I am reading this book, but I am reading it very slowly. I break off after every chapter to read something else. It gives me time to think about the implications of everything that has been written. I find that I need to go back and re-read sections to make sure that I have considered all the implications. Already, the pages that I have read has a mass of those 'post-it' type stickers marking statements that I will need to go back to re-read and reconsider. In a way, I am very glad that I am not reviewing it (though one editor has already indicated to me that he would have preferred that the book had been sent to me by the person who allocates material to reviewers). The reason that I am glad is because I would hate to have to rush through a book that is as dense and meticulously researched as this because of a review deadline. So far, I have resisted the great temptation that Brian Peters failed to do and 'skipped ahead' to read conclusions - but this has been difficult. I want to savour and enjoy what is likely to be the most important book that I have read about the subject that has absorbed my attention all my adult life.
This morning I reached the end of Chapter 7 and we have reached the end of the 18th century in considering a multiplicity of evidence, we reach a section where he, at last, offering some preliminary findings. He has been considering the impact of written and published song and tine material - opera, theatre stage shows, broadsheets, chapbooks and other sources on what was sung by the classes of people from whom the 'folk singers' came:-

.... on what we can surmise to be the state of 'traditional song' of the period. On a superficial reading across the genre, this seems to be true of musicians' tune books in general, which, as far as songs are concerned, are often much closer to the 'art music' of the period than other sources. It may be that their compilers, being semi-professional jobbing musicians, spent time in theatre orchestras and military bands, and playing for middle-class concerts and balls, and that their repertoires reflect this. But this is a superficial impression, and needs to be checked further before it can be accepted as evidence.Page 293

He then goes on to consider the last part of evidence from that century, the manuscripts of Ralph Dunn and in particular the song Poll of Plymouth. This interests Steve because:-

It was repeated in literally dozens of songsters, chapbooks and broadsides, but doesn't seem to have been noted by any of the folk song collectors. Page 294

What Steve seems to be saying here is that there is something about this song (of which he gives the lyrics) that did not attract potential singers. It didn't have the qualities required for it to be taken up by 'The voice of the people'. Without that, the song does not alter and develop on being passed on through entering the repertoires of the common people (whoever they may be). It is when songs start to be altered in this way that they have become more interesting to Steve and other contemporary researchers than the constant haranguing about definition.
The last two paragraphs of Chapter 7 are more revealing about Steve's attitude and his modus operandi than anything else that I have read in the book so far.

As the evidence stands at present, we can reach some tentative conclusions. If the manuscripts are accepted as evidence of vernacular singing, the folk-collectors severely underestimated of higher class art/popular music of the pleasure gardens and theatres in the traditional-song repertoires of a century before their time. The influence of print on traditional song was extensive. The degree of continuity between say, 1790 and 1890 is surprisingly low, and songs did not, in the main, last for a hundred years in the popular tradition, unless the degree of continuity is disguised by the collectors' selection policy. The latter is feasible but does not bear close scrutiny. It seems to argue that all collectors would recognise an eighteenth-century art song at sight, and decline to note it. This may be true of those who had a good working knowledge of popular sing history such as Sabine Baring-Gould. Frank Kidson and Anne Gilchrist, but these collectors are precisely the ones who would have found such survivals interesting and would have noted them. The balance of probability is that these songs simply did not survive to be collected around the turn of the twentieth century.
But the evidence can be read the other way round. As we know that 33 per cent of the 'folk songs' collected later originated in the eighteenth-century of before, the fact that we can find so few of them in the sources investigated simply means that these sources are not sufficiently 'folk' and we are looking in the wrong place. Certainly, compilers of manuscripts will have been, by definition, more literate that the average working person, because they could write as well as read, but we are back to our basic problem. If the 'folk songs' of the time left no tangible trace, we can say little or nothing about them. Page 296

Consider the contributions of all the previous commentators on English folk song. There is little doubt that Sharp and Karpeles knew what they were looking for before they set out the find it. They collected what that wanted to hear and ignored anything that was outside their preconceptions. Then much later we have Lloyd and all the other Marxist commentators, Harker and other de-bunkers, Georgina Boyes and other feminists; they all bought the pre-determined socio-political agendas with them. All have given us invaluable information to help to a greater understanding of the subject but we have to approach all of them with a pre-knowledge of the author's position. The only ones who have radically changed academic opinion have been the ones who have written that the position of women in the collecting work of the first has been seriously understated; they are producing plenty of evidence to support this. The interesting talk by Lizzie Bennett that I heard at a Traditional Song Forum meeting this year produced facts that this happened in Sussex and I had not heard this information before.
The main factor in my (incomplete) interpretation of Steve's approach is that he bends over backwards to omit anything to which he cannot point to a providing evidence. It is this clarity of thought; this abhorrence of assumption that is, I believe, going to provide the way for future academic researchers and writers on the subject of folk song.