Over on his regularly informative ‘stack, Professor Timothy Shanahan, posted a note entitled “Teaching Blending: Are We Overdoing It?” In the post he provided a sensible response to a teacher’s inquiry about blending: “I was just told that the science of reading says that we shouldn’t teach blending or that we teach too much blending and should dial it back. Do you agree with that?”
Professor Shanahan explained that he was certain teaching blending was important in helping children learn the alphabetic principle. He also explained that, “like anything, it would be possible to overdo it.” He recommended moving quickly into larger units of sound such as rhymes.
Professor Mark Shinn dropped a comment on Professor Shanahan’s post in which he suggested that the solution to the problem of teaching blending was to adopt “a well-designed, field-test curriculum.” Professor Shanahan replied (in two parts) that teachers who have such curricula “would benefit from having as clear an idea of why those components are there” and that other teachers may not have not be able to use such would also benefit from having information about blending.

Musings
I agree with Professor Shinn’s comment that it would be better to have educators employ appropriately developed and tested curricula than to continue asking teachers to Roll-Your-Own. There actually is (as Professor Shanahan’s full post indicated) a lot to teaching blending.
Probably the foremost example of a commercially available curriculum that has comprehensive blending instruction is Reading Mastery. When its predecessor (the often-reviled DISTAR Reading) premiered in the late 1960s, it was the only (or one of a very few) curriculum that expressly taught blending at all. Now, thank goodness, there are other curricula that include blending.
Sure, it would be nice to provide professional development that helps teachers understand the why’s-and-wherefores of teaching blending, but a curriculum that has those features programmed into lessons would be a heckuva lot easier on them. One could, for sure, devote a slide or two to those whys-and-wherefores during professional development about using an effective curriculum with fidelity.
And, certainly, some teachers may be teaching in local education agencies or buildings where they do not have (or are even forbidden to use) such curricula. It would be nice, indeed for them to know what to do to modify the shoddy curricula they are requited to use. It would be great if they could supplement that instruction so that what and how they were teaching approximated the programming in carefully developed and tested curricula. But, it’s almost certain that what those teachers who had been schooled in The Science would do would only be an approximation.
To illustrate, sticking with blending for right now, our professional development would need to explain the ideas about connected phonation that were discussed in the Professor Shanahan’s post. I would be encouraged if those teachers charged with instructing children in acquiring the alphabet principle had that knowledge. But I’d be even more encouraged if Roll-Your-Own teachers learned how to design a sequence of lessons that provided
The appropriate number of practice opportunities (massed, spaced, distributed, etc.) to get learners (especially naïve learners, those who came to instruction with few clues about the phonological sound stream) to the point that “they can do it reasonably well.”
The scaffolding (and gradually removed it across lessons) to make that instruction work for naïve learners (those who had little competence in “reading” beyond holding a book right side up);
The plan to avoid stop or plosive consonants (/b/, /d/, /p/) in early blending, to use easily blendable sounds (“continuous” sounds such /f/, /m/, /r/) initially, and to introduce practice with blending words that end with plosives before moving to words that begin with plosives (Engelmann, 1988, 1969);
Appropriate enunciation that minimizes the schwa in letter sounds (we do not want learners saying “muh-aaa-tuh”); this is important both for students and for teachers when they model blending behavior;
Oh, yes...right! And this spills into other issues, like the order in which instruction introduces phoneme-grapheme correspondences; we’ll need to help Roll-Your-Own teachers to sequence the introduction of symbols, too, in harmony with the blending activities;
Oh, yes! There is also the need to coordinate the blending lessons with instruction in its sibling, segmenting, too. (Professor Shanahan mentioned the value of “analytic phonics” in his post). How much should teachers know about segmenting? What about teaching segmenting?
We wouldn’t want the Roll-Your-Own teachers to design their lessons as isolated skills, programming segmenting as skill of the week for three weeks, follow by blending as skill of the week, followed by letter-sound correspondences…and then, once they’re doing with rolling through early decoding instruction, beginning spelling instruction….
Yikes! There’s a lot to designing blending instruction, let alone teaching blending.
As Professor Shanahan noted, “[p]roblems with blending may arise for children when they’re first learning to decode.” So teaching blending is only one part of teaching kids to decode fluently which is only one part of teaching to them to read. It has to be only a part of helping teachers “understand The Science.” Put together, it adds up to a lot that we shall need to cover in our program of professional development.
And, yes, teachers knowing about beneficial blending instruction would be helpful to those teachers who serve on curriculum selection committees. In that regard, because teachers’ voices are likely to be discounted on those committees, it is probably especially important that administrators who hold the purse strings in purchasing curricula know about the intricacies of blending—and actually how to teach blending.1 So, there’s another entire PD program.
Just one more point: Professor Shanahan’s correspondent’s question about overdoing blending instruction is quite sensible in light of the history of reading instruction. It was not that long ago that Helen Yopp published an influential article about phonological awareness. In it she reported about the reliability and validity of 10 test of phonemic awareness. Although about two types of tests (spoiler alert: blending and segmenting) were relevant to reading instruction, there was an explosion of recommendations about teaching kids all 10 types of “skills.” Phonemic awareness was everywhere! Educators for damn sure overdid it. So, let’s presume that Professor Shanahan’s questioner’s question was thinking about the excessive emphasis on phonemic awareness following Yopp’s article. But, our goal is not to teach some broad Blending Skill. We do not to teach substitution (“What word would you have if you changed the /f/ in calf to a /v/?”), deletion (“What word would you have if you said stand without the /t/?”), and etc.. If our goal is to teach decoding (and I think that’s an appropriate goal), we only need to teach the close-in skills of segmenting and blending.
References
Engelmann, S.2 (1967). Teaching reading to children with low MAs. Education and Training of the Mentally Retarded, 2(4), 193-201.
Engelmann, S. (1969). Preventing failure in the primary grades. SRA.
Yopp, H. K. (1988). The validity and reliability of phonemic awareness tests. Reading Research Quarterly, 23(2), 159-177. https://www.jstor.org/stable/747800

Footnote
By “how to teach,” I mean that they really ought to have spent a few weeks-months implementing a well-designed curriculum that has systematic and explicit blending instruction.
This publication appeared in a section of ETMR that was edited by Laura J. Jordon. Successfully searching for the article may be improved if one to puts Ms. Jordon’s name into the search terms. There is also a typescript of the article that Engelmann (or someone) filed with the US educational documents repository, ERIC. It is available for free at https://files.eric.ed.gov/fulltext/ED015020.pdf


I am a huge fan of blending. However, once a student can blend taught sounds, move them on to reading words as a whole for meaning, with fluency and in meaningful text. Morphological units can be easily incorporated into blending instruction. Blending is the tool to unlock the code not the end goal. I’ve seen teachers have students segment and blend too many words once the student has mastered a taught sequence. Always teach blending along with segmenting.
Regarding teaching onset-rime instruction, I read the Shanahan article that you referenced, and agree with him when he stated “I have no doubt that if kids in a synthetic phonics program had trouble with blending, then the teaching of some larger word units (rimes, syllables, multi-letter morphemes) could help.” I have promoted and witnessed success for many students using this approach.
One size doesn’t fit all.