Cognitive Predictors of Reading in Children

The Reading Project

A research partnership to understand how children learn to read, why some struggle, and what we can do about it.

Vision Lab, Centre for Neuroscience, IISc Bangalore X EkStep Foundation

ABOUT

What happens in the brain when child learns to read?

Reading is something the human brain was never designed to do. We evolved to recognise faces and objects — not text. So when a child learns to read, the brain does something remarkable: it repurposes a small patch of visual cortex — originally wired for recognizing objects - and gradually reassigns it to process letters and words. Scientists call this region the Visual Word Form Area (VWFA).

But it doesn't stop there. Learning to read triggers a chain of changes across the brain.

This process happens in children learning to read — and also in adults who become literate later in life. But we still don't fully understand how it unfolds: the sequence, the critical stages, or why some children get stuck. Nearly all reading research has been done in English — we know very little about how this works in Indian languages, which place different demands on the brain.

This project aims to fill that gap.

WHY IT MATTERS

India's reading crisis

India has made enormous strides in getting children into school — enrolment is at historic highs. But being in school is not the same as learning. Across rural India, a vast number of children sit in classrooms every day and still cannot read.

These aren't children out of school. These are children in school — attending class, year after year — who still haven't learned the most basic skill that all further learning depends on. Across a country with 260 million school-age children, even small percentages translate to tens of millions of children left behind.

Many barriers have been addressed — school access, teacher training, textbooks, meals. But one piece is missing: a scientific understanding of why individual children struggle to read, from the brain up. Without that, interventions are designed by intuition, not evidence.

THE PARNERSHIP

IISC & EkStep - a research collaboration

This project — titled "Cognitive Predictors of Reading in Children" — is a formal research collaboration between the Vision Lab at IISc Bangalore and the EkStep Foundation. The partnership is driven by a shared conviction: that improving literacy at scale requires more than better teaching methods or better technology — it requires a fundamental understanding of how reading develops in the brain. By combining rigorous neuroscience with deep experience in education at population scale, this collaboration aims to identify the cognitive factors that determine whether a child learns to read, build tools to measure them, and ultimately design evidence-based interventions that can reach children across India's diverse languages and learning contexts.

Vision Lab, Centre for Neuroscience, IISc Bangalore

Led by Prof. S. P. Arun (Fellow, Indian Academy of Sciences; Senior Fellow, Wellcome Trust / DBT India Alliance).

The lab has various publications on visual neuroscience and has spent over a decade studying how the brain recognises objects, letters, and words. Their prior work on reading — spanning compositionality, jumbled word processing, reading fluency in children, and neural coding of letter position — directly shapes the research questions of this project. India's diverse scripts give the lab a unique natural laboratory for isolating the visual factors in reading

EkStep Foundation

EkStep Foundation is a not-for-profit that creates open digital infrastructure and technology to solve societal challenges at population scale — spanning education, healthcare, and financial inclusion. In the literacy space, their AXL programme uses AI and speech recognition to provide personalised reading and math practice in children's mother tongues, and is already deployed in schools across India. Through years of grassroots work with teachers, state education systems, and communities, EkStep brings a ground-level understanding of why children struggle to read — and supports this research to ensure future literacy interventions are grounded in brain science, not just intuition.

Reading rewires how the brain sees words

By comparing readers of Telugu and Malayalam—different scripts with similar sounds—researchers showed that the brain learns to build words from individual letters rather than storing whole words as fixed pictures. This flexible letter-based coding also helps explain why we can often read jumbled words.

Agrawal, Hari & Arun, 2019 & 2020

Visual processing predicts reading fluency in children

Children who are better at spotting small differences between letter pairs tend to become more fluent readers. This relationship remains strong even after accounting for other well-known reading skills, showing that visual processing itself is an important predictor of reading success.

Agrawal, Nag, Hari & Arun, 2022

Children learn letter shapes before letter names

Young children can tell letter shapes apart before they can say their names. This shows that the brain starts building visual knowledge of letters early, providing the foundation for learning to read.

Das, Nag, Usha & Arun, 2024

Reading experience reshapes individual neurons

A primate brain study showed that learning letter tasks fine-tunes individual neurons. With practice, these neurons become better at recognizing letter identity and letter position, revealing how reading experience reshapes the brain's visual system.

Agrawal, Saha, Munda, Arun & Dehaene, 2025

RESEARCH PLAN

How we'll do it?

1Build a Cognitive Battery

Distill the scientific literature to identify key predictors of reading. Develop automated, child-friendly tests — validated against gold-standard measures. Benchmark eye-tracking and EEG for measuring attention and engagement.

2Map the Stages of Reading

Track the same children longitudinally for one to two years. Use both behavioural and neural measures. Identify which cognitive factors drive reading at each stage — from decoding letters to fluent comprehension .

3Design Interventions

Develop targeted training modules aligned with the core cognitive predictors found in Aims 1 and 2. Test their effectiveness, required practice, and whether gains transfer to broader classroom outcomes

4Extend to Mathematics

Apply the scientific framework developed for reading to the study of mathematics learning. Begin with expert workshops and systematic review. Then identify the cognitive factors that drive math acquisition.

PROJECT OUTCOMES

Why these outcomes matter

🧪 Understanding Why Children Struggle

For the first time, a clear picture of what's happening in a child's brain at each stage of reading — so we can move from guessing to knowing why a child is stuck

🧠 Tools That Work in Any Classroom

Assessments that teachers and schools can use across languages and regions — no specialists or expensive equipment needed

📚 Help That's Targeted, Not One-Size-Fits-All

Interventions designed around the specific cognitive factor a child is struggling with — whether it's visual processing, sound-letter associations, attention, or comprehension

📄 A Path to Literacy at Scale

Evidence-based training modules ready for deployment across India's diverse schools — moving from thousands of children helped to millions

Meet the Team:

Prof. SP Arun - Principal Investigator

Dr.Akshatha Bhat - Project Scientist

Mr. Shreyas Balaji - Research Assistant

Mr.Suyash Sachdev - Research Assistant

Mr.Shashwat Mishra - Research Assistant

FAQ

Questions people ask us

Don't we already know how to teach reading?

We know a lot about teaching methods — phonics, whole language, balanced literacy. But we know surprisingly little about what happens inside a child's brain as they learn, or why two children with the same teacher end up at very different levels. This project aims to move from guesswork to precision — understanding the mechanism, so we can design targeted help.

How is this different from reading research done elsewhere?

Nearly all reading research has been done in English with Western populations. Indian scripts — alphasyllabic, hugely diverse — present different cognitive challenges. And most reading research treats reading as purely a language problem. We uniquely study the visual side: how the brain's object recognition system adapts to process text. India's diverse scripts give us a natural laboratory that doesn't exist anywhere else.

Will this actually help kids, or is it purely academic?

The project is explicitly designed to move from lab to classroom. Step 3 is entirely focused on building, testing, and refining interventions. Our partnership with EkStep — which already runs AI-powered literacy tools in thousands of schools — gives us a direct pipeline from findings to deployment at scale.

Why study reading at a science institute?

Reading starts in the brain's visual system — recognising letter shapes, building them into words, connecting them to sounds and meanings. Understanding this chain requires neuroscience expertise in visual perception, neural coding, and computational modelling. Our lab's prior work has already revealed insights about reading that would never have emerged from an education-only approach.

Does the brain really change when a child learns to read?

Dramatically. Brain imaging studies have shown that reading creates an entirely new functional region — the Visual Word Form Area. It strengthens connections between visual and language circuits, refines speech-sound processing, and shifts how the hemispheres handle face recognition. These are among the most profound examples of experience-dependent brain plasticity known in humans.

What about children with dyslexia?

Understanding typical reading development is a prerequisite for understanding why some children struggle. The tests we develop will help identify where a child is stuck — visual processing, phonological difficulties, attention, or a combination. The interventions in Step 3 can then be tailored to target the specific weak link.

References

Dehaene S & Cohen L (2007). Cultural recycling of cortical maps. Neuron, 56(2): 209–217. doi
Dehaene S, Pegado F, Braga LW, Ventura P, Nunes Filho G, Jobert A, Dehaene-Lambertz G, Kolinsky R, Morais J & Cohen L (2010). How learning to read changes the cortical networks for vision and language. Science, 330(6009): 1359–1364. doi
Dehaene-Lambertz G, Monzalvo K, Dehaene S (2018). The emergence of the visual word form: Longitudinal evolution of category-specific ventral visual areas during reading acquisition. PLOS Biology, 16(3): e2004103. doi
Dehaene S (2009). Reading in the Brain: The New Science of How We Read. Viking Press.

Agrawal A, Hari KVS & Arun SP (2019). Reading increases the compositionality of visual word representations. Psychological Science, 30(12): 1707–1723. doi
Agrawal A, Hari KVS & Arun SP (2020). A compositional neural code in high-level visual cortex can explain jumbled word reading. eLife, 9: e54846. doi
Agrawal A, Nag S, Hari KVS & Arun SP (2022). Letter processing in upright bigrams predicts reading fluency variations in children. Journal of Experimental Psychology: General, 151(9): 2237–2249. doi
Das J, Nag S, Usha MN & Arun SP (2024). Early readers know letter shape without knowing letter names. psyArXiv preprint. preprint
Agrawal A, Saha S, Munda S, Arun SP & Dehaene S (2025). Orthographic training finetunes a neuronal letter position code in primate visual cortex. bioRxiv preprint. preprint

ASER Centre / Pratham (2024). Annual Status of Education Report (ASER) 2024. asercentre.org

EkStep Foundation. AXL — Assisted Language & Math Learning. axl.ekstep.org

Google Sites

Report abuse