The crucial process of learning to read: what does neuroscience say?

Learning to read and write is a key aspect of human development, affecting all life prospects. However we are presently missing a major opportunity in early childhood literacy, say scientist Professor George Ellis and early literacy specialist, Professor Carole Bloch. This power duo set out to challenge current views in  the field in a paper entitled “Neuroscience and literacy: an integrative view” published in Transactions of the Royal Society of South Africa in 2021. Teaching children to read  is now influenced by reductionist neuroscience views supporting a ‘simple view of reading’ which places heavy initial concentration on phonics as an essential first step to reading with comprehension. This often means expecting children to focus on learning and quickly repeating letter-sound combinations, often meaningless, and out of context. Ellis and Bloch argue that this doesn’t actually fit with how the human brain functions.

The brain is prediction-based and non-linear, usually looking at statistical correlations rather than rules. It is structured to recognise patterns. With feedback it corrects predictions over time to be more and more accurate, amounting to a process of successive approximation: the basic process of learning. This is key to perception: what the mind  sees is not the same as the information coming into the eye – it’s what the brain expects to see based on experience. Concentrating on the rapid learning of letter-sound combinations out of context is thus an especially difficult thing for our brains to do. It does not use this predictive ability.

Emotions are critical to learning for our lump of grey and white matter. There are three main emotional drivers for learning: searching for meaning, the need for community or belonging, and play as practice and invention. The first two greatly help with motivation to keep learning when there is a real world purpose for the learning. The third, play, supports self-regulation, and is how many animals learn to do things, by mock attempts which enable them to develop coordination and contextual responses. In humans, play also drives exploring the use of symbols, language being the prime example.

The field currently argues that listening and speaking are evolutionarily evolved and can be picked up naturally without teaching, but reading and writing are not, and need to be broken down and specifically taught in parts. However everything in our brain’s structure, and the resultant abilities, are as evolved as everything else. Ellis and Bloch argue strongly that after spoken language developed, reading and writing also evolved by socio-cultural interactions as a human capacity, and that socio-cultural teaching is involved in both oral and written language learning.

Contra to claims that one necessarily must read words letter by letter, there is evidence that reading often works via the gestalt of words or phrases in the context of surrounding contextual clues. Proficient readers sample text in the given context to get to meaning. The first line of evidence involves jmulbed wrods you can raed because we read with purpose in a specific context. Second, there are eye-tracking experiments that track where readers look when they’re reading. We don’t actually read letter by letter (left to right for English). The eye darts all over the show, missing some letters and sometimes going backwards, sampling what is needed to obtain the desired understanding. A young literacy learner, if encouraged, uses the same contextual strategies to decipher text – what the meaning is suggesting, what clues they can glean from the structure or grammatical aspects, and the grapho-phonic aspects.

This offers  teachers insight into learners’ strategies – as does the third line of evidence, miscues. A child when reading might say out loud the word ‘parent’ when the written word is ‘father’ -  a meaning based miscue gleaned from context. This shows a top down prediction process is occurring whilst reading, and not just decoding letter by letter, a strictly bottom up process. If the child reads ‘fatter’, instead of ‘father’, she’s still clue seeking, but now using the look of the word instead – not as useful in this case.  Successful reading is reading that preserves the meaning of a piece of text. This can occur without having to decode text using phonics. While knowing phonics is indeed a key part of reading and writing, it ought to be taught in service of meaning, and not prior to it.

There are actually two brain pathways at work here: the dorsal ‘indirect’ pathway to reading, and the ventral ‘direct’ pathway. Teaching phonics is stimulating the dorsal or indirect pathway. The reductionist neuroscience claim is that beginner readers must always learn using this dorsal pathway first, to avoid memory overload. Comprehension comes after this. Ellis and Bloch claim that the ventral direct pathway can be stimulated by reading for meaning that sees the words in a gestalt fashion: seeing the word “dinosaur” or “chocolate” as a whole, for example.  Evidence exists that this ability begins with very young children when they have the appropriate stimulation. It is how the brain clumps information into meaningful units, and thereby avoids memory overload. Think about the way we remember songs, rhymes, and stories, says Bloch.

Why is this important? The debate about the ‘science of reading’ is a global one, and teachers in many countries are increasingly being told that there is only one way to teach reading. This case is being made strongly for children from under resourced settings, as is the case in large parts of Africa. In multilingual South Africa, with its legacies  of colonialism and apartheid,  it is a complex  issue. In 2016 it was reported that 78% of South African children could not read for meaning by the end of grade 4 in any language. Ellis and Bloch have brought integrative neuroscience evidence and literacy together to show why following a  model which emphasises initial speed and accuracy over meaning is limiting, and even may be curtailing reading comprehension for huge numbers of children when they lose interest in a tedious  task.

This evidence suggests that to do better, we ought to note how spoken language, which is enormously complex to learn, develops easily in children due to its central communicative function and how immersed in it they are. We praise babies’ babbling attempts, responding with conventional speech to their attempts to communicate. We fill with pride as these start to make sense. In our constant teaching through modelling and responding, we create conditions for language learning. This helps us  to appreciate and understand the mass of experience children require to do complex learning – and it applies to learning written language too. We can see that while phonics is indeed part of the story, it works  best when teachers are oriented to appreciate their crucial role in immersing children in a similar mass of purposeful, compelling and inspiring examples of using the written word. Learners are then more likely to want to own the power of reading and writing for themselves; under such contextualised conditions integrating phonics teaching can make sense.  

Read the article here