Sight Words vs Phonics: What the Reading Research Actually Says

The flashcards came home in a zip-lock bag: the, said, was, because, friend. Your five-year-old is supposed to "just know" them by sight, and most nights you dutifully run through the pile at the kitchen bench. Some words stick. Others, was and saw, of and for, swap places endlessly. Meanwhile a friend's school has banned the word "sight words" altogether and sends home strange little books where every word can be sounded out, and you're left wondering which school has it right.

The sight words vs phonics debate sounds like an argument between two camps, and for decades it genuinely was. But the research has settled more than most parents realise, and the answer isn't a tribal one. It's more interesting: every word eventually becomes a sight word, and phonics is how it gets there.

This article unpacks what "sight words" actually means, why decodable books exist, where memorisation still has a small honest job to do, and how to tell when a stack of memorised words is quietly masking a reading difficulty.

"Sight Words" Means Two Different Things

A surprising amount of confusion comes from one term doing two jobs.

Meaning one: high-frequency words. Teachers often use "sight words" to mean the most common words in English, the, and, said, of, was, usually drawn from lists compiled decades ago. Because these words appear constantly, many schools historically taught them by rote memorisation, on the theory that beginners need them before they can read anything.

Meaning two: words recognised instantly. Reading scientists use "sight word" differently: any word a reader recognises instantly, without consciously sounding it out. For a skilled adult reader, catastrophe is a sight word. So is dog. Your entire reading vocabulary, tens of thousands of words, is sight words in this sense.

Here's the crucial point: skilled readers do not build that enormous instant-recognition vocabulary by memorising words as visual shapes, the way you'd memorise a logo. Decades of research show that words are stored through a process called orthographic mapping, the brain bonds a word's spelling to its pronunciation and meaning, letter-sound by letter-sound. To do that, a child needs two things:

• Phonemic awareness, the ability to hear the separate sounds inside spoken words (that ship has three sounds: /sh/ /i/ /p/).
• Letter-sound knowledge, knowing which letters and letter groups represent which sounds.

When a child sounds out ship a few times, the spelling gets glued to the pronunciation. After a handful of successful decodes, the word is recognised instantly, forever. Sounding out isn't the slow detour on the way to sight reading, it's the mechanism that creates sight reading.

Why Pure Memorisation Runs Out of Road

If decoding builds sight words automatically, why not skip it and just memorise? Plenty of children seem to manage it, for a while.

The problem is arithmetic. Memorising words as wholes, without anchoring them to their sounds, is like memorising phone numbers: each new one is a separate effort, and they interfere with each other. Visually similar words (was/saw, though/through/thought, where/were) blur together because the child is holding shapes, not sound-spelling connections. Most children can hold perhaps a few dozen words this way before the system clogs. English has hundreds of thousands.

A child relying on visual memory typically develops compensations that look like reading but aren't:

• Guessing from the first letter, horse read as "house", went as "want".
• Guessing from pictures or context, fluent-sounding "reading" of familiar books that collapses on an unfamiliar page.
• A plateau around age seven or eight, when texts stop repeating a small pool of words and memory can no longer keep pace.

This is why most English-speaking school systems, including Australia, England and a growing number of US states, have moved towards systematic phonics: explicitly teaching letter-sound correspondences in a planned sequence, and giving children practice applying them. It is also why the conversation about the science of reading has reached so many staff rooms.

None of this means memorisation-heavy children are doomed. Many later pick up decoding implicitly. But for the roughly one child in five who finds the sound structure of words genuinely hard to hear, memorisation isn't a head start, it's a delay in discovering the problem.

Decodable Books vs Predictable Books

This is where the funny little books come in.

Predictable (or repetitive levelled) books were designed for the memorisation era. A typical page: "I can see a cat. I can see a dog. I can see a plane." A beginner can't decode plane, but they don't need to, the sentence pattern and the picture do the work. The child appears to read fluently. What they're practising is guessing.

Decodable books are written so that almost every word uses only the letter-sound patterns the child has already been taught, plus a tiny number of explicitly taught irregular words. If the class has covered s, a, t, p, i, n, the book says "Tim sat. Pip is in a pit." It's not literature, and it isn't meant to be. It's the reading equivalent of training wheels: it makes the correct strategy, looking at the letters and sounding out, the strategy that works.

A few honest points about decodables:

• They're a phase, not a diet. Decodables are scaffolding for the first year or two of instruction. Once a child can decode most regular words, they should graduate to ordinary books quickly. A nine-year-old still on decodables needs investigation, not more decodables.
• They don't replace rich stories. Children still need wonderful books read to them daily, that's where vocabulary, knowledge and the love of stories come from while decoding skills catch up.
• Quality varies. Good decodables have actual plots and humour. Stilted ones still do the core job, but the better series make the practice far more pleasant.

If your child brings home predictable texts and is being taught to "use the picture" or "think what would make sense" as a first strategy for working out a word, the school is using an approach that research has largely moved away from. That's worth a polite conversation.

Where Sight Word Teaching Still Belongs

So is there any honest job left for taught sight words? Yes, a small, specific one.

A handful of very common words are genuinely irregular or use patterns taught much later: the, was, said, of, one, two, who, could. Beginners meet these on page one of almost anything, so most good phonics programs teach a small set early, but not as wholes to be photographed.

The effective method is sometimes called "heart words": the child sounds out everything regular in the word and learns only the odd part "by heart". In said, the s and d behave perfectly; only the ai is strange. That's one exception to remember, not four letters. Taught this way, even irregular words get orthographically mapped, and they stick far better than flashcard drill.

So when you practise the school's word list at home:

1. Say the word, then stretch its sounds together: "said, /s/ /e/ /d/."
2. Match sounds to letters, left to right, and circle or point to the tricky part: "the ai is making an /e/ sound here. That's the funny bit."
3. Have your child spell it aloud and write it, saying the sounds, not chanting letter names in a sing-song they could do with eyes closed.
4. Keep sessions short, five minutes beats twenty, and a few words done well beat a pile done fast.

When Memorised Words Are Masking a Problem

Here's the part that matters most for a worried parent, because strong memorisers can hide a reading difficulty for years.

Children at risk of dyslexia often have excellent visual memory, strong oral vocabulary and sharp intelligence. In a classroom built on word lists and predictable books, they can perform convincingly until the texts outgrow their memory, often in Year 2 or 3, sometimes later. By then, habits of guessing are entrenched and confidence has quietly eroded.

Signs that memory, not decoding, is carrying the load:

• Nonsense words are impossible. Ask your child to read made-up words like vop, slint or chame. A child with solid phonics can do this easily; a memoriser cannot, because there's nothing to remember.
• New words trigger guesses, not sounding out, and the guesses match the first letter or the picture, not the letters.
• Reading the same word differently on different pages.
• Spelling lags far behind reading, with the same word spelled three ways in one piece of writing.
• Familiar books sound fluent; unfamiliar ones fall apart.

Underneath these patterns there's usually a weakness in phonological awareness, the ability to hear and play with the sounds inside words, which is the single most common bottleneck in early reading and is highly responsive to explicit teaching, especially when caught early. (We explain it fully in our guide to phonological awareness.)

If several of these signs feel familiar, don't wait for the school to raise it; schools often notice strong memorisers late. A structured screening can measure decoding, phonological awareness and related skills side by side and show you whether what you're seeing is a normal stage or a pattern worth a professional look, that's exactly the gap a tool like GiraffeLens is built to fill, and you can see what it covers at [/what-we-measure]. A formal diagnosis of a specific learning disorder, if it comes to that, requires a registered psychologist, but knowing whether to take that step shouldn't cost thousands.

Questions Worth Asking Your Child's School

You don't need to arrive with research papers. A few calm questions reveal a lot:

• "What phonics program do you use, and what sequence does it follow?" A confident school can name it. "We do a bit of everything" is a less reassuring answer than it sounds.
• "Are the take-home books decodable or predictable?" And: "What strategy do you teach when a child hits an unknown word?" You want "sound it out from the letters", not "look at the picture and guess".
• "How do you teach the tricky high-frequency words?" Listen for sound-spelling teaching rather than pure flashcard drill.
• "How do you check decoding, and what happens for children who fall behind?" Many systems now run a phonics check around age six; ask what follows a low score.

And whatever the school does, the two highest-value things at home stay the same: listen to your child read a little most days with warm patience for sounding out, and keep reading wonderful books to them long after they can read themselves. Phonics and sight words were never really at war. Sounding out is simply how a beginner turns every word in the language, one by one, into a sight word.