Early Literacy: What Research Shows About How Children Learn to Read

Reading is the foundation of academic learning, and the question of how children best learn to read has been one of the most productive but also most contentious areas in educational research. A decades-long debate between whole-language and phonics-based approaches to reading instruction has gradually been resolved by accumulated evidence, though the transition from research consensus to classroom practice has been slow and uneven. Understanding what the science of reading shows, and why it matters, is essential for parents, teachers, and policymakers. The human brain was not evolutionarily designed to read. Unlike spoken language, which children acquire naturally through exposure, reading is a learned technology that requires explicit instruction. Research by cognitive scientists including Stanislas Dehaene has documented the specific brain networks that literacy training recruits, showing that learning to read literally rewires the brain and that the efficiency of this rewiring depends on the quality of instruction. This neurological perspective supports the conclusion that reading instruction cannot simply be left to emerge naturally through exposure to text. Phonological awareness, the ability to hear and manipulate the sounds within words, is one of the strongest predictors of early reading success. Children who can identify rhymes, segment words into syllables, and manipulate individual phonemes, the smallest units of sound, are better positioned to learn the alphabetic code than those who lack these abilities. Phonological awareness develops partly through oral language experience but also benefits from explicit instruction, particularly for children at risk of reading difficulties. The alphabetic principle, the understanding that letters and letter combinations systematically represent sounds, is fundamental to reading English. Phonics instruction explicitly teaches these letter-sound relationships in a systematic sequence, giving students the decoding tools they need to read unfamiliar words. Research comparing approaches to phonics instruction finds that systematic, explicit phonics instruction produces stronger reading outcomes than embedded or incidental phonics approaches that teach letter-sound relationships only as they happen to appear in texts. The whole-language approach to reading instruction, which emphasizes meaning-making, reading authentic texts, and using context cues to identify unfamiliar words, dominated reading education in many American classrooms through the 1980s and 1990s. This approach is not supported by evidence as a primary method of reading instruction. Research consistently finds that children who are taught to rely on context, pictures, and whole-word memorization rather than alphabetic decoding develop slower and less accurate reading than those who receive systematic phonics instruction. The conflict between whole-language advocates and phonics researchers was sufficiently heated that it became known as the reading wars. Fluency, the ability to read accurately and at an appropriate rate with good expression, develops with practice and is an important bridge between decoding and comprehension. Research on fluency instruction finds that repeated oral reading with feedback produces stronger fluency gains than silent reading practice. Reading aloud to children also supports fluency development and has documented benefits for vocabulary acquisition and comprehension, reinforcing the importance of rich oral language environments alongside systematic decoding instruction. Vocabulary knowledge is among the strongest predictors of reading comprehension. Research by Betty Hart and Todd Risley documented dramatic disparities in vocabulary exposure across socioeconomic groups, finding that children from low-income families had significantly smaller vocabularies entering school than children from high-income families. This vocabulary gap, which contributes substantially to reading comprehension differences, persists and widens across the elementary years without targeted intervention. Direct vocabulary instruction and broad reading are both supported by evidence as approaches to expanding vocabulary. Reading comprehension depends on more than decoding ability. Background knowledge plays a central and often underappreciated role in determining whether readers understand what they decode. Research on reading comprehension consistently finds that domain knowledge moderates comprehension: students understand texts about familiar topics better than texts about unfamiliar topics, and building knowledge about topics before introducing texts about them improves comprehension. This finding has significant implications for curriculum design, suggesting that knowledge-rich curricula that build coherent domain expertise across years support reading comprehension more effectively than approach-based curricula that prioritize strategy instruction over content. The science of reading has gained significant policy traction in recent years, with many states enacting legislation requiring reading instruction based on evidence of effectiveness. These laws typically require phonics-based instruction, discourage the use of three-cueing systems that teach children to guess words from context, and mandate training for existing teachers in evidence-based reading instruction. Implementation of these mandates is underway, with early evaluations showing promise in states that have committed most fully to the transition. The gap between what research shows about reading development and what has been practiced in many classrooms represents one of the clearest examples of the cost of slow knowledge translation in education. The consequences, measured in the millions of students who have left elementary school unable to read proficiently, have been significant and are now driving an overdue correction in both policy and practice.