Summary of “Eyes on Reading: What’s Next in the Science of Reading? with Mark Seidenberg and Maryellen MacDonald”

“Eyes on Reading: What’s Next in the Science of Reading? with Mark Seidenberg and Maryellen MacDonald” (presentation from 1 March 2025; video date 9 April, 2025)

This is the most important presentation I’ve seen so far this year, and indeed that I’ve seen in a long time. The perspectives, insights, knowledge, and wisdom packed into this talk were… beyond words. I want to make people watch it; I can’t do that. I know people are time poor, so hopefully the summary below will at least give you the opportunity to engage with the important points discussed in the very long presentation, and encourage you to perhaps someday listen to it in full. Because this information matters; right here and right now.

This is a talk between three giants: Emily Hanford, Mark Seidenberg, and Maryellen MacDonald. I think I have learned more off Seidenberg about literacy over the years that any other researcher, author, or educator, so I never miss a chance to hear him speak because I know it will be time well spent. I was not disappointed. I hadn’t heard Maryellen MacDonald talk before (that I can recall) and I feel like I have been missing something wonderful.

Seidenberg focuses on the science of reading movement and the need to recalibrate it, and MacDonald discusses a missing piece of the science of reading puzzle (as it’s currently being implemented in particular) - comprehension at the sentence level, rather than at the word or passage level.

Here are some key points from the talk, alongside approximate time stamps for where those sections of discussion began:

8:20 Seidenberg discusses the origins of the “science of reading” movement. There had been resistance to the idea of dyslexia as a real condition, and resistance to change more generally. The science of reading movement grew from these concerns, and was spurred on by his book, by Emily’s documentaries, and by the efforts of parents of dyslexic kids. They were successful at creating demand for change.

11:50 “That’s all great, but the approach has to work.” It’s easier to say what’s wrong than to specify what to do instead. It’s expected that the first take at coming up with what to do instead, wouldn’t be ideal and would need revision.

13:55 “There wasn’t a science of reading curriculum already on the shelf. There was however an approach that could be adopted for the purpose.” “That approach is called Structured Literacy.” It was created by the IDA (International Dyslexia Association). It was explicit, systematic, and thorough. There was so much frustration about a lack of response to dyslexia, that the IDA tried to organise a plan for what to do. But it was then scaled up to be taken to the basis of instruction for all – for general education, not just kids with special needs.

16:12 What is Structured Literacy (SL)? The components from science broken down into teachable pieces. SL was developed to help kids with conditions that interfere with reading, and the claim was that all kids need to learn the same things to learn how to read – that part’s OK. But typically developing children might not need all the instruction involved in SL. The claim was that’s fine too, that it won’t be harmful - that there was no danger of too much of a good thing. Good for dyslexics = good for everyone (was the claim).

18:40 SL works by adding one piece at a time to build up the structure, like trying to learn how to play a piano piece by learning one note each day. “If you treat everyone like they might be dyslexic, you get an intensive, slow, incremental approach to instruction.”

20:00 Phonemes are referred to a lot now – diphthongs and fricatives, 100+ phonics rules, syllable types etc. Teachers learn about the components and try to teach it all in a limited amount of time. “The science of reading has an over-teaching problem.” “The opportunity costs of doing all this instruction, are huge. Instruction time in schools is limited.” Other important activities get sidelined. Some explicit instruction is needed for beginning readers, absolutely; Balanced Literacy and Whole Language didn’t do an adequate job of this. “Science of reading has overshot the target because most kids aren’t dyslexic.”

23:25 “Learning fast and slow.” Explicit instruction is part of the slow system – requiring conscious awareness, intention, and intent. The fast system is also called statistical learning or implicit learning. Components of reading are mainly learned through the implicit part but Structured Literacy tries to teach it all explicitly. The thing that really allows language learning to take off is statistical learning.

26:27 Summary: The science of reading is committed to explicit instruction to make sure everyone succeeds; it isn’t realistic and takes time away from other learning opportunities. “Is it efficient? Definitely not.”

27:16 What can we do about this? Dial it back. Not start over but recalibrate what we’re doing. The function of explicit instruction is to set up implicit learning. The aim is to help the kid crack the code, not teach the whole code itself. Teaching the whole code is costly and not necessary, “Just because something can be taught doesn’t mean it should be.”

The goal is to get the child to escape velocity.

29:50 Maryellen MacDonald starts her talk on an overlooked aspect of reading instruction that could and should be taught, that sits at an intermediate level between whole passages and basic skills: sentence comprehension. There are lots of cognitive mental processes working really hard when trying to comprehend a sentence. Sentence comprehension is understudied in children; it needs more attention.

33:26 What is sentence comprehension? Unconscious decisions are being made at pretty much every single word. People do this work by playing the odds – they lean towards the most likely interpretation of the words and words sequences based on what they know and on the context. To play the odds, you’ve got to know the odds, and this is based on the way language works and the way books work. The odds come from this thing called “book language;” from the vocabulary, sentence structures, and characteristic situations in books etc. These are not the same odds as in speech.

41:04 Why are speech and book language odds so different? Speech is between people who know each other and who share prior knowledge and joint attention. We don’t have to speak about what everyone knows – they can see the situation in front of them. Speech to kids doesn’t have a lot of information transfer. Speech to kids has lots of instructions, guidance, encouragement, and questions. Lots of pronouns.

By contrast, books don’t share detailed prior knowledge, there is no joint attention (in the sense we have in speech), it’s remote, people don’t know each other, so the characteristics of book language are really different. It’s more words, different words, and a different way of using language. Books have many more nouns, emotion words, and descriptions. Books also contain different sentence structures, longer sentences with multiple clauses, and more complex sentences.

Oral language is important but not sufficient for learning all the stuff in book language.

All these differences create difficulties in comprehension.

49:40 How do we learn book language patterns? Someone reads you picture books. Then you got some crucial explicit instruction, then you read decodables, then you read more books. You get lots of practice reading. The value of experience – the fast system of learning.

51:20 Can we replicate our own success in the current era? Kids read a lot less than they used to. Reading outside the class is declining, and likely to be declining in class too.

52:44 What can we do about this? Four things: (1) More shared book reading in and outside of school. Reading to kids has been declining. (2) More reading practice in schools. Long exposure to decodable texts may delay kids’ exposure to book language and the book language statistics that they’re going to need to know. We could build book language into decodable texts (she shares an example of this being done) - we could be thinking about making sure decodables achieve both aims. (3) Audio books. They don’t help with decoding but could be a source of what’s being squeezed out of kids’ lives, and enable exposure to book language. (4) Watch documentaries. Narrated documentaries have the same qualities as book language. Maybe replace some child media with family-friendly narrated documentaries.

1:00:24 Summary and final thoughts. Poverty, language delay etc are also important.

The key is to “get in, get it done, move on.” Don’t take so long on the on-ramp to reading, it leads to problems down the road. Let’s not screw this opportunity up.

1:03:13 Questions: What might we change in policy? Understanding sentences should be in science of reading policy. Kids can read words but not understand sentences; we see this in the “4th grade slump” and in “good decoders, poor comprehenders”. Book language may be the key to addressing this.

1:06:02 Seidenberg briefly on whether we can teach comprehension – the sentence stuff is teachable.

1:07:40 We’re all in trouble, literacy is in trouble, we know this (for example, because of the technology age we find ourselves in). Reading outcomes depend on a lot of factors, we know this too, but we can at least try to get right what we do in schools – that bit we can control.

1:10:10 What have they learned from teachers? Seidenberg responds: It’s eye-opening to visit the actual classrooms. Can see misinterpretation of the science by well-meaning people. Also saw what effective teachers do (shares an example of two teachers working together in a classroom).

1:13:43 “Most words that are common in English are ambiguous”. The more common it is, the more ambiguous it is. Hanford: We need to ask not just what happened in that passage, but what is happening in this sentence.

1:18:32 Audience question: Some children have been “Orton-Gillinghamified for years” and still can’t read much. Dyslexics might never get 100% mastery on certain phonics patterns. Are we over-teaching not just in the general ed setting but also for dyslexics?

Seidenberg’s response: Dyslexia is complex and there are a range of profiles. For some dyslexic kids, they need to slow it down and take more time on different parts. For a lot of kids that is going to be successful, but not for everyone. “The challenge in dyslexia is to know when to stop.” “There are some kids for whom more of the same is just not working, even if it’s done extremely effectively by a highly trained professionals; there’s a point at which you realise they need a workaround. And the challenge is to figure out when to change paths.” Given you can’t assume they’ve all been taught effectively, the thing to try is to improve the basic instruction – not exactly like Structured Literacy but sort of like that. Try that first. For some you have to try something else. Those are difficult choices that professionals have to make.

1:22:52 Seidenberg: “Science of reading” is now the new dogma, and “there are people with vested interests in maintaining some of the problematic things.” “That’s going to take some effort to undo, and they’re not going to undo it just because a few people say they should. It’s going to take work.”

1:27:10 Will technology get us to the point where none of this will be necessary? Hard to say. Experts disagree. Reading more can also improve your speaking skills. Hanford speaks to AI being seen as an add-on rather than a replacement.

1:33:26 Question: Does the Department of Education hear you (hear the concerns of the speakers)? Seidenberg: No, our audience is very limiting. Pointing this stuff out after science of reading got off the ground is really perceived as harmful by some people. MacDonald: Teachers’ deep dissatisfaction with all the things they’re being asked to do is potentially an agent for change. Change happens slowly; this might take years.

1:36:48 Emily Hanford: The pendulum itself shifts; we’re going and back and forth but between different things now than before. The difference is now there is more teacher awareness and knowledge. We need to keep talking about his, writing about this etc.

Don’t give up.

(The above summary was originally shared in the Effective Education Practices NZ Facebook group: https://www.facebook.com/groups/effectiveeducationnz/posts/1012609473671047/ )