Alison Clarke

Certified Practising Speech Pathologist

Speaking at


Not magic, Mem, statistical learning: helping little readers gather Big Data

It’s kind of amazing that so many young children learn to instantly read thousands of words in their first years of primary school, given English’s very complex spelling system. Like most education academics of her generation, Mem Fox asserts “Reading is a grand guessing game” (Reading Magic, p99), but current reading scientists have a different explanation, based on robust evidence.

Small kids are engaged in Big Data Collection, actively gathering statistics on language patterns. We show them patterns to look for, and their little brains go for it, solving what cognitive scientist Mark Seidenberg calls the puzzle of So Much To Learn, So Little Time.

Some children can figure out some patterns for themselves, but a sizeable minority need explicit and systematic teaching about the sounds, major spellings and word parts of English, before they can even work out what data to collect, let alone collect it. This allows them to develop orthographic mapping, and thus instant word recognition, and frees them from having to sound every word out. Having automatised the word recognition task, and thus rightsized its cognitive load, children can then focus more on building vocabulary, comprehension and fluency.

Seen from this perspective, many current early years literacy teaching practices don’t make much sense. Rote memorising high-frequency word lists, ignoring the patterns therein. Predictable/repetitive texts for beginners, rather than decodable texts. Topic-based or error-based spelling lists, not pattern-based ones. The Three-Cueing Method/Multicueing, actively teaching children to guess like weak readers, not decode like strong ones.

When Universities start teaching primary teachers linguistics – in particular phonology, orthography and morphology – and about the reading science, early years teachers will be better positioned to actively help all but about 3-5% of children crunch the necessary numbers, really nail learning to read, and efficiently move on to reading to learn. When upper primary and secondary teachers understand the implications of this for their workloads and classroom behaviour, they’ll demand it.


Alison Clarke has been a Speech Pathologist since 1988, is also an ESL teacher, and has a Masters in Applied Linguistics. She has worked mainly in schools and the disability sector in Australia and the UK, and is now in private practice. She runs the website, which promotes increased (but not exclusive) use of explicit, systematic synthetic phonics with reading and spelling beginners and strugglers. Alison was 2015-16 Vice President of Learning Difficulties Australia, and has also been an administrator, City Councillor and Mayor.