Rifleman: those who actually play the music (which we did last night to great success)

And as I did this afternoon. We seem to be on the same side. Jolly good.

Darowyn - no, it's not about authority. It's partly about having a consistent, comprehensible and usable definition for a recognisable body of material, so that we can agree (in most cases) that song X is traditional and song Y isn't, and then get on with singing them (or not). I think the 1954 definition is consistent, comprehensible and usable, partly because (as Spleen said) it's all about where the material caem from: you can have a traditional song sung in a singaround, played in the Albert Hall or played in a ska-punk stylee through a Marshall stack, the 1954 definition doesn't care.

But mostly, I think, it's about whether we think the other two dimensions of the 'folk' definition Spleen mentioned - marketing and club performance - are fine as they are, or we think they're lacking a certain traditional something. Would you be happier if, on hearing a new album described as 'folk', you could safely assume there'd be some traditional material on it? Would you prefer it if, going out for an evening at a folk club, you could confidently expect to hear a couple of traditional songs? I'd answer Yes to both, and I guess some people would answer No. Ultimately that's what we're arguing about. It's an argument that's never going to be resolved; as I've said before, the only thing that keeps me coming back to it is trying to understand what an alternative definition of 'folk' might be.

I mean, is May You Never a folk song? If so, why? Does it depend on where and when and how it's performed? Was it a folk song when John Martyn recorded it, or has it become one since? Is it a folk song when someone who's never heard it plays the album for the first time? Say what you like about the 1954 definition, it's a lot simpler.