional one, and the lines are clearer than you might expect.
You can say which rule set your materials teach, with the publisher's own naming, and that other published programs organise it differently. All five documents cited above are public.
You can say what the lexicon study found, with its figures and its corpus, and that the author of it still teaches syllable strategies.
You can say that the trials are small and mostly concern older struggling readers inside broader programs, and that no published comparison ranks the three approaches.
You cannot say that a division rule set caused a particular student's progress, because none of the trials isolates one. You cannot say that another provider's rules are wrong — they are a different system, and no body adjudicates between them.
And nothing here is diagnostic: a student who cannot divide a word has told you something about that word, not about a condition. Our post on what a screener can and cannot tell you covers that boundary in detail.
If you are being asked to justify a program choice, the four factual axes this site uses for instruments apply here too: what the publisher says it teaches, how it is sequenced, who may deliver it, and what the independent evidence addresses. Reading intervention programs compared and the structured literacy training programs comparison both work that way.
How this fits the rest of the record
Three patterns from across this site converge on the same point, and together they are more persuasive than any one of them.
First, labels in this field are rarely adjudicated. What "Orton-Gillingham based" means, what "decodable" means, what "the science of reading" names and what "aligned" to the IDA standards means all turned out the same way. Syllable division rules are the fifth instance.
Second, programs specify what the research does not. Mastery criteria, exit criteria, group size, repeated reading protocols and error-correction-adjacent dosage all show manuals carrying numbers the evidence base does not supply. The division rules are a purer case: here the manuals disagree with each other as well.
Third, the measurement posts matter here. A student's progress on multisyllabic word reading is measured with instruments that have their own boundaries — see oral reading fluency norms, rapid automatized naming, phonological awareness assessment, spelling assessment and the standard error on the score. If you change rule systems mid-caseload, the graph will not tell you whether the change helped.
For the role boundaries underneath all of this, the pillar is educational therapist versus learning specialist versus executive function coach, and what educational therapy actually is maps the scope. Progress monitoring decision rules and reading program evidence ratings cover how a change would have to be evaluated. Is educational therapy evidence-based sets out the general version of the problem.
Where the published record runs out
Five gaps, named plainly, because an honest list of what could not be verified is worth more than a confident paragraph.
Take Flight's actual treatment of syllable division is not public. The curriculum is restricted to trained therapists. Two publisher documents do not name it; that is all this page claims.
Wilson's substep numbering is read from Student Notebook pages reproduced in a materials booklet, not from the teacher's manual. The provisions and their substep numbers are as the booklet prints them. Whether the manual adds more is not publicly checkable.
Barton's four numbered rules are named but not stated. The published lesson list gives the titles. What Rule number 1 says is in the materials, which are sold, not published.
The Kearns 2020 article was read through its ERIC abstract, the DOI landing page and two independent secondary summaries, not in full text. The figures are the abstract's own, and the discrepancy noted above is a reason to read the article before relying on the longer-word figure.
No comparison trial exists. Nobody has randomised students between Wilson's provisions, Barton's four rules, Reading Horizons' two skills and a flexible search procedure. Until somebody does, a practitioner choosing between them is choosing on sequencing, training access and fit, not on evidence of relative effect.
Frequently asked questions
How many syllable division rules are there?
There is no agreed number. Wilson's Student Notebook pages set out roughly a dozen provisions organised by consonant count across substeps 3.1 to 6.4. Barton publishes four numbered rules in Level 4 plus separate treatments for silent-e and consonant-le. Reading Horizons publishes two decoding skills. Really Great Reading teaches six syllable types and declines a rule list. Take Flight's published materials name neither.
Are the VCCV and VCV patterns reliable?
Partly. In the only lexicon-wide study, Kearns 2020, VCCV followed the expected short-vowel pattern in 70.6 percent of applicable words and VCV followed the expected long-vowel pattern in 30.5 percent. Both were more reliable in two-syllable words, at 78.8 and 47.3 percent, and less reliable in longer ones.
Should I stop teaching syllable division?
That is not what the sources support. The author of the lexicon study still teaches a syllable-based procedure, and a June 2026 slide of his says syllable division is common in programs but the data about its value are not strong. The consistent recommendation across sources is flexibility over fixed rules, and a moderate rather than a large share of instructional time.
Does the federal practice guide recommend a division rule?
No. The IES guide Foundational Skills to Support Reading for Understanding in Kindergarten Through 3rd Grade contains the phrases "syllable division", "syllabication", "syllabification" and "syllable type" zero times, and "VCCV" and "VCV" zero times. It names closed and open syllables once each, inside a spelling-pattern table, and it advises flexibility with vowel sounds when blending long words.
Why does the rule work and the word still not read?
Most often because the first vowel has reduced to a schwa, which no division rule predicts. Kearns's figures show the size of the problem: VCCV holds 70.6 percent of the time overall and 94.3 percent when only short and long vowels are compared. The next move is to flex the vowel, not to divide again.
What do I do when a student arrives from a different program?
Treat the previous system as a different system rather than a synonym for yours. A student taught Reading Horizons' guardian consonants has not learned a renamed version of Barton's Rule number 1, and a student who met Wilson's substep 3.1 has been taught the short-vowel treatment of a single consonant without yet meeting the long-vowel one at 5.2. Find out which provisions the student actually holds before teaching over them.
Is a flexible procedure less rigorous?
No, and the published record leans the other way. In the studies Timothy Shanahan reviews, the regimes teaching rigid spelling rules for syllabication did not improve reading while those fostering conditionality and flexibility did significantly better. A search with a stopping condition is a procedure. It is simply not a chart.
Working in this field?
Illuminate Educational Therapy Group publishes this kind of source-by-source comparison for educational therapists, learning specialists and executive function coaches — the documents, the figures, the attributions and the gaps, written for practitioners rather than for parents.
Join the Illuminate educator list to get each new comparison as it publishes.




Comments