Once again, Howard Jones' post above seems to make entire sense to me. There is nothing in the post the does not seem to me to reflect entirely the methods and approach that Steve Roud adopts throughout his book. Howard pinpoints the reasons why there are so few songs reproduced in it and to that could be added the reasons why he does not deal with the folk revival - by the time that emerged (circa 1950) the tradition only survived in fragmentary form in a few parts of England. Some, not all, of the traditional performers who became involved in the folk club movement saw it as a potential source of material. Three examples of this that come to mind straight away would be Fred Jordan, George Belton and Sheila Stewart (yes, I know not English, but she learned and subsequently sang a song that she learned at our folk club).
He also asks us not to obsess on origins because that only leads to a chicken & egg situation which we cannot resolve satifactorily. Rather, we should draw conclusions on ideas and theories that we can give evidence for and not what we woukd like to be the case, whilst keeping an open mind of what future research may discover.
On pages 24 - 25 in the chapter on definition, Steve gives a list saying, "The following statements will help us to put these abstract concepts into context, and add some details". The list also gives the thrust of what the book is going to be about. -

* It is not the origin of a song that makes it folk but the transmission within the folk tradition which makes it so.
* It is not where it comes from that matters, but what the folk do with it (what some people summarise as 'folk song by destination' rather than 'folk song by origin').
* Folk song is not only defined partly by its social context; it relies on that context for its existence.
* Folk song does not exist in a cultural or musical vacuum.
* Oral/aural transmission has always been an extremely important component m folk traditions, but since the invention of printing, there has probably never been a purely 'oral' tradition, even among the lower classes.
* However learnt (even if from print or musical notation), performance is normally carried without the aid of written text or notated tune.
* The folk have always taken their material from anywhere they liked, in whatever medium they have found convenient. As soon as new media, such as recorded sound, films and broadcasting became available, these were readily adopted as sources of new songs.
* Within an active tradition, songs are passed from person to person and thereby down the generations. New or different songs can enter a local tradition at any time.
* If nobody in the community likes a song enough to learn it, it will die within that community, but this is not the only possible reason for the 'death' of a song.
* Folk songs are not necessarily very old, but they must have been around long enough to become part of this traditional transmission (two generations might be an acceptable rule of thumb).
* Like all human cultural activities, folk song is not static but is in a continual state of flux, and has always changed over time, A new song usually loses its origin and becomes anonymous and common property.
* The people themselves rarely have an 'original' of a song with which they can compare their own version.
* From the folklorist s point of view, no version of a song is 'better' than another, but singers themselves will have criteria by which they will judge songs and performances.