Phonics, Three Cueing, and the Lost Searchlights: A False Choice in Reading Instruction

Problems arise when people talk about phonics or three cueing.

The only real problem with three cueing is when it becomes the only strategy, and learners have no ability to use partial decoding. This happens when they do not know the core code (the most common grapheme–phoneme correspondences, or GPCs), or when they fail to recode after figuring out unfamiliar words.

Do not treat it as an either/or choice, or you risk the situation now seen in England, where at least one in four children cannot read by age eleven. Since the mandatory adoption of systematic phonics in 2013 and the removal of cue-based strategies (previously known as Searchlights), instruction has narrowed.

The term Searchlights came from the National Literacy Strategy (NLS, 1998), which described reading as drawing on four “searchlights”:

1. Phonic knowledge (sound–symbol relationships)

2. Graphic knowledge (visual recognition of words and letters)

3. Grammatical knowledge (sentence structure)

4. Contextual knowledge (meaning and prediction from context)
   

Teachers were encouraged to help children “use all the searchlights” to make sense of text.

When the Rose Review (2006) was published, it criticised the Searchlights model for encouraging children to guess words from context rather than focusing on phonics as the word mapping (how letters and sounds connect) strategy. Following that, the Department for Education replaced the Searchlights model with the Simple View of Reading (SVR) and made Systematic Synthetic Phonics (SSP) mandatory.

Validated SSP programmes teach around 100 GPCs, leaving children unable to decode unfamiliar words if being able to decode through the whole word is their only option. They are not shown how to use partial decoding, track back, or recode, and most are not taught set for variability skills explicitly. They are told not to use contextual clues or figure out unknown words from pictures. As a result, too many children are not reaching the self-teaching phase because their instruction stops at decoding.

Understanding what is needed for learners to reach the self-teaching phase (Share, 1995; Ehri, 2014, 2015) is essential. Start by asking: What does this learner need, at this stage of their learning pathway? Or with THIS text. What do they already know, and what will they need to do, to figure it out?

Get them to track back and recognise this as recoding, which is central to self-teaching (Share, 1995; Ehri, 2014). Guessing the word is useful, if they then map the speech sounds, spelling and meaning. It's just a 'speech to print' approach to word mapping.

The only issue with three cueing is when it stands alone and is not followed by recoding. Do not treat it as oppositional to phonics, or you replicate England’s ongoing problem. More than a decade of phonics-first instruction has left reading outcomes static (DfE, 2024).

As Seidenberg (2023) and Castles, Rastle, and Nation (2018) explain, skilled reading depends on integrating multiple sources of information to achieve fluent comprehension. Phonological recoding must occur, but context and meaning cues play an important role in verification and self-teaching. Anyone who understands the reading science recognises the value of three cueing with recoding, working alongside systematic phonics, not against it.

In England, where synthetic phonics is mandated, one in four children still cannot read or spell at an adequate level by age eleven (DfE, 2024). These children have typically been taught only around 100 GPCs. In my school visits, I find that strict adherence to “no guessing” prevents many from developing set for variability skills. When I introduce the concept in Word Mapping Mastery teacher training, teachers often worry they might be contravening DfE guidance. I use this image and ask them to work out which of the GPCs in this image they don't teach explicitly.

It's not a complex piece of text. But what does a child do if they aren't taught all the GPCs, and are told not to guess - or not given this in class as it is not fully decodable to the child? Outside of the structured phonics lessons, where materials are designed around being 'decodable' to the child, according to where they are in their phonics lessons, what happens for the rest of the day, when they have to read and also write words with other GPCs. This is often the first time teachers in England have been asked this by anyone.

What is dismissed as three cueing can actually be more effective than relying solely on limited GPC instruction. As the example shows, when a child reads, “Sam has so **** apples. Poor Fred does not have *** apples,” they can deduce the missing words many and any. They then map the phonemes and graphemes (m–a–n–y to /m ɛ n iː/ and a–n–y to /ɛ n iː/). Trying another word, such as a number, would not fit the graphemes or the initial phoneme. Dyslexic learners who work this way often reach the self-teaching phase earlier. Three cueing can be highly effective if it leads to phoneme–grapheme mapping.

Our word mapping technology now makes this process visible by showing the grapho-phonemic structure of any word. Using Phonemies® and our patented Code Mapping® algorithm, it reveals both the phoneme values and their grapheme markers, mapping all words of English. This shows the graphemes, for example - teachers can map their materials with the Code Mapping Tool in the MyWordz® tech.

And when the Phonemies are added they can see the different sound value for /a/ on those two words (any and many) and check the Spelling Clouds® for others.

Skilled readers use cueing to reach self-teaching. But this is generally implicit learning. The key difference between those who map words easily and those who struggle lies in phonemic awareness rather than word knowledge. That is why this topic matters so much, because teachers, often under intense time pressure, may misinterpret three cueing debates without delving into the underlying research. Let's explore it rather than just accept it by those taking a 'phonics v whole word' stance. We can take a 'what does this child need, at this time' stance instead. Those using the Speech Sound Pics approach teach all of these GPCs systematically and explicitly- but with tech, so they work through the 4 Core Code Levels at their pace. However we do not rely on this in our quest to help every child move into the self-teaching phase, or expect them to WANT to read. This is why we use the One, Twe, Three and Away! series now - the first 50 books mapped to show the orthographic code. Miss Emma

The Neurodivergent Reading Whisperer®