The two wires of the light were connected, the light was hung and all worked as expected. Having not come across a double insulated light before, and being nervous of a metal light, I tested the wire shade with a non contact detector pen and saw that it indicated the shade was live.
Thinking the light must be dangerously faulty, I took it down and powered it up via a plug (with appropriate precautions). This time the shade was not live.
I then tested the light fixing (see pic) and saw that *this* was the cause of the live signal despite being screwed into wood. This led me to scan the surrounding ceiling and I saw that quite a large area was flagged as live by the detector (50cm2). Using an endoscope I inspected inside the hole and, whilst there were wires visible, they were not significantly close to the ceiling plaster, nor had I screwed through any of them. Measuring the voltage of the fixing using a multimeter measured 1.6v.
There was no obvious signs of damp in the plaster. but does this mean there is definitely a short somewhere causing this? (I'm guessing it's an obvious yes!) Are there any techniques for diagnosing where the problem may originate without making more holes in the ceiling or lifting the floor above?