Interview - Andrew Davis, Research fellow at Durham University's School of Education

Can you explain why you say that the evidence base for synthetic phonics, or for any other prescriptive method to teach children to read, is 'fundamentally flawed'?

Researchers investigating synthetic phonics approaches often try to use 'scientific' methods appropriate for testing drugs or fertilisers. Imagine comparable levels of precision in a description of how reading should be taught.

Imposing such a strategy would turn teachers into mere technicians, supposedly implementing a 'teacher proof' or 'pupil proof' method. For if, contrary to our supposition, it could be modified in the face of pupil responses, it would no longer conform to a specification. By definition, 'pupil proof' methods don't cater for differing pupil needs and stages of development, and hence are not teaching.

In actual fact, research into teaching reading usually investigates strategies following broad and abstract 'principles'. Yet the teachers concerned cannot avoid taking many decisions on the hoof, in the light of children's varying responses. There are no simplistic rules.

Teachers could not and should not say: whenever pupil accents differ from yours, do this; whenever she recognises the whole word while supposedly blending sounds, do that; if he can already read silently, always do the following. Teachers reach sensitive practical verdicts in each case. The allegedly researchable 'method' fails to outlive the exigencies of real teaching.

What do you think the effects of synthetic phonics will be?

Suppose the phonics check remains, and Government refuses to insert into the Programmes of Study crucial phrases such as 'Teachers remain responsible for decisions about how and when they use phonics'.

If so, I fear that 'reading for meaning' will be at least temporarily undermined for some pupils, in favour of mere letter sounds blending, with consequent damage to motivation and cognitive development.

What do you think is the best way to teach children to read?

There is no 'best way'. Teachers should continue to share the joy of books and stories and to draw flexibly on a repertoire of tools including varieties of phonics. They should not feel pressurised at any one time to use one 'method' exclusively.