Thanks for writing this, Carl! As the co-author of The Writing Revolution book, I've thought quite a bit about why there's been so much less research on writing than on reading, particularly by cognitive psychologists. I've asked a couple of them about that and haven't gotten a satisfactory answer, but I think it may have something to do with the difficulty of evaluating writing as opposed to reading. Reading COMPREHENSION ability isn't actually that easy to evaluate, but psychometricians, policymakers, and others are convinced they can do it.
Writing evaluation, on the other hand, is notoriously subjective, and the writing rubrics intended to standardize the process can easily go awry. One promising alternative approach is comparative judgment as applied to writing evaluation, which is being pioneered by No More Marking, based in the U.K. That approach hasn't really taken hold in the U.S., but I hope someday it will--and maybe even lead to more research on writing.
One other comment: In your introductory paragraph, you say that the "science of reading" has come to mean "a large body of converging evidence about how the brain learns to decode print, the centrality of phonemic awareness, and the role of background knowledge and vocabulary in comprehension." I wish that were the case in the U.S., where "the role of background knowledge and vocabulary in comprehension" is rarely included in the definition. Many states have adopted legislation or policies designed to promote the "science of reading," but practically none of them make any mention of the role of knowledge, and it's often missing from the conversation about reading here.
With new writing CBM work by Adrea Truckenmiller, we have pretty solid CBM writing tools for K-8 in terms of assessment to drive instruction and monitoring of growth. Seeing how low our writing scores are on national measures, this is a great start. I wish more schools were aware of and using these metrics. Like reading, poor writing starts at the earliest stages and even our long standing CBM tools here that have been around for decades are almost nowhere to be found in our schools. While Comparative Judgment is interesting, CBMs are much more scalable and have considerable research behind them. I'm not as clear on the evidence base for Comparative Judgement, but there is far less history there to comment on as afar as I am aware.
When my daughter was a bit younger, I volunteered to lead an after-school creative writing club at her elementary school. There were a half dozen or so kids and we met once a week in the school library for about an hour and a half after school.
I assumed going in that I would have to come up with things for the kids to do. This assumption proved incorrect. For the most part, the kids just wanted time and space to write stories. Some of them would write, and then read, and then chat a bit, and then write some more. Others would just write away steadily until our time was up. Some of them would ask me to read what they had written, and some of them preferred to keep it to themselves. One, the youngest, a very serious child, filled a notebook with a story that she never showed me. A little while ago I learned that another, who had been flagged as writing below grade level, just completed her first draft of a novel.
This was a small and self-selected group of students, and my role in their development as writers was small and largely pretextual. (I specialized in staying out of the way and letting them do their thing.) But what I take away from that experience, and what I regularly see with my community college students, is that people have an innate drive to create and communicate (even if only for and with themselves). When we think about how to teach writing, we should be thinking in the first place about how to connect with and cultivate that drive. That, to me, is the foundation of writing (and of reading, too).
I found much to think about in this piece, particularly your discussion of writing as a means of shaping thought rather than merely recording it.
What stood out to me was the section on “cognitive levers.” If words like although, because, unless, and even though encourage perspective-taking, causal reasoning, consideration of exceptions, and the revision of first impressions, then writing instruction is doing something much larger than helping students produce correct sentences. It is helping them develop habits of mind.
That raises an interesting question for me.
For the past decade, much of the literacy conversation has focused on what students know. Your article seems to point toward something equally important: how students think with what they know.
A student may possess information, but can they weigh competing perspectives? Can they reconsider an initial impression? Can they reason about causes and consequences? Can they recognize exceptions to a rule? These feel less like questions of knowledge and more like questions of judgment.
As a classroom teacher, I often found that students needed both structure and immersion. They needed explicit instruction, but they also needed opportunities to wrestle with ideas, develop a voice, take risks, and discover what they actually wanted to say. Writing seemed to flourish not through structure alone or freedom alone, but through the interaction of the two.
Reading your piece, I found myself wondering whether the next literacy conversation may be less about knowledge versus skills and more about how schools cultivate judgment, wisdom, and the capacity to participate thoughtfully in the “long human conversation” you describe.
This is a most needed post. 56 years ago just beginning my course work to be a high school teacher, I had the fortune to come across Francis Christensen and the Generative Rhetoric of the Sentence. When others were talking about the writing process I was learning about the sentence. Thank you, Carl.
I once read the following: you will only ever write as well as you read. I agree with the explicit instruction and rehearsal and practice a teacher must deliver in the classroom but the more a person reads and imbibes hundreds, thousands of sentences in all shapes and forms and sizes, the more their brain intuitively works out and produces like sentences; there is a limit to instruction that only the consumption of sentences through reading can exceed.
I wouldn’t argue, yet miscommunication seems inherent whenever people try to express their thoughts in words, whether written or spoken.
Maybe that’s why Goethe wrote: “To know someone here or there with whom you can feel there is understanding in spite of distance or thoughts unexpressed, that can make of this world a garden.”
Lingo: the coin of the realm. I used to lurk on the coders websites before the clock turned over at Y2K…it was a lot of fun once I understood their language ;)
This is why at the elementary level in particular 'mentor texts' are so important so that students have models of sentence and text structure to analyze and utilize--in addition, of course, to doing their own independent reading. I remember reading Gone with the Wind in the 9th grade for my recreational reading and subsequently striving to produce "like sentences". There's no bypassing the need for wide reading.
"its the simultaneous orchestration of transcription, sentence-level syntax, discourse structure, domain knowledge, audience awareness and self-regulation, all of it held in working memory at once, and none of these components reduces to or substitutes for the others." Thank you for this important reminder of the complexity involved.
This is why revision guides are so important, the topic of my master's thesis in writing instruction. From Stoddard and MacArthur (1993): "First, the writer notices a problem or discrepancy between the actual text and the intended or ideal text. Second, the writer diagnoses the problem, decides what needs to be changed, and selects a strategy for making the change. Finally, the writer makes the actual change. Viewed from this perspective skilled revision requires goals, knowledge, strategies, and specific writing skills."
If you’ve taught high school the way I did for eight years, you’ll soon find out that you are the exception and writing must be taught, not just assigned. When I was head of the English department, I structured writing instruction so that the eight types of writing students were held accountable for in their senior year (interpretation, evaluation, controversial issue, autobiographical incident, reflection, process, reporting, and biography) received repeated practice spread out over their four years in high school, five major compositions each year. Most students need to be taught the difference between evaluating a book and interpreting it—between narrating a story and arguing a position. You’re not a freak. But you’re in the minority.
I was referring to literacy on a scale. Basic literacy is just knowing phonetic sounds that letters make. However, literacy is never something that you "arrive" at. It is a skill that continuously improves with reading. It is not just knowing how to read words. It is fully mastering language, understanding continuity of thought, and being pulled into "the great conversation." It is understanding that all writing is interacting with other writing, responding and reacting to each other and to the world around them.
Agree about problems upstream. Science writer Ed Yong suggests that problems at the level of individual sentences often arise from problem at the level of structure, and problems at the level of structure often arise from problems at the level of story (he's speaking of story in journalistic terms). It follows that finding problems at one level can help to diagnose problems at the next level up, and solving problems one level up can help to solve problems at the next level down. I've found Yong's suggestion very helpful for the students I work with. (I teach writing at a community college.)
And of course, a lot of writing problems are even more upstream than that:
life circumstances, stress and anxiety, prior bad experiences with writing for school, etc.
There are more complications at the other end, which I discovered teaching kindergarten. At that age, writing using invented (estimated) spelling promotes phonemic awareness, which improves reading. Restricting writing to 'correct' sentences can be problematic for this reason, though we are dictating sentences to them to practice correctness--but that's separate from self-generated composition. I have found that Steve Graham's recommendations are the most balanced on this point: he understands the importance of correctness, but also understands the need to allow students 'free expression' at this age so that they become comfortable with mapping sound to print and can achieve a certain level of fluency before imposing correctness on them.
Very interesting! The challenges at the early end are fascinating and really do require a distinctive expertise (which I confessedly lack, though as it happens my mother was also a kindergarten teacher).
- A big reason why the science of reading (SoR) movement took off is the influence behind it. Hanford's series is the most prominent example of this. There has been a concerted effort to influence states to adopt language that mandates practices and programs that reduces reading instruction down to a set of skills. See Dr. Elena Aydarova's research (https://elenaaydarova.com/); she follows the money invested in the SoR movement and connects the dots between the different power players.
- The quest for an exclusive approach to teaching readers reflects a desire for a simple solution, which also explains why the SoR movement took such as strong hold. In their excellent resource Learning to Improve, Anthony Bryk and colleagues refer to this desire as "solutionitis". What does work: investing in teacher quality through time, resources, and opportunities to support professional collaboration around shared problems of practice. Linda Darling-Hammond refers to this process as developing "adaptive expertise".
I agree that reading outcomes need to improve, and that a primary driver for that improvement is an increase in high-quality instruction.
Where I push back here is when you state that the reason for these outcomes is "the incapacity of so many teachers to turn out students who can read". There are too many causal factors to assign the blame to teachers.
For example, in my state of Wisconsin, I ran a correlation between reading scores on last year's state test and students living in financial hardship. 63% of the influence on reading achievement scores can be attributed to whether or not a student qualifies for free or reduced lunch.
We need to be mindful of the simple stories we create due to not examining additional factors that are leading to the outcomes we are seeing. It's what leads to solutionitis and passing mandates that don't address the root causes.
About 25 years after I started teaching (1997?), I had the fortuitous privilege of participating in a series of workshops taught by Judith Hochman. After implementing the strategies I learned for teaching sentence writing, my elementary level SPED students improved significantly. After years of failure, I had finally been able to help students be better writers; I no longer had to feel like a failure. A few years later, I had the opportunity to conduct staff development and I chose to focus on the sentence writing strategies I had learned. Unfortunately, none of the teachers I worked with recognized its value and never followed through. Fortunately, many years later, as an assistant professor, I taught these strategies to my pre-service graduate students with the help of "The Writing Revolution" text.
Steven, I’d be interested in your take on the use of SRSD with the Writing Revolution/Judith Hochman approach as discussed by Karen Harris who suggests there are benefits in using both and.
“… the data, there are no experimental studies of the writing revolution. There are no controlled studies….There's a lot of good stuff in the writing revolution, but it is not teaching students how to write effectively across genres.”
The data and information you need for kids who struggle is more in the levels of language as well as the cognitive profiles research of Troia. Kids who struggle are going to need a lot of work on transcription, spelling and sentence level work before heavy emphasis is laid on larger text. While there might not be direct evidence from the Writing Revolution *itself*, I feel these SRSD adjacent people are doing a disservice to how much work *some* kids need at these levels. There is a lot more to all of this, but even Grahm himself said a student should be pretty proficient at the sentence level before engaging with SRSD strategies. That is straight from the horses mouth.
FBV108, do you have a reference for where Graham said a student should be proficient at the sentence level before engaging with SRSD strategies?
[AI Overview Steve Graham did not state that students must be proficient at the sentence level before engaging with SRSD. In fact, his research contradicts this sequence, finding no evidence to support starting with small parts of writing and moving up. Instead, he recommends embedding sentence-level work within the revising stage of the writing process.]
Thanks for the question and helping me clarify this. I double checked, and the actual reference was to a co-authored text for which Grahm was only one author on (the Powerful Writing Strategies book), but I originally came across this in Shapiro's Academic Skills Problems:
"It is important to consider transcription interventions first because transcription represents a potentially significant limiting factor of writing output and quality (Limpo et al., 2017). Students will be significantly constrained in drafting high-quality written content if they are severely limited in their ability to get words onto the page. Even among students in the seventh and eighth grades, Limpo et al. (2017) found that better handwriting fluency was associated with stronger writing planning, which in turn was predictive of overall writing quality. In fact, developers of SRSD recommend that students should be able to write at simple sentences (e.g., “The dog ran”) before starting SRSD (Harris et al., 2008). Transcription need not be perfect; if students can only write just very short and simple sentences, they can start interventions like SRSD while transcription is addressed. But if students have severe difficulties with handwriting and spelling that impair theirability to form even simple sentences, then transcription should be addressed first. " (pp.320-321)
There may be another quote in there that adds more details, but to go to the direct source it says, from the introductory material to Powerful Writing Strategies (p.7):
"SRSD is meant to supplement, not supplant, the larger writing curriculum. It does not represent a complete writing program. Although writing and self-regulation strategies are essential to writing development, they are not the only abilities students must develop to improve their writing skills. Furthermore, to profit from SRSD instruction, young students must be able to write complete sentences. It is therefore inappropriate to ask students to engage in these writing and self-regulation strategies before they can
write sentences. For young students, a simple sentence (e.g., The dog ran.) is fine. Writing skills such as sentence construction and usage, spelling, and handwriting must also be developed. SRSD, while not a panacea, is an important tool for teachers' repertoires—one that makes a difference"
The Shapiro quote actually understates what is in there and I may have overstated, but we are in this zone. I work with kids with disabilities and have long grappled with the advice from modern proponents of this method and it is clear that some (or most of the kids I work with) should be focusing on heavy foundations like the above. Transcription includes not only handwriting but spelling and basic punction for sentences. For impacted kids that can easily take up to a year or two to get on the line and be fluent enough to start to generalize to larger form writing. This would also extend to kinder and possibly first grade in many schools, unlike the push you see in the modern SRSD iterations that leave kids I work with, and others, in the dust. In some settings with less careful or trained teachers, I would guess that it could potentially create complex problems for both classroom teachers and interventionists imho as well.
Further, Troia et al. (2022) that I referenced, identified that transcription and spelling skills are empirically linked to overall writing proficiency, and struggling writers form a distinct, identifiable profile. They found about 39% of writers in the 4th and 5th graders studied were in this camp and needed definite work on transcription skills. Pairing this with the above, it is hard to say to what extent these even older kids needed to solidfy sentence level skills, but you can see these concerns cover more than just young children and the lack of kids moving up to middle school never punctuating sentences is likely just as much to do with our instruction as it does with any other modern excuses.
While SRSD has an evidence base, specifically for whom and when needs a considerable amount more scrutiny imho. I am sure they are getting results over "busines as usual", but if you look at the other groups in Troia's study there are 40% who primarily need strategy instruction, which could be enough to bump up plenty of effect sizes across many schools and districts. Along with research, please be extremely careful with AI, particularly Gemini/Google and Chat GPT. They are only of so much use if you don't know how to prompt it well, but it will give you what you want to hear with great ease.
References:
Harris, K. R., Graham, S., Mason, L. H., & Friedlander, B. (2008). Powerful writing strategies for all students. Paul H. Brookes.
Shapiro, E. S., Clemens, N. H., & Shapiro, E. S. (2022). Academic skills problems: Direct assessment and intervention (5th ed.). Guilford Press.
Troia, G. A., Wang, H., & Lawrence, F. R. (2022). Latent profiles of writing-related skills, knowledge, and motivation for elementary students and their relations to writing performance across multiple genres. Contemporary Educational Psychology, 71, Article 102100. https://doi.org/10.1016/j.cedpsych.2022.102100
I didn’t rely on the AI quote and had already quoted and referenced Dr Graham in earlier comments, as had others. Your unreferenced claim contradicted those referenced quotes.
I agree with Dr. Leslie Laud (who has also commented here) that it helps to have a variety of teaching foci occurring in parallel, rather doing “ a lot of work on transcription, spelling and sentence level work before heavy emphasis is laid on larger text”.
Teaching of text structure can start before students write sentences, where kindergarten students are using pictures and invented spellings as the foundational skills of spelling, decoding and sentence writing is taking place. You can see the details in this presentation given at the 2024 PaTTAN Literacy Symposium website:
I'm aware of all these opinions. You can do anything you want with kids, but give me some more robust and well designed studies with the populations I tend to work with and I'd be happy to read through it to see if there is any actual weight here. Otherwise the quotes above are crystal clear and in line with other educational, intervention, and cognitive literature. Understanding how to put it in the context of Grahm's other quotes and a deeper read of what the SRSD or adjunct literature suggests is important instead of claiming any of this settled science at the foundational levels. It is not.
"The first thing to consider, which is common to all of these traditions [SRSD and writing revolution] is to privilege the sentence as the unit of instruction."
I don't think this is exactly true for SRSD. Do you have a reference? Or a reference for any peer reviewed research which states this?
Matt Writer, that was also my impression, although there may have been a shift towards doing sentences and paragraphs in parallel with ThinkSRSD now Releasing Writers.
In Dr. Karen Harris and Dr. Steve Graham’s Self-Regulated Strategy Development (SRSD) framework, paragraph-level planning (the overall text structure) takes instructional priority over sentence-level grammar.Rather than waiting for students to achieve sentence mastery before they write, SRSD teaches them to brainstorm and build whole paragraphs first using mnemonic organizers like POW (Pull apart the prompt, Organize your notes, Write) and TREE (Topic sentence, Reasons, Explanations, Ending).The Sentence vs. Paragraph PriorityThe Cognitive Approach: SRSD research shows that tackling paragraph structure and higher-order planning first actually builds a scaffolding that improves sentence writing and vocabulary. The paragraph gives the student context and purpose for their sentences.Sentence-Level Work is Integrated: When sentence structure is taught, it is integrated within the paragraph-writing process (such as using transition words) so that foundational mechanics and text composition develop together.Independence and Fluency: Students are equipped with cognitive strategies and self-regulation procedures (like self-talk and goal-setting). This prevents the "bottlenecking" that happens when a student tries to perfect an isolated sentence before knowing the flow of their argument.
Agreed, this is an oversight in this blog. The research suggests that sentence level skills load the lowest on writing quality. They are better incorporated at the lower end for kids who need a lot of work with the word and sub word level, or at the upper level for kids who can jump right into full text production. Sentence level work is more adjunct in a way.
Lively discussion, and so good to see this much engagement with the actual research.
I have been tracking the writing research closely for over three decades now, and contributing to it for about twenty years. I come at this as a classroom teacher who uses the research every day with students, so my perspective, and my views, come from that angle, and they keep evolving.
Yes, as several people here have said, teaching text structure is the most well researched starting point we have, paired with teaching the writing process step by step: picking ideas, planning, writing, revising.
One reason sentence level work gets less attention may simply be that research is often short term and costly to run. Sentence level growth takes longer to show up. Both matter, though. (As an aside, our field really struggles with holding both/and two things at once. Teach text structure AND sentence, both, and at once from the start.)
My takeaway is that if we start by teaching the writing process and text structure (something like Topic, Information, Details, End), students see fast gains. That matters a great deal, especially early on. But alongside that, we need to teach sentence writing too. Sentence expansion and sentence combining both have real evidence behind them.
We should also be teaching students to strengthen every level of language, including word choice. A planner like TIDE-L (Topic, Information, Details, End, Language/vocabulary) can help students hold all of that together.
Here is the point I care most about. I would ask that we call this body of work EBP-W (evidence based practices for writing) rather than naming it after one branded, commercialized approach, whether that is The Writing Revolution or SRSD. SRSD itself goes back much further than the name suggests, and it is built directly on earlier researchers who get erased once a single brand becomes the shorthand for the whole field. Deshler's six step model is where SRSD's six stages come from. Englert's POWER is where SRSD's POW comes from. I say this as someone who misunderstood that history myself for a long time, and contributed to erasing the people who came before. Englert's Cognitive Strategy Instruction, Olson's Cognitive Strategies Toolkit, and so many others did foundational work that gets lost once the science of writing gets named after one approach by only using SRSD rather than EBP-W for the wider based that includes it and so many others with similar findings.
On a hopeful note, it is wonderful to see research going back to the 1970s finally entering these conversations. Let's keep the focus on the findings, EBP-W, rather than the researcher brand names. From there, text structure is the MVP, but sentence level instruction can and should be taught right alongside it.
Considering the overlays of all education, I think just talking about doing "evidence based practice" period, is key. These movements and brands are the death toll.
"Graham's meta-analytic evidence converges on the same point from the other direction, with sentence combining showing reliable effects on writing quality across grades and learner profiles."
While Graham's support for sentence combining is clear, there's some confusion about his recommendation that we teach this skill within the context of a piece of writing. In his keynote address at last year's Multlit conference, he stated:
"So I understand that The Writing Revolution has gotten a good bit of press in your country [Australia]. That they often cite me as being support for what they do. They're right in a way, as … teaching sentence construction skills makes a difference for kids. However, what they don't have evidence to suggest is that starting with a small piece of writing and moving up--like a sentence, moving up to a paragraph, then moving up to a larger text--makes any difference whatsoever. That's not been tested systematically. What we do know is that when you bring various aspects of writing, when you have multi-component writing instructional programs, that's where we get our effects. So, on the one hand, they could be right, they just don't have evidence to support that.”
FBV108, as discussed below, it would help if you quoted this with a reference as I don’t believe this statement is accurate. I haven’t read everything Steve Graham has written in his long career so my beliefs may be mistaken so I am happy to be corrected.
I gave full references in the other statement, please refer there. It was in a co-authored book, so partial attribution - but stated in conjunction with Harris and others.
Explicit writing instruction in K-2 is my passion and built on 20 years of building step by step lessons. I wonder if we can connect for a discussion about this in Eugene? I’ll be at the NIFDI conference all week. There IS a science of writing!
I'm going to be up there. I'd love to pick your brain on this topic. Your approach seems to stand in contrast to some of what I see being pushed with some of the strategy instruction that has gained so much traction thanks to social media. I think you can take the research in either direction, but it's a critical topic imho.
Great writers come to us through a sliver in the universe. They are never born and certainly not created, but rather emerge.
That said, a great writer will always need to be capable of terrible writing - and perhaps even solid writing.
Just as Monet mastered the classical style, so too much all writers. But if they ever write the way they were taught, they never be great, or even interesting.
Only a gun to the head can make some writing interesting.
When most people talk about writing instruction they mean sentences and paragraphs.
You mentioned that we need to "privilege the sentence as the first unit of construction".
Actually the first unit of construction is the letter, then the word. Then words without any sentence form. A lot of students miss an adequate introduction in K-1 and don't get anywhere near the hours of practice they need just on letters and words.
These are the years (K-2) for which research is missing, because I believe there's still a belief that writing is "emergent" in these grades and will just appear after being exposed to activities. We're no longer teaching whole language reading, but K-1 teachers are doing a lot of what I would consider whole language writing. Either that or trying to expect little kids to meet ridiculous standards.
Reading research skews towards the early grades, whereas writing research skews towards 5th grade and up. Reading has become a science partly because of the studies of those earliest years.
Kindergarten and first grade are filled with activities, but not instruction. I have created what I called The Pyramid of Writing Skills for K-2, a way to ease children into writing without cognitive overload. My writing curriculum is based on this and teachers breathe a sigh of relief to finally have appropriate instruction. I know that SRSD and The Writing Revolution can "work" in the youngest grades (not sure how that goes or if it works for everyone) but I think they would have the most impact when kids have some foundational skills.
Let me know through a comment if you'd like a copy of my article.
Have you looked much at the SRSD literature? Particularly for any of it possibly pertaining to younger grades? I would be curious on your viewpoint here.
Yes, I have. There is evidence that it is successful in second grade and up. It has also been used in K-1 and I have seen anecdotal evidence of success. The question is: who is the strategy benefitting? Kids who are already confident writers? Or all novice writers? Also, are the studies measuring how well kids can use the strategy? Or their writing skills in general? Kids can do anything with a lot of scaffolding, help and guidance. That doesn't necessarily mean they have stored learning in long term memory. I
I’m in the process of helping my 5 YO grandson to read & write. He’s “tutored” by a lady who relies on my long past experience as a reading tutor (via Phono-Graphix) as she’s mostly been a babysitter because my stubborn son won’t enroll him in a proper homeschooling program (forget public school). Our rural location & his single Dad status are partly why but that’s just part of the problem my grandson faces.
Reading the many replies to this article sadly reminded me a bit of different arguments for religious perspectives. I’m more of the big tent type. I’m also reminded of my training experience in spelling where I was informed that in Shakespeare’s time it was considered creative to spell words differently (even within the same piece) as long as the writers intent was discernible!
I’d like to recommend Justin Deschamps’ recent Substack article: Cultology: The Pen and the Sovereign Mind. I consider it a brilliant piece that is so much more than its parts. We can all learn about how “writing by hand” makes us choose and order our thoughts and words in a meaningful way.
I’ve had to choose my remarks here due to time constraints but I look forward to hearing, reading & writing more as time allows as I’m thankful for everyone’s thoughtful contributions.
My traveling days are mostly behind me but I spent many years traveling to conferences of learning that interested me. Fortunately the internet now provides access to the like…kinda like the Buckee’s of learning ;)
I do question your remark about speakers & writers reason for doing so is to steal from people?
We teach everyone how to paint. It’s excellent to learn the basics of painting.
We teach everyone to write. Great, everyone should know how to string a sentence.
BUT where once necessary to be a solid, professional grade writer, now it’s not necessary. And this is a great thing!
However, will professional grade writers have more power than others? Definitely before. You needed to be a great writer to be a professional success.
Not anymore. And in fact, there’s only one reason to become a great writer now, just as with the advent of writing, there was only one good reason to become a great speaker: to enchant and confide…or rather: to steal from people.
Oracy and literacy are NO LONGER the dominant medium for knowing: algorithmacy is.
And if you don’t know about algorithmacy, come down to Trinidad this fall: algorithmacy.org
I really appreciate this piece, especially the emphasis on sentence-level instruction and the role of conjunctions as cognitive levers. That is very close to what I see clinically. Children who struggle with writing are often not struggling because they lack ideas. They are struggling because they are being asked to coordinate spelling, syntax, vocabulary, sentence structure, background knowledge, and working memory all at once.
I do think this also raises a question for the reading conversation. If writing cannot be reduced to a single mechanism, I’m not sure reading can either. The public version of the science of reading has often leaned too hard on the idea that reading is primarily the mapping of letters to sounds. But English orthography represents language through meaning, structure, history, and phonology together. That becomes obvious in spelling and writing, but it is also true in reading.
So perhaps part of the reason writing has been left out is that writing makes the system harder to oversimplify. It forces us to look at words, sentences, and meaning together.
Writing and reading can’t be reduced to a single mechanism but this does not mean research can’t be done on multicomponent approaches and also on the components required for effective literacy teaching.
With writing, Drs. Karen Harris and Steve Graham, along with other researchers, have conducted more than 170 studies on SRSD for writing instruction. And they have conducted studies on difficulties with text transcription skills such as handwriting or spelling which may constrain young children’s writing development. It’s not either or, it is yes and.
Reading can’t be reduced to a ‘single mechanism’ and yet it is still possible to conduct research on PA, phonics, knowledge of the alphabetic principle, morphology, fluency and other components separately and when included in multicomponent approaches to teaching.
The newer models of reading (David Share, Universal Combinatorial Model 2025; Nell Duke and Kelly Cartwright, Active View of Reading 2022; Maryanne Wolf POSSuM (2021) are not moving away from teaching phonics to beginning readers. M. Wolf’s POSSUM model (https://www.shankerinstitute.org/sites/default/files/documents/elbow-room-report.pdf)“… the teaching of these multiple processes should not be conceptualized as a checklist to do in isolation, but rather as the component parts that are to be integrated daily. … Phonemes need letters. Phonics needs semantics, syntax, and morpheme knowledge. Words need stories. The reading brain connects all of these processes, and so should our teaching.”
Similarly, with multicomponent approaches in teaching writing, teachers are not moving away from teaching components of writing systematically and explicitly. Researchers are not moving away from studying the components to find out what is necessary (but not sufficient) in writing instruction.
I agree that this is not either/or. Components can and should be studied, and explicit instruction matters in both reading and writing.
My concern is not with component research itself. My concern is with what happens when one component becomes the public organizing frame for the whole system.
In the reading conversation, that has often meant treating reading as though it is primarily a matter of mapping letters to sounds, with morphology, syntax, semantics, and etymology added later or treated as enrichment.
The quote from Wolf actually points to the distinction I’m trying to make: “Phonics needs semantics, syntax, and morpheme knowledge. Words need stories.” That is not phonics standing alone. That is phonology functioning inside a larger language system.
So yes, we can study components. But instruction still has to help children understand how those components work together.
Writing makes that especially visible because children cannot write well by coordinating only one part of the system. I would argue that reading is the same.
English orthography represents meaning, structure, history, and phonology together, and our teaching should reflect that from the beginning.
This is where you part ways with people like Wolf who (based on research evidence) emphasise the importance of knowledge of the alphabetic principle and phonics at the beginning. [Phonics taught in meaningful morphemes (cat, mat, sat) which come from meaningful stories (Pete the Cat, The Cat in the Hat, and importantly, children’s own stories of their pet cat which they write about using evidence based SRSD approaches)]
I don’t think I part ways with Wolf as much as that framing suggests. When I listen to Wolf present, I hear an integrated model of reading, not a phonics-only model.
I agree that children need alphabetic knowledge. I agree that they need to understand that letters can represent phonemes. My concern is with making that relationship the organizing frame for beginning instruction, especially when English orthography represents meaning, structure, history, and phonology together.
Using words like cat, mat, and sat in meaningful stories is certainly better than isolated drill. But the grouping is still organized by sound pattern, not by shared meaning or structure. That distinction matters.
So the question for me is not whether phonology belongs in early instruction. It does. The question is whether beginning readers should be taught from the start that spelling is a meaningful, structured system, rather than first being taught to expect spelling to work primarily by sound.
That is an interesting hypothesis that could be tested. Write a program for beginning readers so a controlled study can be done comparing your approach with something like UFLI Foundations. Publish and have the research replicated.
At the same time, “build a full program, run a controlled study, publish it, and get replication” is not the entry fee for raising a legitimate theoretical and instructional question.
My claim is not that research should stop at hypothesis. My claim is that the phonics-first frame rests on assumptions about English orthography that deserve scrutiny. If English spelling represents meaning, structure, history, and phonology together, then it is reasonable to question whether beginning instruction should be organized primarily around letter-sound mapping.
A comparative study would be useful. But the linguistic question does not disappear while we wait for one.
What stood out to me is the idea that writing is not simply a way of recording thought but a way of creating it. Reading lets us borrow the conclusions of others, but writing forces us to confront our own half-formed ideas. A vague feeling becomes a sentence, and a sentence becomes something we can question, refine, and understand more clearly.
I think I understand an argument until I struggle to write it independently. Writing helps clarify what is ‘in my mind.’
Writing and the extended mind thesis.
Clark and Chalmers used the famous "Otto's Notebook" thought experiment to develop the concept of the extended mind:
Imagine a man named Otto who has Alzheimer's. To help him remember things, he carries a notebook everywhere he writes down important information, effectively using it as his biological memory. According to the extended mind thesis, because Otto relies on the notebook exactly the way a healthy person relies on their brain's memory, the notebook and Otto's brain constitute a single, coupled cognitive system. Taking the notebook away is functionally equivalent to taking away a part of his mind.
I recently learned about Steven Pinker’s “The Sense of Style” that uses insights from cognitive science to improve writing. It may contribute to the “science of writing” (not just writing instruction).
Thanks for writing this, Carl! As the co-author of The Writing Revolution book, I've thought quite a bit about why there's been so much less research on writing than on reading, particularly by cognitive psychologists. I've asked a couple of them about that and haven't gotten a satisfactory answer, but I think it may have something to do with the difficulty of evaluating writing as opposed to reading. Reading COMPREHENSION ability isn't actually that easy to evaluate, but psychometricians, policymakers, and others are convinced they can do it.
Writing evaluation, on the other hand, is notoriously subjective, and the writing rubrics intended to standardize the process can easily go awry. One promising alternative approach is comparative judgment as applied to writing evaluation, which is being pioneered by No More Marking, based in the U.K. That approach hasn't really taken hold in the U.S., but I hope someday it will--and maybe even lead to more research on writing.
One other comment: In your introductory paragraph, you say that the "science of reading" has come to mean "a large body of converging evidence about how the brain learns to decode print, the centrality of phonemic awareness, and the role of background knowledge and vocabulary in comprehension." I wish that were the case in the U.S., where "the role of background knowledge and vocabulary in comprehension" is rarely included in the definition. Many states have adopted legislation or policies designed to promote the "science of reading," but practically none of them make any mention of the role of knowledge, and it's often missing from the conversation about reading here.
With new writing CBM work by Adrea Truckenmiller, we have pretty solid CBM writing tools for K-8 in terms of assessment to drive instruction and monitoring of growth. Seeing how low our writing scores are on national measures, this is a great start. I wish more schools were aware of and using these metrics. Like reading, poor writing starts at the earliest stages and even our long standing CBM tools here that have been around for decades are almost nowhere to be found in our schools. While Comparative Judgment is interesting, CBMs are much more scalable and have considerable research behind them. I'm not as clear on the evidence base for Comparative Judgement, but there is far less history there to comment on as afar as I am aware.
When my daughter was a bit younger, I volunteered to lead an after-school creative writing club at her elementary school. There were a half dozen or so kids and we met once a week in the school library for about an hour and a half after school.
I assumed going in that I would have to come up with things for the kids to do. This assumption proved incorrect. For the most part, the kids just wanted time and space to write stories. Some of them would write, and then read, and then chat a bit, and then write some more. Others would just write away steadily until our time was up. Some of them would ask me to read what they had written, and some of them preferred to keep it to themselves. One, the youngest, a very serious child, filled a notebook with a story that she never showed me. A little while ago I learned that another, who had been flagged as writing below grade level, just completed her first draft of a novel.
This was a small and self-selected group of students, and my role in their development as writers was small and largely pretextual. (I specialized in staying out of the way and letting them do their thing.) But what I take away from that experience, and what I regularly see with my community college students, is that people have an innate drive to create and communicate (even if only for and with themselves). When we think about how to teach writing, we should be thinking in the first place about how to connect with and cultivate that drive. That, to me, is the foundation of writing (and of reading, too).
To put this another way, I'd suggest that the teaching of writing needs to start not with the sentence, but with the writer.
They are mutually necessary.
Carl,
I found much to think about in this piece, particularly your discussion of writing as a means of shaping thought rather than merely recording it.
What stood out to me was the section on “cognitive levers.” If words like although, because, unless, and even though encourage perspective-taking, causal reasoning, consideration of exceptions, and the revision of first impressions, then writing instruction is doing something much larger than helping students produce correct sentences. It is helping them develop habits of mind.
That raises an interesting question for me.
For the past decade, much of the literacy conversation has focused on what students know. Your article seems to point toward something equally important: how students think with what they know.
A student may possess information, but can they weigh competing perspectives? Can they reconsider an initial impression? Can they reason about causes and consequences? Can they recognize exceptions to a rule? These feel less like questions of knowledge and more like questions of judgment.
As a classroom teacher, I often found that students needed both structure and immersion. They needed explicit instruction, but they also needed opportunities to wrestle with ideas, develop a voice, take risks, and discover what they actually wanted to say. Writing seemed to flourish not through structure alone or freedom alone, but through the interaction of the two.
Reading your piece, I found myself wondering whether the next literacy conversation may be less about knowledge versus skills and more about how schools cultivate judgment, wisdom, and the capacity to participate thoughtfully in the “long human conversation” you describe.
Thank you for a thought-provoking article.
This is a most needed post. 56 years ago just beginning my course work to be a high school teacher, I had the fortune to come across Francis Christensen and the Generative Rhetoric of the Sentence. When others were talking about the writing process I was learning about the sentence. Thank you, Carl.
I once read the following: you will only ever write as well as you read. I agree with the explicit instruction and rehearsal and practice a teacher must deliver in the classroom but the more a person reads and imbibes hundreds, thousands of sentences in all shapes and forms and sizes, the more their brain intuitively works out and produces like sentences; there is a limit to instruction that only the consumption of sentences through reading can exceed.
…or only the hearing of many spoken words from birth can impact a large vocabulary.
Yes but even then written syntax is different to spoken syntax. There’s a reading arrangement of words that is different from speaking arrangement.
I wouldn’t argue, yet miscommunication seems inherent whenever people try to express their thoughts in words, whether written or spoken.
Maybe that’s why Goethe wrote: “To know someone here or there with whom you can feel there is understanding in spite of distance or thoughts unexpressed, that can make of this world a garden.”
Lovely!
Lingo: the coin of the realm. I used to lurk on the coders websites before the clock turned over at Y2K…it was a lot of fun once I understood their language ;)
This is why at the elementary level in particular 'mentor texts' are so important so that students have models of sentence and text structure to analyze and utilize--in addition, of course, to doing their own independent reading. I remember reading Gone with the Wind in the 9th grade for my recreational reading and subsequently striving to produce "like sentences". There's no bypassing the need for wide reading.
Yes. I think it was C S Lewis who was asked by a child what advice he could give about becoming a writer. His reply: read, read, read
"its the simultaneous orchestration of transcription, sentence-level syntax, discourse structure, domain knowledge, audience awareness and self-regulation, all of it held in working memory at once, and none of these components reduces to or substitutes for the others." Thank you for this important reminder of the complexity involved.
This is why revision guides are so important, the topic of my master's thesis in writing instruction. From Stoddard and MacArthur (1993): "First, the writer notices a problem or discrepancy between the actual text and the intended or ideal text. Second, the writer diagnoses the problem, decides what needs to be changed, and selects a strategy for making the change. Finally, the writer makes the actual change. Viewed from this perspective skilled revision requires goals, knowledge, strategies, and specific writing skills."
If you’ve taught high school the way I did for eight years, you’ll soon find out that you are the exception and writing must be taught, not just assigned. When I was head of the English department, I structured writing instruction so that the eight types of writing students were held accountable for in their senior year (interpretation, evaluation, controversial issue, autobiographical incident, reflection, process, reporting, and biography) received repeated practice spread out over their four years in high school, five major compositions each year. Most students need to be taught the difference between evaluating a book and interpreting it—between narrating a story and arguing a position. You’re not a freak. But you’re in the minority.
The more literate you are, the more writing becomes second nature.
I was referring to literacy on a scale. Basic literacy is just knowing phonetic sounds that letters make. However, literacy is never something that you "arrive" at. It is a skill that continuously improves with reading. It is not just knowing how to read words. It is fully mastering language, understanding continuity of thought, and being pulled into "the great conversation." It is understanding that all writing is interacting with other writing, responding and reacting to each other and to the world around them.
Agree about problems upstream. Science writer Ed Yong suggests that problems at the level of individual sentences often arise from problem at the level of structure, and problems at the level of structure often arise from problems at the level of story (he's speaking of story in journalistic terms). It follows that finding problems at one level can help to diagnose problems at the next level up, and solving problems one level up can help to solve problems at the next level down. I've found Yong's suggestion very helpful for the students I work with. (I teach writing at a community college.)
And of course, a lot of writing problems are even more upstream than that:
life circumstances, stress and anxiety, prior bad experiences with writing for school, etc.
There are more complications at the other end, which I discovered teaching kindergarten. At that age, writing using invented (estimated) spelling promotes phonemic awareness, which improves reading. Restricting writing to 'correct' sentences can be problematic for this reason, though we are dictating sentences to them to practice correctness--but that's separate from self-generated composition. I have found that Steve Graham's recommendations are the most balanced on this point: he understands the importance of correctness, but also understands the need to allow students 'free expression' at this age so that they become comfortable with mapping sound to print and can achieve a certain level of fluency before imposing correctness on them.
Very interesting! The challenges at the early end are fascinating and really do require a distinctive expertise (which I confessedly lack, though as it happens my mother was also a kindergarten teacher).
Carl, thank you for sharing your thinking here.
I have a few thoughts as well.
- Reading does, in fact, draw on several factors: word recognition and language comprehension, but also and not limited to executive functioning, motivation and engagement, a wide knowledge base, and self-efficacy. There is not a narrow biological pathway to becoming a reader. See Peter Afflerbach's book, Teaching Readers, Not Reading (https://www.guilford.com/books/Teaching-Readers-Not-Reading/Peter-Afflerbach/9781462548613?srsltid=AfmBOoqVJZwaGGCrVGGgKfm-fA3QpaCUcju9MmN3DCZ3FuEdSkFy86B4) for a thorough summary of the research, supporting these different domains.
- A big reason why the science of reading (SoR) movement took off is the influence behind it. Hanford's series is the most prominent example of this. There has been a concerted effort to influence states to adopt language that mandates practices and programs that reduces reading instruction down to a set of skills. See Dr. Elena Aydarova's research (https://elenaaydarova.com/); she follows the money invested in the SoR movement and connects the dots between the different power players.
- The quest for an exclusive approach to teaching readers reflects a desire for a simple solution, which also explains why the SoR movement took such as strong hold. In their excellent resource Learning to Improve, Anthony Bryk and colleagues refer to this desire as "solutionitis". What does work: investing in teacher quality through time, resources, and opportunities to support professional collaboration around shared problems of practice. Linda Darling-Hammond refers to this process as developing "adaptive expertise".
Perhaps an even bigger reason the ‘movement’ took hold was the incapacity of so many teachers to turn out students who can read. Just a thought.
I agree that reading outcomes need to improve, and that a primary driver for that improvement is an increase in high-quality instruction.
Where I push back here is when you state that the reason for these outcomes is "the incapacity of so many teachers to turn out students who can read". There are too many causal factors to assign the blame to teachers.
For example, in my state of Wisconsin, I ran a correlation between reading scores on last year's state test and students living in financial hardship. 63% of the influence on reading achievement scores can be attributed to whether or not a student qualifies for free or reduced lunch.
We need to be mindful of the simple stories we create due to not examining additional factors that are leading to the outcomes we are seeing. It's what leads to solutionitis and passing mandates that don't address the root causes.
Quite. The incapacity of so many teachers to teach effectively is historically / ideologically explicable, and systemic.
I’m looking at kids who can read, and those who can’t … and the ways they’ve been taught. I don’t need to consult a library of references.
About 25 years after I started teaching (1997?), I had the fortuitous privilege of participating in a series of workshops taught by Judith Hochman. After implementing the strategies I learned for teaching sentence writing, my elementary level SPED students improved significantly. After years of failure, I had finally been able to help students be better writers; I no longer had to feel like a failure. A few years later, I had the opportunity to conduct staff development and I chose to focus on the sentence writing strategies I had learned. Unfortunately, none of the teachers I worked with recognized its value and never followed through. Fortunately, many years later, as an assistant professor, I taught these strategies to my pre-service graduate students with the help of "The Writing Revolution" text.
Steven, I’d be interested in your take on the use of SRSD with the Writing Revolution/Judith Hochman approach as discussed by Karen Harris who suggests there are benefits in using both and.
“… the data, there are no experimental studies of the writing revolution. There are no controlled studies….There's a lot of good stuff in the writing revolution, but it is not teaching students how to write effectively across genres.”
https://podcasts.apple.com/au/podcast/education-research-reading-room/id1200068608?i=1000651365009
Turning Broccoli Into Ice Cream Sundaes: Self-Regulated Strategy Development for Persuasive Writing Using Informational Text
August 2014
DOI:10.1598/0353.
The data and information you need for kids who struggle is more in the levels of language as well as the cognitive profiles research of Troia. Kids who struggle are going to need a lot of work on transcription, spelling and sentence level work before heavy emphasis is laid on larger text. While there might not be direct evidence from the Writing Revolution *itself*, I feel these SRSD adjacent people are doing a disservice to how much work *some* kids need at these levels. There is a lot more to all of this, but even Grahm himself said a student should be pretty proficient at the sentence level before engaging with SRSD strategies. That is straight from the horses mouth.
FBV108, do you have a reference for where Graham said a student should be proficient at the sentence level before engaging with SRSD strategies?
[AI Overview Steve Graham did not state that students must be proficient at the sentence level before engaging with SRSD. In fact, his research contradicts this sequence, finding no evidence to support starting with small parts of writing and moving up. Instead, he recommends embedding sentence-level work within the revising stage of the writing process.]
Thanks for the question and helping me clarify this. I double checked, and the actual reference was to a co-authored text for which Grahm was only one author on (the Powerful Writing Strategies book), but I originally came across this in Shapiro's Academic Skills Problems:
"It is important to consider transcription interventions first because transcription represents a potentially significant limiting factor of writing output and quality (Limpo et al., 2017). Students will be significantly constrained in drafting high-quality written content if they are severely limited in their ability to get words onto the page. Even among students in the seventh and eighth grades, Limpo et al. (2017) found that better handwriting fluency was associated with stronger writing planning, which in turn was predictive of overall writing quality. In fact, developers of SRSD recommend that students should be able to write at simple sentences (e.g., “The dog ran”) before starting SRSD (Harris et al., 2008). Transcription need not be perfect; if students can only write just very short and simple sentences, they can start interventions like SRSD while transcription is addressed. But if students have severe difficulties with handwriting and spelling that impair theirability to form even simple sentences, then transcription should be addressed first. " (pp.320-321)
There may be another quote in there that adds more details, but to go to the direct source it says, from the introductory material to Powerful Writing Strategies (p.7):
"SRSD is meant to supplement, not supplant, the larger writing curriculum. It does not represent a complete writing program. Although writing and self-regulation strategies are essential to writing development, they are not the only abilities students must develop to improve their writing skills. Furthermore, to profit from SRSD instruction, young students must be able to write complete sentences. It is therefore inappropriate to ask students to engage in these writing and self-regulation strategies before they can
write sentences. For young students, a simple sentence (e.g., The dog ran.) is fine. Writing skills such as sentence construction and usage, spelling, and handwriting must also be developed. SRSD, while not a panacea, is an important tool for teachers' repertoires—one that makes a difference"
The Shapiro quote actually understates what is in there and I may have overstated, but we are in this zone. I work with kids with disabilities and have long grappled with the advice from modern proponents of this method and it is clear that some (or most of the kids I work with) should be focusing on heavy foundations like the above. Transcription includes not only handwriting but spelling and basic punction for sentences. For impacted kids that can easily take up to a year or two to get on the line and be fluent enough to start to generalize to larger form writing. This would also extend to kinder and possibly first grade in many schools, unlike the push you see in the modern SRSD iterations that leave kids I work with, and others, in the dust. In some settings with less careful or trained teachers, I would guess that it could potentially create complex problems for both classroom teachers and interventionists imho as well.
Further, Troia et al. (2022) that I referenced, identified that transcription and spelling skills are empirically linked to overall writing proficiency, and struggling writers form a distinct, identifiable profile. They found about 39% of writers in the 4th and 5th graders studied were in this camp and needed definite work on transcription skills. Pairing this with the above, it is hard to say to what extent these even older kids needed to solidfy sentence level skills, but you can see these concerns cover more than just young children and the lack of kids moving up to middle school never punctuating sentences is likely just as much to do with our instruction as it does with any other modern excuses.
While SRSD has an evidence base, specifically for whom and when needs a considerable amount more scrutiny imho. I am sure they are getting results over "busines as usual", but if you look at the other groups in Troia's study there are 40% who primarily need strategy instruction, which could be enough to bump up plenty of effect sizes across many schools and districts. Along with research, please be extremely careful with AI, particularly Gemini/Google and Chat GPT. They are only of so much use if you don't know how to prompt it well, but it will give you what you want to hear with great ease.
References:
Harris, K. R., Graham, S., Mason, L. H., & Friedlander, B. (2008). Powerful writing strategies for all students. Paul H. Brookes.
Shapiro, E. S., Clemens, N. H., & Shapiro, E. S. (2022). Academic skills problems: Direct assessment and intervention (5th ed.). Guilford Press.
Troia, G. A., Wang, H., & Lawrence, F. R. (2022). Latent profiles of writing-related skills, knowledge, and motivation for elementary students and their relations to writing performance across multiple genres. Contemporary Educational Psychology, 71, Article 102100. https://doi.org/10.1016/j.cedpsych.2022.102100
I didn’t rely on the AI quote and had already quoted and referenced Dr Graham in earlier comments, as had others. Your unreferenced claim contradicted those referenced quotes.
I agree with Dr. Leslie Laud (who has also commented here) that it helps to have a variety of teaching foci occurring in parallel, rather doing “ a lot of work on transcription, spelling and sentence level work before heavy emphasis is laid on larger text”.
Teaching of text structure can start before students write sentences, where kindergarten students are using pictures and invented spellings as the foundational skills of spelling, decoding and sentence writing is taking place. You can see the details in this presentation given at the 2024 PaTTAN Literacy Symposium website:
https://youtu.be/NY3LX84Hx-E?si=9ex50wdO6BTjE1Yg
I'm aware of all these opinions. You can do anything you want with kids, but give me some more robust and well designed studies with the populations I tend to work with and I'd be happy to read through it to see if there is any actual weight here. Otherwise the quotes above are crystal clear and in line with other educational, intervention, and cognitive literature. Understanding how to put it in the context of Grahm's other quotes and a deeper read of what the SRSD or adjunct literature suggests is important instead of claiming any of this settled science at the foundational levels. It is not.
"The first thing to consider, which is common to all of these traditions [SRSD and writing revolution] is to privilege the sentence as the unit of instruction."
I don't think this is exactly true for SRSD. Do you have a reference? Or a reference for any peer reviewed research which states this?
Matt Writer, that was also my impression, although there may have been a shift towards doing sentences and paragraphs in parallel with ThinkSRSD now Releasing Writers.
Interesting comments here from Steve Graham-
https://www.ozlitteacher.com.au/blog/writingresearch
A quick search supports this:
AI Overview
In Dr. Karen Harris and Dr. Steve Graham’s Self-Regulated Strategy Development (SRSD) framework, paragraph-level planning (the overall text structure) takes instructional priority over sentence-level grammar.Rather than waiting for students to achieve sentence mastery before they write, SRSD teaches them to brainstorm and build whole paragraphs first using mnemonic organizers like POW (Pull apart the prompt, Organize your notes, Write) and TREE (Topic sentence, Reasons, Explanations, Ending).The Sentence vs. Paragraph PriorityThe Cognitive Approach: SRSD research shows that tackling paragraph structure and higher-order planning first actually builds a scaffolding that improves sentence writing and vocabulary. The paragraph gives the student context and purpose for their sentences.Sentence-Level Work is Integrated: When sentence structure is taught, it is integrated within the paragraph-writing process (such as using transition words) so that foundational mechanics and text composition develop together.Independence and Fluency: Students are equipped with cognitive strategies and self-regulation procedures (like self-talk and goal-setting). This prevents the "bottlenecking" that happens when a student tries to perfect an isolated sentence before knowing the flow of their argument.
This is interesting and confirms my understanding of Harris and Graham as well as my experiences teaching writing to K-1 students. Thank you!
Agreed, this is an oversight in this blog. The research suggests that sentence level skills load the lowest on writing quality. They are better incorporated at the lower end for kids who need a lot of work with the word and sub word level, or at the upper level for kids who can jump right into full text production. Sentence level work is more adjunct in a way.
Lively discussion, and so good to see this much engagement with the actual research.
I have been tracking the writing research closely for over three decades now, and contributing to it for about twenty years. I come at this as a classroom teacher who uses the research every day with students, so my perspective, and my views, come from that angle, and they keep evolving.
Yes, as several people here have said, teaching text structure is the most well researched starting point we have, paired with teaching the writing process step by step: picking ideas, planning, writing, revising.
One reason sentence level work gets less attention may simply be that research is often short term and costly to run. Sentence level growth takes longer to show up. Both matter, though. (As an aside, our field really struggles with holding both/and two things at once. Teach text structure AND sentence, both, and at once from the start.)
My takeaway is that if we start by teaching the writing process and text structure (something like Topic, Information, Details, End), students see fast gains. That matters a great deal, especially early on. But alongside that, we need to teach sentence writing too. Sentence expansion and sentence combining both have real evidence behind them.
We should also be teaching students to strengthen every level of language, including word choice. A planner like TIDE-L (Topic, Information, Details, End, Language/vocabulary) can help students hold all of that together.
Here is the point I care most about. I would ask that we call this body of work EBP-W (evidence based practices for writing) rather than naming it after one branded, commercialized approach, whether that is The Writing Revolution or SRSD. SRSD itself goes back much further than the name suggests, and it is built directly on earlier researchers who get erased once a single brand becomes the shorthand for the whole field. Deshler's six step model is where SRSD's six stages come from. Englert's POWER is where SRSD's POW comes from. I say this as someone who misunderstood that history myself for a long time, and contributed to erasing the people who came before. Englert's Cognitive Strategy Instruction, Olson's Cognitive Strategies Toolkit, and so many others did foundational work that gets lost once the science of writing gets named after one approach by only using SRSD rather than EBP-W for the wider based that includes it and so many others with similar findings.
On a hopeful note, it is wonderful to see research going back to the 1970s finally entering these conversations. Let's keep the focus on the findings, EBP-W, rather than the researcher brand names. From there, text structure is the MVP, but sentence level instruction can and should be taught right alongside it.
I like the idea of EBP-W.
EBP-Reading + EBP-Writing + EBP-Oral language EBP-Learning rather than brands and gurus sounds good.
Considering the overlays of all education, I think just talking about doing "evidence based practice" period, is key. These movements and brands are the death toll.
"Graham's meta-analytic evidence converges on the same point from the other direction, with sentence combining showing reliable effects on writing quality across grades and learner profiles."
While Graham's support for sentence combining is clear, there's some confusion about his recommendation that we teach this skill within the context of a piece of writing. In his keynote address at last year's Multlit conference, he stated:
"So I understand that The Writing Revolution has gotten a good bit of press in your country [Australia]. That they often cite me as being support for what they do. They're right in a way, as … teaching sentence construction skills makes a difference for kids. However, what they don't have evidence to suggest is that starting with a small piece of writing and moving up--like a sentence, moving up to a paragraph, then moving up to a larger text--makes any difference whatsoever. That's not been tested systematically. What we do know is that when you bring various aspects of writing, when you have multi-component writing instructional programs, that's where we get our effects. So, on the one hand, they could be right, they just don't have evidence to support that.”
He also said that students should be fairly proficient with sentence level skills BEFORE engaging with things like SRSD. Let's not forget that either.
FBV108, as discussed below, it would help if you quoted this with a reference as I don’t believe this statement is accurate. I haven’t read everything Steve Graham has written in his long career so my beliefs may be mistaken so I am happy to be corrected.
I gave full references in the other statement, please refer there. It was in a co-authored book, so partial attribution - but stated in conjunction with Harris and others.
FBV108, your references do not include the quote or meaning of what you have said here.
Thanks for your opinion.
Explicit writing instruction in K-2 is my passion and built on 20 years of building step by step lessons. I wonder if we can connect for a discussion about this in Eugene? I’ll be at the NIFDI conference all week. There IS a science of writing!
I'm going to be up there. I'd love to pick your brain on this topic. Your approach seems to stand in contrast to some of what I see being pushed with some of the strategy instruction that has gained so much traction thanks to social media. I think you can take the research in either direction, but it's a critical topic imho.
Great! You can email me via my website Growing Writers to figure out a good time and place to meet up at the conference.
Great writers come to us through a sliver in the universe. They are never born and certainly not created, but rather emerge.
That said, a great writer will always need to be capable of terrible writing - and perhaps even solid writing.
Just as Monet mastered the classical style, so too much all writers. But if they ever write the way they were taught, they never be great, or even interesting.
Only a gun to the head can make some writing interesting.
When most people talk about writing instruction they mean sentences and paragraphs.
You mentioned that we need to "privilege the sentence as the first unit of construction".
Actually the first unit of construction is the letter, then the word. Then words without any sentence form. A lot of students miss an adequate introduction in K-1 and don't get anywhere near the hours of practice they need just on letters and words.
These are the years (K-2) for which research is missing, because I believe there's still a belief that writing is "emergent" in these grades and will just appear after being exposed to activities. We're no longer teaching whole language reading, but K-1 teachers are doing a lot of what I would consider whole language writing. Either that or trying to expect little kids to meet ridiculous standards.
Reading research skews towards the early grades, whereas writing research skews towards 5th grade and up. Reading has become a science partly because of the studies of those earliest years.
Kindergarten and first grade are filled with activities, but not instruction. I have created what I called The Pyramid of Writing Skills for K-2, a way to ease children into writing without cognitive overload. My writing curriculum is based on this and teachers breathe a sigh of relief to finally have appropriate instruction. I know that SRSD and The Writing Revolution can "work" in the youngest grades (not sure how that goes or if it works for everyone) but I think they would have the most impact when kids have some foundational skills.
Let me know through a comment if you'd like a copy of my article.
Have you looked much at the SRSD literature? Particularly for any of it possibly pertaining to younger grades? I would be curious on your viewpoint here.
Yes, I have. There is evidence that it is successful in second grade and up. It has also been used in K-1 and I have seen anecdotal evidence of success. The question is: who is the strategy benefitting? Kids who are already confident writers? Or all novice writers? Also, are the studies measuring how well kids can use the strategy? Or their writing skills in general? Kids can do anything with a lot of scaffolding, help and guidance. That doesn't necessarily mean they have stored learning in long term memory. I
Indeed, these are the things that keep me up at night.
Yes please Sally. Send away!
I’m in the process of helping my 5 YO grandson to read & write. He’s “tutored” by a lady who relies on my long past experience as a reading tutor (via Phono-Graphix) as she’s mostly been a babysitter because my stubborn son won’t enroll him in a proper homeschooling program (forget public school). Our rural location & his single Dad status are partly why but that’s just part of the problem my grandson faces.
Reading the many replies to this article sadly reminded me a bit of different arguments for religious perspectives. I’m more of the big tent type. I’m also reminded of my training experience in spelling where I was informed that in Shakespeare’s time it was considered creative to spell words differently (even within the same piece) as long as the writers intent was discernible!
I’d like to recommend Justin Deschamps’ recent Substack article: Cultology: The Pen and the Sovereign Mind. I consider it a brilliant piece that is so much more than its parts. We can all learn about how “writing by hand” makes us choose and order our thoughts and words in a meaningful way.
I’ve had to choose my remarks here due to time constraints but I look forward to hearing, reading & writing more as time allows as I’m thankful for everyone’s thoughtful contributions.
Speaking and writing do not convey knowledge anymore. They package it, and only to exploit others.
No one doing actual knowledge work relies on Oracy and literacy anymore.
Only those trying to profit from people who aren’t at the edge.
My traveling days are mostly behind me but I spent many years traveling to conferences of learning that interested me. Fortunately the internet now provides access to the like…kinda like the Buckee’s of learning ;)
I do question your remark about speakers & writers reason for doing so is to steal from people?
Like great songwriters also come to us through a “sliver in the universe.” I do believe!
We teach everyone how to paint. It’s excellent to learn the basics of painting.
We teach everyone to write. Great, everyone should know how to string a sentence.
BUT where once necessary to be a solid, professional grade writer, now it’s not necessary. And this is a great thing!
However, will professional grade writers have more power than others? Definitely before. You needed to be a great writer to be a professional success.
Not anymore. And in fact, there’s only one reason to become a great writer now, just as with the advent of writing, there was only one good reason to become a great speaker: to enchant and confide…or rather: to steal from people.
Oracy and literacy are NO LONGER the dominant medium for knowing: algorithmacy is.
And if you don’t know about algorithmacy, come down to Trinidad this fall: algorithmacy.org
I really appreciate this piece, especially the emphasis on sentence-level instruction and the role of conjunctions as cognitive levers. That is very close to what I see clinically. Children who struggle with writing are often not struggling because they lack ideas. They are struggling because they are being asked to coordinate spelling, syntax, vocabulary, sentence structure, background knowledge, and working memory all at once.
I do think this also raises a question for the reading conversation. If writing cannot be reduced to a single mechanism, I’m not sure reading can either. The public version of the science of reading has often leaned too hard on the idea that reading is primarily the mapping of letters to sounds. But English orthography represents language through meaning, structure, history, and phonology together. That becomes obvious in spelling and writing, but it is also true in reading.
So perhaps part of the reason writing has been left out is that writing makes the system harder to oversimplify. It forces us to look at words, sentences, and meaning together.
Writing and reading can’t be reduced to a single mechanism but this does not mean research can’t be done on multicomponent approaches and also on the components required for effective literacy teaching.
With writing, Drs. Karen Harris and Steve Graham, along with other researchers, have conducted more than 170 studies on SRSD for writing instruction. And they have conducted studies on difficulties with text transcription skills such as handwriting or spelling which may constrain young children’s writing development. It’s not either or, it is yes and.
Reading can’t be reduced to a ‘single mechanism’ and yet it is still possible to conduct research on PA, phonics, knowledge of the alphabetic principle, morphology, fluency and other components separately and when included in multicomponent approaches to teaching.
The newer models of reading (David Share, Universal Combinatorial Model 2025; Nell Duke and Kelly Cartwright, Active View of Reading 2022; Maryanne Wolf POSSuM (2021) are not moving away from teaching phonics to beginning readers. M. Wolf’s POSSUM model (https://www.shankerinstitute.org/sites/default/files/documents/elbow-room-report.pdf)“… the teaching of these multiple processes should not be conceptualized as a checklist to do in isolation, but rather as the component parts that are to be integrated daily. … Phonemes need letters. Phonics needs semantics, syntax, and morpheme knowledge. Words need stories. The reading brain connects all of these processes, and so should our teaching.”
Similarly, with multicomponent approaches in teaching writing, teachers are not moving away from teaching components of writing systematically and explicitly. Researchers are not moving away from studying the components to find out what is necessary (but not sufficient) in writing instruction.
Reductionistic and Holistic Science
https://pmc.ncbi.nlm.nih.gov/articles/PMC3067528/
I agree that this is not either/or. Components can and should be studied, and explicit instruction matters in both reading and writing.
My concern is not with component research itself. My concern is with what happens when one component becomes the public organizing frame for the whole system.
In the reading conversation, that has often meant treating reading as though it is primarily a matter of mapping letters to sounds, with morphology, syntax, semantics, and etymology added later or treated as enrichment.
The quote from Wolf actually points to the distinction I’m trying to make: “Phonics needs semantics, syntax, and morpheme knowledge. Words need stories.” That is not phonics standing alone. That is phonology functioning inside a larger language system.
So yes, we can study components. But instruction still has to help children understand how those components work together.
Writing makes that especially visible because children cannot write well by coordinating only one part of the system. I would argue that reading is the same.
English orthography represents meaning, structure, history, and phonology together, and our teaching should reflect that from the beginning.
This is where you part ways with people like Wolf who (based on research evidence) emphasise the importance of knowledge of the alphabetic principle and phonics at the beginning. [Phonics taught in meaningful morphemes (cat, mat, sat) which come from meaningful stories (Pete the Cat, The Cat in the Hat, and importantly, children’s own stories of their pet cat which they write about using evidence based SRSD approaches)]
I don’t think I part ways with Wolf as much as that framing suggests. When I listen to Wolf present, I hear an integrated model of reading, not a phonics-only model.
I agree that children need alphabetic knowledge. I agree that they need to understand that letters can represent phonemes. My concern is with making that relationship the organizing frame for beginning instruction, especially when English orthography represents meaning, structure, history, and phonology together.
Using words like cat, mat, and sat in meaningful stories is certainly better than isolated drill. But the grouping is still organized by sound pattern, not by shared meaning or structure. That distinction matters.
So the question for me is not whether phonology belongs in early instruction. It does. The question is whether beginning readers should be taught from the start that spelling is a meaningful, structured system, rather than first being taught to expect spelling to work primarily by sound.
That is an interesting hypothesis that could be tested. Write a program for beginning readers so a controlled study can be done comparing your approach with something like UFLI Foundations. Publish and have the research replicated.
I would welcome that research.
At the same time, “build a full program, run a controlled study, publish it, and get replication” is not the entry fee for raising a legitimate theoretical and instructional question.
My claim is not that research should stop at hypothesis. My claim is that the phonics-first frame rests on assumptions about English orthography that deserve scrutiny. If English spelling represents meaning, structure, history, and phonology together, then it is reasonable to question whether beginning instruction should be organized primarily around letter-sound mapping.
A comparative study would be useful. But the linguistic question does not disappear while we wait for one.
What stood out to me is the idea that writing is not simply a way of recording thought but a way of creating it. Reading lets us borrow the conclusions of others, but writing forces us to confront our own half-formed ideas. A vague feeling becomes a sentence, and a sentence becomes something we can question, refine, and understand more clearly.
Yes, Tyler. Reading and writing to learn.
I think I understand an argument until I struggle to write it independently. Writing helps clarify what is ‘in my mind.’
Writing and the extended mind thesis.
Clark and Chalmers used the famous "Otto's Notebook" thought experiment to develop the concept of the extended mind:
Imagine a man named Otto who has Alzheimer's. To help him remember things, he carries a notebook everywhere he writes down important information, effectively using it as his biological memory. According to the extended mind thesis, because Otto relies on the notebook exactly the way a healthy person relies on their brain's memory, the notebook and Otto's brain constitute a single, coupled cognitive system. Taking the notebook away is functionally equivalent to taking away a part of his mind.
I recently learned about Steven Pinker’s “The Sense of Style” that uses insights from cognitive science to improve writing. It may contribute to the “science of writing” (not just writing instruction).