The Science of Reading, explained without jargon

A mother in our parent group sent the same question twice in one week. Her older daughter, in a district one town over, was learning to read with decodable books and explicit sound-letter drills. Her younger son, two miles down the road in a different district, was being handed picture books and told to guess unfamiliar words from context. Both schools said they were teaching reading "the right way." She wanted to know which one was lying.

Neither was lying, exactly. But only one was teaching what reading researchers have spent fifty years confirming actually works. The phrase parents keep hearing — the Science of Reading — is the short name for that body of evidence. It is not a curriculum, not a brand, and not a political position. It is what we know about how the human brain turns squiggles on a page into meaning.

Here is what's in it, in plain English.

A very short history of the Reading Wars

For most of the twentieth century, two camps argued about how children learn to read. One camp, broadly called phonics, taught children the relationship between letters and sounds first — c-a-t says cat — and built up from there. The other, which evolved into whole language in the 1970s and 80s, taught that reading was a natural process like learning to speak. Surround children with good books, the theory went, and they would absorb reading the way they absorbed talking.

Whole language lost ground after researchers measured what was actually happening in classrooms. Kids exposed only to immersion were guessing at words from pictures and context, and many were not catching up. By the early 2000s, a softer version called balanced literacy had taken over most American schools. It mixed some phonics in with a lot of leveled-text reading, where children read books matched to their current "level" and were taught a three-cueing strategy — guess from meaning, sentence structure, or the first letter.

That model is what most parents under 45 were taught with. It is also the model the Science of Reading movement is actively pushing back against — not because phonics was missing entirely, but because three-cueing was teaching the wrong habits. A child who guesses pony when the word is horse because both fit the picture has not learned to read. They have learned to bypass reading.

The five pillars

In 2000, the United States Congress asked the National Reading Panel (NRP) to settle the question with evidence. The panel reviewed more than 100,000 studies and named five things that reliably matter when children are learning to read. These are now usually called the five pillars:

1. Phonemic awareness — hearing and manipulating individual sounds in spoken words. A child who can tell you that cat starts with /k/ and rhymes with bat has phonemic awareness. This is purely an ear skill; no letters involved yet.
2. Phonics — connecting those sounds to letters and letter patterns on the page. This is where /k/ meets the letter c (or k, or ck).
3. Fluency — reading accurately, at a reasonable pace, with expression. A child who can decode every word but reads at one word per second is not yet fluent, and comprehension will suffer.
4. Vocabulary — knowing what words mean. You can decode photosynthesis perfectly and still have no idea what you just read.
5. Comprehension — pulling meaning out of a text, including things the text does not say directly.

The original NRP report is still publicly available through the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) at nichd.nih.gov. It is dry reading, but the executive summary is short and worth a look if you want to see the raw conclusions before anyone marketed them to you.

Scarborough's Reading Rope

A year after the NRP report, a researcher named Hollis Scarborough published a diagram that has done more to clarify reading instruction than almost anything since. She called it the Reading Rope.

The rope has two main strands twisted together.

The first strand is word recognition. It contains phonological awareness (hearing sounds), decoding (mapping letters to sounds), and sight recognition of familiar words. As a child practices, these three threads tighten into something automatic — they see a word and read it without conscious effort.

The second strand is language comprehension. It contains background knowledge, vocabulary, language structures (grammar and syntax), verbal reasoning, and literacy knowledge (knowing how books and stories work). These threads tighten more slowly, over years of conversation, being read to, and exploring the world.

Skilled reading is the rope — both strands, braided. A child who is strong on one side and weak on the other will struggle. A kid who can decode anything but has a thin vocabulary will hit a wall around third or fourth grade, when the books stop being about familiar topics. A kid with a rich vocabulary but shaky decoding will guess and fake their way through for years before someone notices.

The International Dyslexia Association maintains a clean overview of the rope at dyslexiaida.org. If you only look at one outside source from this post, look at that diagram.

Why even "sight words" are phonics

Here is the part that surprised me when I first read it as a parent.

Linnea Ehri, a developmental psychologist who spent decades studying how children store words in memory, showed something counterintuitive. Even words that look irregular — said, was, the, of — are not stored as visual pictures. They are stored phonetically, mapped sound by sound onto their letters, with the irregular parts flagged as exceptions. Ehri called this orthographic mapping.

What this means in practice: a child does not memorize said by staring at it 200 times until the shape sticks. They learn /s/ /e/ /d/, they connect those sounds to s-a-i-d, they notice that the middle bit is weird, and that small act of phonetic anchoring is what makes the word permanent. The brain is a sound-mapping organ, not a flashcard.

This is why the "heart words" approach you may have heard about — teach the regular parts phonetically, and only flag the irregular part as something to "know by heart" — has quietly replaced the old sight-word flashcard drill in most evidence-aligned classrooms. I wrote more about that distinction in Sight words vs heart words: what your kid actually needs to memorize.

What's actually still debated

The Science of Reading is not a closed file. A few things are still being worked out.

Decodable vs leveled texts, especially after the first year. Almost everyone now agrees beginning readers need decodable books — books written using only the letter patterns the child has been taught. The argument is about when, and how quickly, to transition to richer, less controlled texts. There is no clean line.

The exact order to teach phonics patterns. Programs differ. Some teach short vowels first; some teach high-utility consonants; some start with the letters in the child's name. The research supports explicit, systematic instruction; it does not strongly favor one specific scope and sequence over another.

How much of this generalizes outside English. English is unusually messy — about 50 phonemes mapped onto 26 letters with a thousand-year pile-up of borrowed spellings. Spanish, Italian, Finnish, and other "shallow" orthographies are far more predictable, and the phonics phase is shorter. Some of the Science of Reading findings are about reading in general; some are specifically about reading in English. It is worth knowing which is which.

How to teach kids with dyslexia, or English Language Learners, or both. The pillars still apply, but the dosage, pacing, and intensity all change. The University of Oregon's Center on Teaching and Learning publishes some of the cleanest open materials on this at reading.uoregon.edu. For dyslexia specifically, the IDA Knowledge and Practice Standards are the document to know.

A short checklist for parents

When you look at any reading program — school, after-school, app — these are the questions that separate evidence-aligned work from rebranded balanced literacy.

• Is phonics taught explicitly and systematically? Not "embedded when it comes up." Taught, in order, with practice.
• Do beginning readers get decodable texts? Books that match what they've been taught so far, not books they have to guess through.
• Is three-cueing absent? No "look at the picture, guess from context" as a primary strategy for unknown words.
• Is there frequent, low-stakes assessment? Quick checks every week or two, not a high-stakes test at the end of the year.
• Does it cover all five pillars? Phonics alone is not the whole picture — vocabulary, fluency, and comprehension still matter.

If a program ticks four out of five, it is probably fine. If it ticks one out of five and tells you the others happen "naturally," ask harder questions.

Where Lumi sits

Lumi is one tool, not a full reading program. Its job is the part it can do well over voice: phonemic awareness games, explicit phonics practice, decodable sentences, and the kind of one-on-one back-and-forth that builds fluency. The vocabulary, background knowledge, and language comprehension side of the rope still comes from being read to, talked with, and let loose in the world. We are deliberate about that boundary. An app that claims to do all five pillars on its own is selling something.

If you want to see how the adaptive side of this works in practice, Adaptive learning isn't a setting — it's the whole product walks through it.

The Reading Wars are mostly over. The research is clear enough that the question for parents is no longer which side but which programs are caught up. That is a different and much more answerable question.

Image brief

• Hero image: A warm illustrated diagram of a braided rope splitting into two strands labeled "word recognition" and "language comprehension", drawn in the Lumi cocoa and sun palette.
• Inline image 1: A simple side-by-side panel showing a child guessing "pony" from a picture vs. the same child decoding "horse" letter-by-letter. Place after the "very short history" section.
• Inline image 2: A clean five-block diagram of the NRP pillars (phonemic awareness, phonics, fluency, vocabulary, comprehension) stacked like building blocks. Place inside the "five pillars" section.

Internal link suggestions

• Sight words vs heart words: what your kid actually needs to memorize — anchor text: "Sight words vs heart words: what your kid actually needs to memorize" (already linked in body).
• What early dyslexia looks like — and what helps — suggested anchor text: "what early dyslexia looks like".
• Adaptive learning isn't a setting — it's the whole product — anchor text: "Adaptive learning isn't a setting — it's the whole product" (already linked in body).

Editor's note

Two things to verify before publishing. First, the NICHD URL for the NRP small book is stable as of last check, but the agency has reorganized publication pages before — please confirm the link still resolves to the executive summary [VERIFY]. Second, the IDA "Scarborough's Reading Rope" infographic page URL has moved at least once in the past; if it 404s, swap to the IDA homepage and update the anchor text [VERIFY]. The Ehri orthographic mapping summary is paraphrased from her published work over the 1990s and 2000s; if you'd rather cite a specific paper, the 2014 Scientific Studies of Reading piece is the usual go-to.