Whenever we hear an educator begin a statement with the words “Research says…,” we must attend with caution. This is because what is said is an interpretation or a translation of research. The interpretation may or may not accurately reflect the findings of specific research studies. In this article, I express concerns about what is lost in translation in the interpretation of what the science of reading (SoR) research says about particular approaches to and strategies for reading and writing instruction from my positionality as a bilingual teacher educator and researcher in dual language education. In the February 2023 edition of Language Magazine, Kari Kurto, SoR project director for the Reading League, defined the science of reading as “a vast interdisciplinary body of scientifically based research about reading and issues related to reading and writing.” This expansive definition of SoR includes interdisciplinary studies from the multiple fields that comprise the knowledge base for educating multilingual learners. Research on multilingual students is conducted in the academic disciplines of applied linguistics, psycholinguistics, second-language acquisition, second-language reading, cross-linguistic transfer, and the sociocultural contexts of literacy learning. Additionally, international research on literacy in the languages of instruction of target and partner languages in dual language programs informs multilingual literacy instruction. Studies of how monolingual speakers of a language such as English or Spanish give educators a knowledge base for effective approaches to teaching bilingual learners to read and write in both their first and second languages (Mora, 2016).
A theory is an explanation of an aspect of the natural world that is formulated from a body of evidence that has been repeatedly tested and corroborated using scientific methods and protocols of observation, measurements, and evaluation. In his book titled Rethinking the Education of Multilingual Learners, Professor Jim Cummins (2021) proposes three criteria for evaluating the legitimacy and credibility of theoretical constructs and claims that are advanced using the analytical processes common to all scientific inquiry (p. 141):
• Empirical adequacy—to what extent is the claim consistent with all the relevant empirical evidence?
• Logical coherence—to what extent is the claim internally consistent and noncontradictory?
• Consequential validity—to what extent is the claim useful in promoting effective pedagogy and policies?
Cummins’s criteria enable educators to distinguish between evidence-free ideological claims and evidence-based, logically coherent, and pedagogically useful claims that support effective instructional practices for multilingual learners. Education researchers observe patterns of findings that suggest some stable underlying principle at work across various contexts with different student populations in search of refined hypotheses with broader explanatory and predictive power. Theories and constructs in research are the products of systematic observations and rigorous empirical evaluation through peer review before they gain legitimacy as contributions to a scientific knowledge base. A research theory is very different from what may be called “just a theory” in the common vernacular.
The two criteria of empirical adequacy and logical coherence apply to all theoretical constructs, while the criterion of consequential validity is context specific. Isolated research findings become relevant for education policy and practice only when they are integrated into coherent theoretical frameworks. The three criteria are also applicable for analyzing claims from journalists and the media about the effectiveness, or lack thereof, of certain methods, approaches, and strategies for reading instruction. As an example, let us examine the claim that a strategy popularly known as three-cueing is ineffective and possibly harmful to emergent readers.
According to Hanford (2019) in her podcast titled Sold a Story, “The cueing theory provided justification for not teaching children how to sound out words.” Hanford attributes this theory to Marie Clay, a New Zealand literacy educator, and also to “an influential academic in the United States.” Hanford claims that Clay’s theory is a “debunked idea about how children learn to read.” Subsequently, three-cueing is alleged to be a practice that “runs counter to how the brain processes print and language” (Kurto, 2023). Again, this claim is made without citing any studies or SoR research base. In actuality, cueing is not “Marie Clay’s theory.” It is a construct that is the subject of investigation in multiple academic disciplines by many respected academic researchers. Researchers who find these claims to be problematic are unable to identify any empirical data to examine to either affirm or challenge such assertions. Journalists and other critics of three-cueing fail to articulate a theoretical framework or to reference specific research studies to support their condemnation of its effectiveness as a practice in reading instruction. Consequently, such claims lack empirical adequacy to support the argument that teachers should reject cueing as a tool in their toolkit of instructional strategies.
A clear definition of a construct that is identifiable as associated with a theoretical framework or theoretical model and empirical data base is required before a theory or practice can be declared to be scientifically based. The notion of cueing systems comes from linguistics, where subsystems of language are distinguished according to their roles in meaning making from oral language and written text. Three language subsystems are identified in multiple bodies of research as sources of cues to comprehension of words, phrases, and sentences in reading: graphophonics (letter–sound associations; spelling), semantics (meanings of words and phrases), and syntax (word order and grammar). Clearly, graphophonics cues are the basis for orthographic mapping, the technical term for decoding. But what happens when there is no meaning for the word in the reader’s mental lexicon? Or what does the teacher do when the reader does not understand the word because spelling alone does not provide a cue to its meaning within the sentence? A significant percentage of English words have the same spelling but multiple meanings. Does the teacher simply let the reader wander into a linguistic cul-de-sac without providing any guidance for figuring out the meaning of an unfamiliar word?
The case of the critique of cueing as a “debunked practice” in reading instruction (Hanford, 2019) is particularly problematic from the vantage point of curriculum and instruction for multilingual learners. The theoretical framework for miscue analysis and eye-movement research utilizes these terms to articulate the construct of the sources of knowledge from a text and from background knowledge that enable readers to construct meaning from written language. There are decades of second language acquisition research into the theoretical construct of lexical inferencing. The construct is the product of observations of how readers utilize linguistic sources to derive the semantics of words in the process of meaning making in reading (Wesche and Paribakht, 2009). Lexical inferencing is a cognitive and linguistic process that supports vocabulary acquisition and reading comprehension while enhancing bilingual learners’ understanding of syntax and grammar in their first and second languages. Advocates for the science of reading who now criticize teachers for using what is popularly termed the three-cueing system do so without referencing the corpus of multidisciplinary empirical data from which the construct emerged. This omission of reference to a widely utilized theoretical perspective on language and literacy learning processes poses a risk of a loss of credibility of the science of reading for the multilingual learner education research community.
The second criterion for assessing the legitimacy of claims based on scientific research requires that the claim must demonstrate that it is internally consistent and noncontradictory. To examine the criticisms of three-cueing ostensibly based in theories about literacy learning and neuroscientific studies, we must look at established theoretical frameworks. One is the simple view of reading (SVR) (Hoover and Gough, 1990). Kurto (2023) cites this study in a Reading League article, calling the SVR theory “the research that the framework used to describe the reading process is built upon” (p. 34). The SVR proposes that reading is the product of decoding and linguistic comprehension. A combination of skills in both processes is needed to comprehend text. Furthermore, instruction that advances skills in either process advances overall achievement in reading comprehension. Decoding instruction and comprehension instruction are not in competition with each other. The empirical data that support the SVR suggest a reciprocity between decoding and comprehension skills development. Consequently, to discourage teachers from using a set of strategies based on the false notion that their use detracts from students’ learning of decoding skills is not supported by scientific reading research. Therefore, claims against the use of three-cueing systems for meaning are internally inconsistent and contradictory to established theory.
There is also neuroscience research to support the assertion that instruction in language-subsystem cueing is an effective practice. The data base from research on reading in the brain is especially informative when combined with related data bases of observations of reading behaviors. A large data base has been generated using eye-movement studies overlayed onto miscue analysis of oral reading performance (Strauss, 2013).
This triangulated data provides detailed descriptions of eye fixations and their relationship to reading miscues to explain readers’ perceptual and informational processing of text. The corpus of eye movement/miscue analysis (EMMA) data confirms that meaning construction relies on psycholinguistic strategies far more efficient and effective than letter–sound conversion instruction. This scientific technology challenges the claim that a meaning-making theorical model of reading is disconfirmed by neurobiological research into brain-region activation in phonological processing.
Through brain imaging (fMRI) technology, a dual route to the mental lexicon, one relying on grapheme–phoneme correspondences and the other on direct lexical access from the visual word form, is supported by neuroimaging data (Strauss, 2013; Strauss et al., 2009). It is misleading to infer that neuroscience has discredited the theoretical framework of meaning centered, psycholinguistic models of reading. On the contrary, neuroscientists argue that emerging concepts from the neuroscientific study of brain function both support and are supported by psycholinguistic research on the reading process.
It is inconceivable that teachers could be told that it is not aligned with the science of reading to teach vocabulary or grammar during interactive oral reading with their multilingual students. It is unimaginable that teachers would be advised against scaffolding texts to support their English language learners’ comprehension of challenging academic texts because it detracts from their decoding skills acquisition. It is difficult to imagine that the SoR does not promote students’ use of metacognitive strategies to enhance self-monitoring and problem solving during oral and silent reading to comprehend text. Yet this is the inevitable consequence of journalists’ inexpert and uninformed claims, made without clearly identifying the theoretical framework and research foundations for the approach that they condemn. The censure of a common instructional practice popularly known as three-cueing is not valid. In fact, readers’ use of language subsystems as cues in meaning making in reading is essential for comprehension. Alternative theoretical perspectives cannot invalidate the findings from empirical research data bases without establishing a causal relationship between a construct and intervention program outcomes. MacPhee et al. (2021) make this observation about media portrayals of the science of reading: “The media have asserted a direct connection between basic research and instructional practice that, without sufficient translational research that attends to a variety of instructional contexts and student populations, may perpetuate inequities” (p. S145).
It is counterproductive for advocates of SoR to allow the media to create practitioner lore condemning practices associated with theoretical frameworks that are not favored by a certain group of reading scientists. If science of reading advocates seek to maintain legitimacy with the multilingual learner education research community, we must set a very high bar for judging any theoretical framework to be “disproven,” “debunked,” or “wrong.” Educators must not make judgments about the utility of approaches and strategies based on a caricature of typical practice. The full corpus of scientific research on literacy learning must inform instruction, without privileging a narrow range of methodologies or discrediting empirical data bases because of ideological rather than scientific criteria. Professor Cummins (2021, p. 131) warns that theoretical constructs cannot be made the scapegoat for problematic instructional practices or misguided broader policy initiatives.
We academic researchers cannot be held responsible for misinterpretations, misrepresentations, and misapplications of the findings of our studies. Instead, we encourage dialogue with our practitioner colleagues to improve the translation of research findings into effective instruction to ensure that reading and writing programs are accessible to all students.
Cummins, J. (2021). Rethinking the Education of Multilingual Learners: A Critical Analysis of Theoretical Concepts. Multilingual Matters.
Hanford, E. (2022). “Sold a Story: How teaching kids to read went so wrong.” APM Reports. https://features.apmreports.org/sold-a-story/?fbclid=IwAR2NWgzZSOemLf7Ctg4fW3bBiGuvKF9DdljD14swHpFbPQkjNBTz7-QXRlk
Hoover, W. A., and Gough, P. B. (1990). “The Simple View of Reading.” Reading and Writing, 2(2), 127–160.
Kurto, K. (2023). “Clarifying the Science of Reading.” Language Magazine, 32–35.
Mora, J. K. (2016). Spanish Language Pedagogy for Biliteracy Programs. Montezuma Publishing.
Strauss, S. L. (2013). “We Need a Paradigm Shift in Research on Reading and Dyslexia: Fundamental problems with fMRI studies of written language processing.” Journal of the Neurological Sciences, 333, e579–e628.
Strauss, S. L., Goodman, K. S., and Paulson, E. J. (2009). “Brain Research and Reading: How emerging concepts in neuroscience support a meaning construction view of the reading process.” Educational Research and Review, 4(2), 21–33.
MacPhee, D., Handsfield, L. J., and Paugh, P. (2021). “Conflict or Conversation? Media portrayals of the science of reading.” Reading Research Quarterly, 56(S1), S145– S155.
Wesche, M. B., and Paribakht, T. S. (Eds.). (2009). Lexical Inferencing in a First and Second Language: Cross-Linguistic Dimensions. Multilingual Matters.
Jill Kerper Mora, EdD, is associate professor emerita at San Diego State University. Dr. Mora disseminates her international research and expertise in Spanish language and literacy instruction for dual language educators through her MoraModules website at moramodules.com. Dr. Mora is a member of the California Committee for Effective Literacy.