Children and adults learn new words when reading, an effect robustly demonstrated through the self-teaching paradigm.
Teachers and tutors using the Speech Sound Pics (SSP) Approach in Australia marvel that children transfer their learning of new words to that of other words, even though they are not sure how. The 'how' doesn't matter to them as much as the evidence that this takes place. The SSP is approach is not 'synthetic phonics' - the approach of choice by the DfE.

​

In the UK all children learning in state maintained schools are explicitly taught 'grapheme to phoneme' correspondences' and additional correspondences are restricted: most teaching is by a teacher at the front of the class who teaches all children at the same time, pace and in the same way. Differentiation is rarely considered or individual learning differences catered for. Little attention is really paid to understanding exactly what each child is learning during their time in class.  Teachers will discuss what they have 'taught' but are often totally unaware of what, if anything, each child actually learned. This is because the organisation of 'instruction' is designed to make life easier for the teacher - eg by creating a 'Scope and Sequence'. Although we can introduce correspondences (the 4 Code Levels) in a systematic way, and offer code level texts - to read and write - this is not a 'Scope and Sequence' and the phase must be over with as quickly as possible. A Scope and Sequence is not possible as Individual differences and needs MUST be catered for - as seen in her Speech Sound Pics (SSP) Approach classroom program, used across Australia.

 

  The correspondences covered within DfE validated synthetic phonics programmes are predominantly used from print to speech - with the GPC shown. The Phonics Screening Check is prime example: when testing phonics knowledge the children are never asked to give phoneme to grapheme correspondences - and yet phonic instruction should focus on both grapheme to phoneme and also phoneme to grapheme correspondences ie decoding AND encoding skills. Encoding is far more difficult - as it involves retrieval of words from orthographic lexicon. Miss Emma has always argued that the best way to avoid difficulties with decoding fluency is to spend a LOT of time in the early stages developing encoding (spelling) skills. It is easier to see which students are at risk: they may blend the sounds into words but struggle to 'Duck Hand' the words. They may be able to 'follow the Monster sounds to say the word' but struggle to do the monster routine - where retrieval is essential. When we show the word we help them to develop phonemic AND orthographic awareness in a systematic and scaffolded manner. We give them the support needed until new pathways are created in their brains, and they can learn about English orthography without help - and move towards being able to say 'I Can Read Without You'. The better they are at spelling, the earlier they reach that phase. 'Phonics' programs only touch the surface of learning to read. 
Not only do ALL children thrive, the children who are usually failed receive the intervention they need.
At least 95% are at the 'self-teaching' phase before they turn 6: with most progress made via implicit learning. The development towards orthographic mapping is portrayed by Ehri (2005a) as a sequence of overlapping phases, each characterized by the predominant type of connection linking spellings of words to their pronunciations in memory; she outlines the process, Miss Emma ensures that ALL learners are able to achieve it. 
Orthographic mapping is enabled by phonemic awareness and grapheme to phoneme knowledge. According to Ehri OM to support sight word reading is facilitated when beginners are taught about articulatory features of phonemes and when grapheme-phoneme relations are taught with letter-embedded picture mnemonics: she researched Letterland, a DfE validated synthetic phonics programme.

Miss Emma argues that this 1:1 pairing is short-sighted: embedded pictures should relate directly to the phonemes, as she does with the Speech Sound Monsters, and not a picture that creates a firm connections between ONE grapheme to phoneme correspondence and that relies on phonemic awareness. eg /a/ shown with an apple. Some children may memorise that connection but without phonemic awareness that connection between a sound they can't isolate, with fruit, can create confusion. Ehri focused on on the very beginning stages, and Miss Emma argues that this is a fools errand: English has an opaque orthography and concepts matter. Within ICRWY children can see /a/ and know the sound value based on the embedded monster, from term 1 of kindergarten. Children are capable of understanding this with the Speech Sound Monsters. The HOW of developing phonemic awareness and orthographic knowledge matters.