I still have not finished reading Folk Song In England - being out of the country for over a month hasn't helped but my instinctive feeling is that this book is best read in small doses, a chapter at the time, to allow the brain time to absorb the implications before moving on to read and digest another aspect.

The day after reading the review by Vic Gammon which I posted above, I read this passage (pages 442 ? 444) which seem to encapsulate very much of the attitude and approach that Steve Roud brings to his book: -

It has been reliably claimed that 90 to 95 per cent of the items at Victorian and Edwardian collectors noted as 'folk songs' had appeared on broadsides in the eighteenth and nineteenth centuries. This in itself no solid indicator of a direct link between print and oral tradition, but coupled with examples of direct testimony from singers about the quality of the songs from broadsides and songbooks, and the growing number of studies using internal evidence, the trend is abundantly clear.
Most of the folk-song collectors were scathing about the quality of the: broadside songs but were well aware of the fact that many of their singers had definitely gathered their material from print. Lucy Broadwood, for example, stated, 'The words of many country ballads are derived. directly or indirectly, from broadsides and Alfred Williams:-
. The songs were mainly obtained at the fairs. These were attended by the ballad-singers, who stood in the market-place and sang the new tunes and pieces, and at the same time sold the broadsides at a penny each. The most famous ballad-singers in the Thames Valley in recent times were a man and woman, who travelled together, and each of whom had but one eye. They sang at all the local fairs, and the man sold the sheets, frequently wetting his thumb with his lips to detach a sheet from the bundle and hand it to a customer in the midst of the singing.
This is not to argue that all singers learntall their songs from print - far from it. Henry Burstow, the singer from Horsham, Sussex, gives direct evidence on this question in his Reminiscences of Horsham (1911). After writing of learning songs from his parents and other people he knew or met, continues:-
The remainder I learnt from ballad sheets I bought as they were being hawked about at the fairs, and at other times from other printed matter. I remember, when quite a boy, buying for my mother of a pedlar, as he sang in the street, the old ballad 'Just Before the Battle Mother'. This was her favourite song.
We have less direct evidence for the earlier centuries, but it is clear the manuscripts which are analysed in earlier chapters that people regularly copied songs from broadsides into their own notebooks.
Two things are now abundantly clear. Firstly, once printing had been invented, there was never again a pure 'oral' tradition, but oral and print were: intimately interwoven. Secondly, the songs that the ordinary people turned into 'traditional' or 'folk' songs were normally written by outsiders and reached them first in printed form.
For these reasons alone it would be essential for us to fully understand the genre, but a close knowledge of the broadside and chapbook trade is also important for more practical reasons. Whatever the characteristics of an 'oral tradition' may be, its undeniable failing for historical enquiry is its almost complete lack of a datable evidence trail, and the temptation this offers for wishful speculation on the part of commentators is enormous.
A song collected from a shepherd or a dairymaid in 1903 might have been knocking around the village for 200 years, or they might have learnt in the previous week, and without this information we have no basis on which to assess or investigate the workings of 'the tradition'.
As the accumulated evidence mounts up, it seems increasingly that the broadside texts were indeed the originals of many songs, because they were written specifically for that medium, and we therefore have a welcome opportunity to get to grips with questions of what really happened to songs when they entered a local tradition. If we know how they started and how they ended up, we can at least start to investigate what happened in between.
Supporters of 'oral tradition' are often understandably wary of such comparative work, because it is so easily couched in terms designed to prove the degenerative and unreliable nature of 'oral tradition', but used sensitively it could actually demonstrate what is built into many a definition of 'folk song' ? that transmission within a healthy tradition is a positive force and, by selection and variation, results in 'better' songs. Or it could simply reveal the essentially conservative nature of the singers? attitudes to song texts and the fundamental fidelity of their memories.