Left to right isn’t the norm for about half the world’s population and looking at the letters can be a problem. Recent research into learning difficulties suggests that many children fail to read because they don’t see letters properly at all. To understand what children experience, imagine reading the Financial Times in a stuffy car being driven at speed around winding roads by Jeremy Clarkson.

Teaching reading is all about strategies and skills, but which ones? Reading research has more factions than the Afghan opposition - and the similarities don’t end there. Researchers seem to spend as much time rubbishing each other as they do on learning what works, and their enthusiasm for testing out their latest theories on the hapless school population is undimmed by failure.

Theories abound. There’s real books, Reading Recovery, reading schemes, Look and Say, phonics I The Government tried to strait-jacket the debate in 1997 with the literacy hour, a structured session focused on phonics and “fast-paced activities”. Cat-sat-on-the-mat reading schemes were out, as was the quaint idea that children developed reading skills through “exposure” to real texts.

Known by some researchers as the Tamsin principle, this approach looked at middle-class children’s reading success and assumed that having lots of real books was the key factor, rather than the money, the culture, the parental support and so on. The literacy hour’s focus on group work hasn’t satisfied teachers who believe children need the support that comes from adults hearing individual children read, while some militant researchers argue that the shift towards phonics hasn’t gone far enough.

Phonics, the focus on letter sounds rather than whole-word recognition, gained the upper hand in the last battle. The latest wheeze is synthetic phonics, which claims to get children reading within weeks. It is said to be the bees’ knees with boys, dyslexics and ethnic-minority children. It also cures the plague and makes children eat their greens. Allegedly.