Place names are signposts to the past. And how revealing they are

I WAS RECENTLY BACK IN DEVON. LONG AGO, ON childhood holidays from the industrial north, this seemed to me the ‘real’ England.

- Michael Wood

Still today, a rusting cast-iron signpost peeping through the foliage in a sunken lane gives me an inexplicable thrill. It’s the idea that the deep past might still be touched, just around the corner.

Place names are signposts to the past. Chulmleigh, Woolfardisworthy, Nymet Rowland... what names! And how revealing they are. Take Ipplepen, a village with Roman and Iron Age roots in the hills near Torquay. A recent dig here turned up a small post-Roman cemetery, in use till the eighth century, and a Roman butchers shop, with coins and broken amphorae that once held olive oil, wine and fish sauce imported from the Mediterranean. The name Ipplepen was first recorded in the 10th century, and contains a personal name that appears to be Roman-British or Celtic: Epilus or Ipela. Ina region where the Celtic (pre-Welsh) language may have died out only after 1066, the name of the pre-English landowner was, incredibly, passed down.

In the Home Counties, ancient British (or Welsh) place names are mainly gone - except for river names. Here they linger, from the Thames to smaller watercourses such as the wonderfully named Rib, Beane and Maran (or Mimram) in Hertfordshire. Sometimes English newcomers didn’t understand these words. Avon, of course, has the same derivation as