Since 1945 the musical scene in East Africa has been characterised by a series of important developments which ran parallel to an extensive political–cultural change of consciousness. All the musical forms which arose in this connection were ‘popular’ in the sense that they were appreciated by broad sections of society.

There is an apocryphal legend that when Louis Armstrong was asked whether he considered his music to be folk music he replied: ‘Wal, yeah, I guess so: leastways, I never heard of no horse making it.’ If and when we talk of Bob Dylan as a folk singer, we do so partly of course in this fundamental sense: his music is of, about and for people. But we also mean — as Satchmo's interviewer must have done — that Dylan calls on resources and techniques that pertain to oral cultures rather than ostensibly to our own. Unlike ‘real’ folk song, Dylan's words and music are written down; yet what is notated is no more than an approximation to the sounds heard, which are created empirically for each performance, from verbal inflection and from the body's gestures. Dylan is a ‘folk’ artist in that his sources are absorbed at a more or less pre-conscious level: from the monody and banjo-picking of ‘poor white’ Americans and the traditions of white British folk music behind them; from the black blues and the African heritage it stemmed from; and from some more sophisticated, or at least literate, sources such as white hymn, march and parlour ballad. To these musical roots there are literary parallels: Dylan's verses have been deeply influenced by rural folk ballads, by the poetry of the blues, by the Bible, by Bunyan's Pilgrim's Progress (scarcely less pervasive in the Bible belts than the Good Book itself), by Hymns Ancient and Modern, by children's rhymes and by runic verse of all kinds, including Blake's Songs of Innocence and Experience, which are close to the core of Dylan's experience and which, although literary, were sung by the poet himself to improvised, presumably folk-like, tunes. In his early days Dylan operated with his voice — untrained, ‘natural’, veering between speech and song — and his acoustic guitar.

If the North East is typical of other English regions, then the following propositions will have to be severely modified or given up: that there was a distinct ‘break’ between pre-industrial and industrial workers’ culture; that concert hall was a London phenomenon; that it originated mainly from petit-bourgeois ‘free-and-eases’; and that it reached its peak in Edwardian England, as music hall. Further, I will contend that ‘leisure’ and ‘work’ cannot rationally be separated in the study of workers’ culture, that social historians and students of culture need each other and that the concept of ‘folk song’ does more harm than good.

There is good evidence that for over ninety-nine per cent of human history, and for ninety-seven per cent of the time since the emergence of our own species (homo sapiens sapiens) approximately 70,000 years ago, all music was popular, in so far as it was shared and enjoyed by all members of a society. If there were distinctions of style within a society's music, they were accepted as sings of functional or social differentiation rather than as barriers to mutual communication. Distinctions between sacred and secular music, between music for young and old or men and women, were generally drawn within the style of each music culture, and, at least in principle, a member of one group could perform and appreciate the music of another group in the same society.

Rock, the saying goes, is ‘the folk music of our time’. Not from a sociological point of view. If ‘folk’ describes pre-capitalist modes of music production, rock is, without a doubt, a mass-produced, mass-consumed, commodity. The rock-folk argument, indeed, is not about how music is made, but about how it works: rock is taken to express (or reflect) a way of life; rock is used by its listeners as a folk music — it articulates communal values, comments on shared social problems. The argument, in other words, is about subcultures rather than musicmaking; the question of how music comes to represent its listeners is begged. (I develop these arguments further in Frith 1978, pp. 191–202, and, with particular reference to punk rock, in Frith 1980.)