Posted April 14, 2022
by Dr. Claude Goldenberg, Nomellini & Olivier Professor of Education, emeritus, Stanford University
The newest front in the never-ending wars and squabbles over how to teach reading involves English language learners (EL), students learning to read (and write) in English while simultaneously learning to speak and understand it.
In the wake of renewed prominence of research supporting an early focus on phonics and decoding (aka the “science of reading”), EL advocates around the country have been sounding alarms.
For example, a recent EdSource article led with the following: “As California launches a new literacy campaign, some advocates worry that for English learners, a focus on sounding out words will come at the expense of learning the meaning of words.”
In Colorado: “Colorado’s emphasis on phonics in reading could hurt English language learners, advocates say.”
The concern is understandable. Learning to decode words if you don’t understand what they mean seems pointless. In addition, solely focusing on phonics during reading instruction provides a meager literacy diet.
But the concern misses the larger point that learning to read requires integrating written language with oral language. This integration is impossible if you can’t recognize—read—the words on the page accurately and automatically followed by verifying the words’ accuracy and meaning.
Word recognition is foundational to reading. Knowing word meanings is important for confirmation. Among skilled readers, the process works automatically, effortlessly, and efficiently. Beginning and early readers need varying degrees of support, scaffolding, and direct teaching to acquire these skills. Some require very little and appear to “catch on” easily; others need a great deal of direct teaching; most are somewhere in between.
This larger point applies to readers in general, not just ELs. I’ll use a metaphor to try to illustrate.
Imagine two roads that begin separately and eventually converge at some desired destination. One road is, let’s say, the “road to the code,” where you learn how the sounds of a spoken language are represented in writing. Progress along this road is relatively straightforward, although it has its challenges. It involves developing phonological awareness, knowledge of letters, letter combinations, and corresponding sounds, and how to use that knowledge to identify written words, often referred to as “phonics” or “decoding.”
I’ll call the other road the “road to understanding.” Here’s where you learn how to make sense of the world as you experience it either directly or through someone’s oral retelling or through electronic media or in written texts. To proceed successfully along this road, you need to learn aspects of language that carry meaning (e.g., vocabulary, morphology, syntax), background knowledge and knowledge of the world, and comprehension and thinking skills and strategies.
Eventually, the roads must converge at their destination: full and competent literacy. The roads start out separately because they proceed in different parts of the brain. Metaphorically, they originate in different geographic locations and traverse distinct landscapes.
Successful travel through the code road in an alphabetic language essentially requires knowing and being able to apply the sound-symbol mapping system of the language’s orthography. Letters are human inventions developed to represent the sounds of speech. There is nothing inherently meaningful about letters, although since their names typically contain the sounds they represent, learning letter names helps connect them to those sounds. The whole idea of learning to read is to make those letters and sounds meaningful by connecting them to meaningful oral language.
Here’s a critical point: Research has shown that the best way to help learners master the road to the code is through direct and explicit instruction in what is sometimes called the “alphabetic principle,” that is, the basic understanding that printed letters represent speech sounds and using that understanding (aka phonics and decoding) to read words. (The issues are somewhat different in nonalphabetic writing systems.)
Mastering the code road is no guarantee of full literacy. Learners must also successfully travel along the path to understanding. This path can be complex and winding. It typically commences even before the road code appears. Explicit teaching contributes to progress along the path to understanding, but so do—perhaps even to a greater extent—direct experiences with the world, interactions with others, and planned and unplanned events and activities.
The path to understanding matters immensely if you are to get to the ultimate destination of literacy. If you can read accurately but without understanding, full literacy is impossible.
Similarly, if you don’t travel well and far on the code road—e.g., someone tells you to take a shortcut by guessing at words—your path to uniting with the road to understanding will be truncated. You won’t get to your destination of full literacy. Why? Because if you don’t successfully traverse the road to the code, your world will be limited to what doesn’t require accurate and fluent reading of written language.
The necessity of traveling along two distinct but ultimately converging paths is as true for English learners as it is for learners who already speak English. This has been amply demonstrated by neurocognitive and neurolinguistic studies involving many different first and second languages and learners around the world.
For English learners or, more generally, students learning to read in a language they are simultaneously learning to speak and understand, there is an important qualification: Care must be taken that words (and text) they are being taught to read are meaningful to them.
We can typically assume that students learning to read in a language they already know already understand words used in beginning and early reading instruction, words such as run, dog, me, and can. We can’t make that assumption with second-language literacy learners. As we take them along the code road, we must also make sure they’re making corresponding progress on the road to understanding the words the code road is helping them learn to read.
This has been well documented in studies by Sharon Vaughn, Linnea Ehri, and their colleagues. These researchers have found that students learning to read in English while gaining English proficiency, but experiencing early reading difficulties, make greater progress when instruction is focused on helping them progress along both roads.
Children were taught the foundations of word recognition directly and explicitly. They were also taught the meaning of the words they were being taught to read and to use those meanings to cross-check what they read using their developing word reading skills.
Note that I’m referring here to English learners learning to read in English. If they are in a bilingual program and learning to read in their home language, learning to read while learning the language is not an issue. While learning to read in your home language might be preferable for any number of reasons, the fact is a large majority of English learners learn to read in English as they are learning English.
The implication is straightforward: English learners need to be helped—taught—to understand the words and text they are being taught to read. But they need to learn, explicitly and directly, the foundations of word recognition to accurately read the words. Both are essential if these students are to become fully, functionally, and successfully literate.
Claude Goldenberg, Ph.D. is Nomellini & Olivier Professor of Education, emeritus, Stanford University. He is author of Successful school change: Creating settings to improve teaching and learning (Teachers College); co-author with Rhoda Coleman of Promoting academic achievement among English Learners: A guide to the research (Corwin); and co-editor with Aydin Durgunoglu of Language and literacy development in bilingual settings (Guilford). He has published and been on the editorial boards of various literacy and education academic and professional journals. Previous projects focused on improving literacy achievement among English Learners in elementary and middle school, language and literacy development among Mexican children in Mexico, and a randomized control trial of an early literacy intervention in Rwanda. Current projects include consulting for the US Department of Justice on English Learner issues and chairing a research advisory panel on early childhood education for Arizona’s First Things First.
Claude Goldenberg recommends “direct and explicit instruction” in phonics and decoding, as the “best way” to master the code. This “instructional path” involves conscious learning of the rules of phonics. There is good evidence, however, that the instructional path cannot do the entire job: Scholars have not discovered all the rules of phonics, and even among rules that have been described, some are extremely complex with numerous exceptions 1: only the more straight-forward phonics rules can be consciously learned.
How do we acquire the complex rules that are not taught? It takes place on the other path described by Goldenberg, “direct experiences with the world.”
This path is not only the “road to understanding” but also promotes “acquisition,” subconscious absorption of the rules of phonics. Most likely, the kind of “direct experience” that does this is reading, reading comprehensible and interesting books and other kinds of texts.
Reading is also an efficient and pleasant way of developing competence in grammar, spelling, and vocabulary 2 and builds knowledge in a number of areas: those who read more, for example, know more about history, science, economics, and practical matters.3 This helps make input more comprehensible and results in more acquisition of language, literacy, and knowledge.
(1) Clymer, T. (1963/1966). The utility of phonics generalizations in the primary grades. The Reading Teacher, 16/50, 252-258/182-185.
Johnson, F. (2001). The utility of phonics generalizations: Let’s take another look at Clymer’s conclusions. The Reading Teacher, 55, 132-143.
Smith, F. 2004. Understanding Reading. L. Erlbaum Associates. (Sixth edition).
(2) Krashen, S. (2004). The Power of Reading. Libraries Unlimited.
(3) Stanovich, K. and Cunningham, A. 1993. Where does knowledge come from? Journal of Educational Psychology 85, 2: 211-229.
Thanks for your comment. What you suggest is also supported—albeit possibly (or maybe not) with different mechanisms, pathways, etc.—by Share’s “Self-teaching hypothesis.” See https://fivefromfive.com.au/the-self-teaching-hypothesis/; or from the horse’s mouth: Share D. L. (1995). Phonological recoding and self-teaching: Sine qua non of reading acquisition. Cognition, 55, 151–218. Then see a more recent updating/reworking, aka the phonological-decoding self-teaching theory, in Ziegler J. C., Perry C., Zorzi M. (2014). Modelling reading development through phonological decoding and self-teaching: Implications for dyslexia. Philosophical Transactions of the Royal Society B: Biological Sciences, 369(1634), Article 20120397. doi:10.1098/rstb.2012.0397.
Then there’s the “computational model” of reading acquisition, which further builds on Share’s model. See Ziegler et al. (2020). Learning to Read and Dyslexia: From Theory to Intervention Through Personalized Computational Models. Curr Dir Psychol Sci. 2020 Jun; 29(3): 293–300. doi: 10.1177/0963721420915873. (Available at https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7324076/#bibr37-0963721420915873.)
I do think it’s clear, as I believe your comment implicitly acknowledges (“the instructional path cannot do the entire job”), that direct instruction helps learners get a “toe-hold” on the foundational knowledge and skills required to become literate, but that is by no means the whole show, even for word recognition/reading itself. And certainly not for the many things required to get beyond basic word reading skills.
BUT… we can’t underestimate the extent to which for many kids (and for a variety of different and sometimes multiple reasons) getting that foundational toe-hold is no easy matter. I hope you’d agree that meeting their needs is at least as important as meeting the needs of kids who learn to read with relative ease. I’m a huge fan of “reading, reading comprehensible and interesting books and other kinds of texts.” But as far as reading development goes, these are of limited utility unless you can actually *read* them. Would you agree? Self-teaching and computational models presume some threshold level of reading knowledge/skill before reading development can take off with less, minimal, or even no direct teaching.
I’d be interested in your further thoughts.
I don’t have a bibliography to accompany my comment, but over the years I have known many ELL students who learn to understand English, develop background knowledge, think critically, and persuasively use evidence to argue their point. But they still struggle to read. I can’t think of one who became a strong reader in English and failed to develop comprehension skills. Both roads are necessary, but (from what I see) the road to the code is not offered to most ELLs, while ESL programs often provide good roads to understanding. Is my experience unusual?
How to make sure less advanced readers reach the stage where reading can be done and students can continue to progress, with minimal or no direct teaching
The problem is that we can only teach the straight-forward phonics rules and the simplest grammar. After that, we need to acquire from input.
The solution: Provide comprehensible yet highly interesting reading (and listening) material.
My hypothesis is that if the input is interesting and comprehensible, and there is enough of it, it will have what the acquirer needs and is ready for (“i+1”) (Krashen, 2013).
The most obvious way: tell stories (Mason and Krashen, 2020) and provide “graded readers.” (A better term is interesting/comprehensible readers; they don’t have to be graded or sequenced). The good news is that some of the graded readers of today are genuinely interesting; in fact, in my opinion they deserve the term “literature.”
We can guarantee interest if we provide a large collection and allow students to select what they want to read, encourage students to stop reading if the book is not interesting or is too hard, and help them find a more suitable book. No testing, and no form-focused instruction. Instead, more reading and listening. (For supporting evidence, see Mason and Krashen, 2017).
Another suggestion: comic books! (See Krashen, 2004).
Krashen, S. 2004.. The Power of Reading. Englewood, CO: Libraries Unlimited. http://www.sdkrashen.com/content/books/the_power_of_reading.pdf (second edition)
Krashen, S. 2013 The Case for Non-Targeted, Comprehensible Input. Journal of Bilingual Education Research & Instruction 15(1): 102-110. https://tinyurl.com/y6boor3q
Mason, B., & Krashen, S. 2017. Self-selected reading and TOEIC performance: Evidence from case histories. Shitennoji University Bulletin, 63, 469-475. https://tinyurl.com/yc9tc8ha
Mason, B. & Krashen, S. 2020. Story-Listening: A brief introduction. CATESOL Newsletter, June, 2020. http://beniko-mason.net/content/articles/2020-story-listening-introduction.pdf
I don’t think your experience is at all unusual, but I would love to hear from other educators with relevant experience (no bibliography required!). I believe what you say about “the road to the code … not [being] offered to most ELL” is true for at least 2 reasons.
One is that people–particularly if they’ve not taught beginning/early literacy, and particularly particularly if they’ve not taught kids who have trouble learning “the code” and/or who don’t come from homes with abundant literacy and literacy-related experiences–assume that learning phonics and decoding is simple and easy and “kids will eventually get it.” This is true for some but certainly not for all, and probably not for a whole lot, whether EL or not. Longitudinal studies (unfortunately none as far as I know with ELs, but still relevant) show very conclusively that kids who don’t get the code part (ie foundational reading skills) early on, say, in general by 2nd grade, have a low probability of becoming at least average readers by the end of elm. school. And if they’re not at least average by then… well, you know the rest.
The second reason (and related to the first) is that the most prominent set of beliefs animating EL literacy education has been closely aligned with what used to be called whole language or literature-based reading (aka “you learn to read by reading”), more recently called balanced literacy, and now I’m not sure what… but basically the idea that foundational knowledge and skills of PA, letter-sound associations, phonics, and decoding, all with accuracy and automaticity, don’t need to be prioritized. That is a fundamental error. They DO need to be prioritized. They’re certainly not all there is to becoming fully literate, but there’s a reason they’re called FOUNDATIONAL. To mix my metaphors, think of a foundation for a building. If all you have is the foundation, you haven’t got much… but without a foundation, you have at best a very creaky building and at worst, none at all.
So I think the best way to think of foundational skills instruction is that they’re (very) necessary but not sufficient for full literacy. Furthermore, different kids will require different amounts of code instruction and at different levels of intensity in order to get those foundational skills nailed down. The issue is providing them with the instruction they need in order to get firmly on the road to reading, and literacy more generally, back to the road metaphor.
Thanks for your comment. I hope others will share their thoughts and experiences.