"The term does not cover composed popular music that has been taken over ready-made by a community and remains unchanged, for it is the re-fashioning and re-creation of the music by the community that gives it its folk character".

This statement has been misrepresented in the past by various people focusing on the "composed" and ignoring the "unchanged". What then of Webber's May Song or Parting Song, or of John Connolly's Fiddlers Green? Both have seen variation, both have been passed on orally in many cases (despite the fact that several recordings exist, some people still learn songs the old way)? Purists will tell us that these are not folk songs... I beg to differ.

Further, the 1954 definition is arbitrary and imperfect, in my view, fashioned with the knowledge of the day and does not pass the test of time. The moment a collector records a source singer, the oral-transmission cycle is broken (something the definition conveniently, but erroneously, ignores); this shows the arbitrariness of the "oral" rule.

But any definition that uses external trappings to define a social process is bound to be anchored to the time of its making and hence imperfect. It's like trying to describe dance as "the synchronous movement of body and limbs to the rhythm of accompanying mysic" - i.e. bollocks (think of marching, for example). Folk and its meanings are only expressed by the externalisations, and should not be defined by them.