Why Doesn’t the Science of Reading Work?

A consensus built across 167 languages

The history of literacy education has never lacked methodological debates. For decades, advocates of learning to read through immersion in complete texts clashed with those who insisted on the need to teach the relationship between letters and sounds explicitly. At the same time, another debate was unfolding in ministries of education across much of the world: could research conducted primarily in English-speaking countries really be applied to education systems with different languages, different writing systems, and different social realities?

The report Effective Reading Instruction in Low- and Middle-Income Countries: What the Evidence Tells Us was produced precisely to answer that question. Its purpose is neither to present a new method nor to reopen long-standing pedagogical controversies. Instead, it seeks to determine whether the existing body of evidence is robust enough to guide literacy policy. The answer is based on a review of 151 studies conducted across Africa, Asia, Latin America, and the Middle East, covering more than 167 languages.

Its conclusion is striking because it combines two ideas that might initially seem contradictory. The first is that learning to read follows universal principles. The second is that not all languages are learned in the same way. A child learning to read in Spanish, where the relationship between letters and sounds is highly regular, requires less phonics instruction than a child learning English, whose spelling system is far less consistent. Someone learning to read in Hindi or Amharic must become familiar with hundreds of different symbols, while a student using the Latin alphabet faces a very different challenge. The pathways differ, but the architecture of learning remains remarkably stable.

This is probably the report’s most important finding. There is no single method that works identically across all languages, but there is a set of principles that consistently emerge whenever research examines how children become skilled readers. The consensus no longer rests on studies carried out in just a handful of countries. It speaks dozens of languages and has been tested in some of the world’s most diverse educational contexts.

If we know how children learn to read, what is going wrong?

Around 70 percent of ten-year-old children in low- and middle-income countries cannot read and understand a simple age-appropriate text. The figure is well known. What is truly alarming is the context in which it occurs. Never before have so many children been enrolled in school, and never before has so much research been devoted to understanding how they learn to read.

According to GEEAP, part of the problem lies in a deeply rooted misconception: the belief that reading develops in much the same way as spoken language. It does not. Children learn to speak because the human brain is biologically prepared to do so. Growing up surrounded by people who speak is enough. Reading belongs to an entirely different category. Writing emerged only about six thousand years ago—far too recently for the brain to have evolved structures specifically dedicated to it. Every reader must build new connections between written symbols, speech sounds, and word meanings. Reading is a cultural achievement, not a biological one.

This distinction has profound implications for teaching. A child does not naturally discover that words can be broken down into sounds, that those sounds are represented by letters, or that the two can be combined to decode unfamiliar words. Someone must teach these relationships explicitly. That is why the report places such emphasis on skills such as phonological awareness, systematic phonics instruction, and reading fluency. These are not fashionable pedagogical trends; they are the mechanisms through which the brain transforms written symbols into language.

The report dismantles a false dichotomy that has shaped countless debates in education. For years, the discussion revolved around whether teaching children to read meant helping them understand texts or teaching them the written code. Research shows that these two abilities are inseparable. A student may fully understand a story when someone reads it aloud and yet be unable to read it independently because they have not mastered decoding. The opposite can also happen: a student may read every word accurately without understanding the text. Reading proficiency emerges only when both dimensions develop together.

For this reason, the report highlights six components that must be developed in an integrated way: oral language, phonological awareness, phonics, fluency, comprehension, and writing. The last of these deserves special attention because it challenges a widespread misconception. Writing is not presented as the result of learning to read; it is part of the learning process itself. Writing letters, words, and short texts helps consolidate word recognition, improves spelling, and strengthens reading comprehension. Reading and writing are not parallel skills—they reinforce one another.

Seen in this light, the question changes entirely. The persistence of learning poverty does not reflect a lack of knowledge about how reading should be taught. It reflects the difficulty of ensuring that this knowledge reaches millions of classrooms consistently. This is where the report shifts the focus of the debate.Su conclusión resulta llamativa porque combina dos ideas que podrían parecer incompatibles. La primera sostiene que aprender a leer responde a principios universales. La segunda recuerda que no todos los idiomas se aprenden igual.

What does an effective literacy system actually do?

A reading program may have a sound curriculum, well-designed materials, and reasonable teacher training, yet still fail. The GEEAP report identifies six conditions that enable a reform to move beyond policy documents and become a consistent reality in classrooms.

The first is alignment. The most successful programs do not consist of disconnected initiatives; they organize a coherent package of lesson plans, student books, teacher guides, training, and ongoing support. Every element reinforces the others. Six of the eight highest-performing reading programs examined in the report followed this integrated model.

The second concerns instructional materials. A teacher’s guide may incorporate every appropriate principle and still prove useless if it is difficult to follow. The report recommends lessons that are clear, contain only a limited number of activities, bring together all essential information on a single page, and include examples that are relevant to the local context. In Malawi, overly dense lesson plans significantly reduced their actual use in classrooms. For this reason, materials should be tested with teachers before being rolled out at scale and revised according to the difficulties that emerge.

The third condition is teacher training focused on concrete classroom practices, combined with sustained follow-up support. One-off workshops have limited capacity to change well-established routines. Teachers need opportunities to practice new techniques, receive feedback, and benefit from coaching as they begin applying them in their classrooms. The report extends this recommendation to pre-service education: faculties of education should also incorporate the teaching of reading into their curricula and provide high-quality practical experiences.

The fourth strategy addresses something many reforms overlook: how teachers actually adopt new practices. Limited time, unclear instructions, excessive workloads, gaps in knowledge, and familiarity with previous routines all influence the decisions teachers make in the classroom. Asking teachers to implement too many changes simultaneously, or relying on vague guidance, can undermine a reform before it has the chance to take root.

The fifth condition requires political, administrative, and financial support. Successful programs depend on coordination among ministries of education, local authorities, school leaders, and assessment systems. They also require stable funding for instructional materials, training, coaching, and monitoring. An initiative sustained solely through external funding is vulnerable to disappearing once the project ends.

The sixth involves embedding the program within the regular structures of the education system. This means transferring expertise to local teams, adapting supervision procedures, integrating the changes into the curriculum, and ensuring that public administrations can continue the work without relying indefinitely on external advisers. In India, the SERI program initially combined external trainers with government coordinators, who gradually assumed responsibility for delivering the training until they were able to lead it independently.

The report adds an important warning: scaling up can dilute results. A program that is highly effective in a small group of schools requires strong systems of monitoring, continuous evaluation, and quality assurance if it is to maintain its impact as it expands.

An effective literacy system, therefore, does not place the entire responsibility on teachers, nor does it rely on a single intervention. Instead, it builds a complete chain: it decides what should be taught, designs practical tools, prepares and supports those who will use them, removes implementation barriers, secures the necessary resources, and embeds the program as a permanent part of the public education system.

Five patterns that appear to work

If this report demonstrates anything, it is that countries that succeed in improving literacy do not do so because they have discovered a secret method or an exceptional pedagogical innovation. They succeed because they make a series of coherent decisions and sustain them over time. Contexts, languages, and available resources may differ. What changes very little are the priorities.

The report avoids presenting a universal formula. Education systems start from very different realities, and they do not all have the same resources or face the same challenges. However, when the reforms that have successfully improved literacy are compared, a pattern emerges that is difficult to ignore. The same decisions recur, even though they take different forms depending on the context.

Language of instruction: it is better to begin with one children know

The evidence shows that children learn to read more easily when literacy instruction begins in a language they already know. Where this is not possible, the development of oral language becomes even more important, so that linguistic difficulty does not become an additional barrier to learning.

Clear and well-defined curricula

The most effective programs do more than establish broad objectives. They clearly define which skills should be acquired at each stage, how much time they require, and how they relate to one another. This sequence gives teachers and students a shared roadmap.

Teacher training: examples and ongoing support

The studies reviewed agree that practices change more easily when teachers have access to concrete examples, opportunities to try out new strategies, and support as they incorporate them into their daily work. Training therefore stops being understood as a one-off course and becomes an ongoing process.

Clear, aligned, and context-appropriate materials

Far from being a secondary element, the report shows that the design of instructional materials shapes the quality of teaching. Clear guides, books aligned with the curriculum, and activities adapted to the local context make it easier to apply the principles of the science of reading consistently in the classroom.

Ongoing support

Reforms produce better results when mechanisms are in place to observe what is happening in schools, support teachers, and address difficulties during implementation. The evidence shows that expanding a program without maintaining this level of follow-up usually reduces its impact.

These five recommendations reflect a way of understanding literacy, and they all share the same logic: reducing the gap between what research knows about how children learn to read and what happens in classrooms every morning. The science of reading works. But education policies must be made to work with it.