Italian has the term aria, followed by French and English air, whose more central meaning reminds us of the way in which tunes were originally produced. The term occurs in ‘Londonderry Air’, the name of a paradigm popular tune, half folk-tune, collected by Jane Ross in County Londonderry in the 19th century, and published, with the words of ‘Danny Boy’ added, in 1855. German borrows the term too, but in the Aria mit verschiedenen Variationen that Bach composed for Mr Goldberg, it denotes a piece of melodic and harmonic thinking that is very far from what the English know as a tune.
‘Song’, Lied, chant and canto denote an entire musical episode, and there are tuneless songs, just as there are tunes that cannot easily be sung, like the first subject of Rachmaninov’s Second Piano Concerto – tuneful though it certainly is – or which contain leaps that can be managed only by the trained voice – like the first subject of Strauss’s Don Juan. And by using the term ‘subject’ to describe those examples I have implied something about their inner nature – as the starting points of musical arguments. The Greek word ‘theme’, which has been adopted into French, English, German and Italian, expresses a similar idea, and the tradition of Western classical music abounds in themes which, for all their intrinsic interest as musical units, are far from being tunes – the theme that opens Beethoven’s 5th symphony for example (more gesture than tune), the theme of the Passacaglia from Brahms’s 4th or the opening subject of Mozart’s D Minor piano concerto, K 466. A theme can comprise many tunes or tuneful fragments – like the opening theme of Bruckner’s 7th, or that of Elgar’s 2nd Symphony. And a tune can govern works, like those that comprise Schumann’s Dichterliebe, which have nothing deserving the name of a theme, and may in itself be incapable of thematic development.
Of course, the distinctions here are in the first instance verbal, and may have little bearing on the underlying musical reality. Nevertheless, the questions that they naturally provoke tell us something important about music, and in particular about the idea of a musical individual, which we can know and love as a whole. The tune is an extended episode in the history of melody, and one that illustrates the way in which musical individuals have arisen from the human need to sing and dance. Not every musical culture lays stress on the tune, but all – or almost all – have some form of melody, and regard melody as essential to the musical individual that contains it. (The most important exception to that generalisation is African drum music, which reproduces in the dimension of rhythm some of the complexities that we know from melodic voice-leading.)
We can divide melodies roughly into the melismatic, the thematic and the tuneful. The first kind includes plainchant, raga, and certain kinds of Rock – idioms in which the momentum pushes easily through harmonic and rhythmic boundaries, and has no defining closure. The second kind includes the melodies of our own classical tradition, in which elements develop in new directions, and move towards closures not contained within the original statement. The third kind, which principally interests me, is bounded at each end, contains a distinctive and recognizable internal order, and is regarded as a complete individual, to be memorized as a whole. And among tunes we can make a further division, between those that arise from setting words (the logogenic), those that arise from dancing (the orchegenic), and those that arise from the harmonic relations of the tones that they contain (the harmonegenic). Folk music contains tunes of all three kinds, though the harmonegenic are more typical of ‘art music’, or music influenced by art music, like Blues and Ragtime.
Plainchants are often instantly recognizable and thrilling to the ear – like the Dies Irae, the Hodie Christus natus est or the Veni Creator Spiritus. They are melodious, but are they tunes? Berlioz turns the Dies Irae into a tune, but only by squeezing it into diatonic harmony and destroying its modal character. A plainsong chant exhibits another kind of order from the tune – it is a fragment of eternity which has neither beginning nor end in the scheme of things. It flows endlessly, and that which we know as the beginning is simply the point where the voices enter. We hear the movement of the chant as preceding the entry of the voices, which are heard as ‘taking up’ a melody that flows in the cosmos unendingly. Plainsong presents us with another idea of the melodic individual, one that seems not to belong to us but to exist eternally in another realm. And of course, that is part of its meaning, and one reason why these melodies, and the performance tradition that protects them, and which has been many times lost and recovered, are important to us.
A tune is a melodic unit, and typically one that can be sung. It has a beginning, a middle and an end. It may also begin with an up-beat, which both is and is not a part of the tune, and the question whether a tone or a phrase is an up-beat is one on which much may hang and which does not yield to an easy answer. (Consider the first three notes of the Londonderry Air – to describe them as an up-beat is to misrepresent the significance they assume as the song proceeds, with just such three-note entrance figures repeated four times, and three note exit figures occasionally matching them: Ex. 1.)
A tune may be diatonically tonal, like ‘The Sweet Nightingale’, modal, like the English folksong ‘Scarborough Fare’, pentatonic, like the Scottish folksong ‘Over the Sea to Skye’, even atonal like the tune that opens Schoenberg’s Violin Concerto. But it must have a beginning, and it should work towards a conclusion, even if, like the tune from Rachmaninov mentioned earlier, it peters out, rather than settling on some final full stop. One of the complaints made by Adorno (see ch. 13) against the American music industry is that it highlights the tune in place of the theme, and so curtails all possibility of extended musical thinking. His point is that the American popular song has opted for melodic statement against thematic development – in other words that it belongs to the tradition of strophic singing rather than patient listening. And its tunes, moreover, come from a fixed repertoire, assembled from familiar phrases, and present no challenge to the ear. Pick up the songbook from any year between 1920 and 1970 and you will find a selection of 32-bar ‘numbers’: strophic songs consisting of two internally related tunes, the one reserved for the verse, the other for the chorus. This format speaks of the social meaning of the songbook, which consists of so many dialogues between narrator and audience, between the interesting outsider with a story to tell and the community waiting to absorb him. But the American songbook consists only of tunes – alone or paired with a chorus – and tunes which seem so firmly encased in their own unified movement as to stand alone, refusing the very idea of development.
The baroque ‘air’ is frequently the subject of variation, in which the underlying harmonic progression provides the anchor (as in the Goldberg variations); the classical theme is the subject of development, which may involve breaking down the theme into its elements, diminishing and augmenting its rhythmical order, elaborating particular phrases, and so on. The American song cannot easily be treated in that way: it bursts through any attempt to vary it, wearing the same sociable expression, and demanding the same harmonic sofa on which to sit. It refuses to be discomposed into thematic fragments, and is intrinsically resistant to the classical forms of development. Performers will ‘improvise around a tune’, without really varying or developing it, and even if it is the harmonic sequence, rather than the melodic line, that is taken as the subject, the improvisations will preserve the outline of the tune and seldom work towards a tune of their own.
The American songbook is only one example of a tradition of popular music centred on the tune. The folksong traditions of Europe and America are similar, and so are many liturgical traditions. The hymn-book is a particularly interesting case, since it shows in all clarity just why tunes are needed. The strophic tune is the companion of the strophic verse form. And whether the verse tells a story or praises God, it can be easily memorised only if it has a repeatable pattern: and this pattern is most easily memorised when sung. And this gives a clue to the nature of tunes. The paradigm use of a tune is in a song, with or without chorus. Its beginning, middle and end mirror the beginning, middle and end of verse stanzas. It has a form analogous to a sentence – a clear statement of where we have got to in a narrative, which advances by rhythmic and sentential repetition. A tune can bear being many times repeated without variation; and if is so rarely varied this is in part because of its solid and repeatable character, which resists the attempt to dissolve it in a narrative larger than itself.
Folksongs in the European tradition are also strophic, and this form survives in the verses long after the music has been lost – as in so many of the border ballads and other relics collected in the 18th century. (The term ‘ballad’ is cognate with ‘ballet’, Italian ballata – not something sung, but something danced. Again we are seeing through etymology into the social origins of music in all its popular forms.) Tunes appeal to us because they return us to the strophic order of the hymn and the ballad – the memorable melody that carries a burden of meaning in its words. But their order is a musical order: their devices are those whereby one phrase answers, continues or concludes another. And in all forms they exemplify the way in which tones organize themselves in musical units, so as to create boundaries at both ends of a sequence, and an intelligible contour between them.
It is this experience of the boundary that is most mysterious and most fundamental to the evolution of melody in Western music. It has an evident relation to a phenomenon that we know from our experience of language. Sentences in a language begin and end, and the words that occur in the middle are are heard as embedded in a continuous process. We can explain this experience in terms of the cognitive processes whereby we recuperate from the surface structure the deep structure that generates it, and theories of deep structure will eventually be semantically grounded. The very process whereby we create meanings in words imposes the boundaries and continuities that we hear in language.
This has led to the suggestion – taken seriously by Fred Lerdahl and Ray Jackendoff – that musical order exhibits the same kind of deep structure that we witness in language, and that boundaries and continuities arise from the cognitive segmentation required by the attempt to recuperate deep structure. However there are two insuperable objections to this approach. The first is that music has no semantic structure, so that the search for deep structure cannot be grounded, as it is grounded in linguistics, in a semantic theory. The second is that the only candidates for a generative theory of musical structure, which might assimilate the surface segmentation to that which we observe in language, is the theory of Schenker, which applies only to diatonic tonal music, and which relies upon the harmonic function of scale degrees. Yet tunes and their boundaries are not always tonal and certainly not always diatonic. To those objections we might add the evident truth that, while the boundary experience in language is a human universal, the ability to discriminate sentences being an essential part of linguistic competence, the boundary experience in music is subject to considerable cultural variation. Tunes as we know them are tightly wrapped in their boundaries: plainchant melodies, Indian ragas, the melodies of the traditional Arabic maqamat, and so on, have weak or non-existent boundaries, so that one strophe flows into the next without a comma or a stop.
On the other hand, it is surely plausible to suggest that scales and modes have a powerful input into the boundary experience. There are places where melodies typically start, and (in the Western tradition) places where they typically end. One explanation of the ‘up-beat’ experience is that the melody is being led to its starting point, both rhythmically, by a ‘left-over’ beat, and melodically, by an entry on the dominant, as in many English, German and American folk-songs. There are also tunes that start beyond the boundary, as though preceded by a kind of musical glottal stop. Such tunes are normal among folk-songs originating in places where the spoken language places the accent on the first syllable of a word, as in Czech or Hungarian. When we listen to a song we cross a boundary into the tune, either after a few measures or at the very outset: and we know that we have crossed that boundary immediately, by recognizing that each note that occurs after that boundary is linked to its successor as a note in a tune. What are we hearing when we hear this? The first thing that we hear is direction: each note leads to its successor. This is a distinctively musical experience and not one that is easily replicated in language, where the position of words is determined by their meaning, rather than by any apparent movement from word to word along the time-dimension.
We also hear that the tune is moving to a goal: a place where it comes to an end, so that anything occurring after that point does not belong to the tune, unless the tune has begun again. Scales, modes and harmonic relations will make it more natural for a tune to end on certain notes – usually on the tonic, if there is one. But not every tonic is heard as a conclusion, and even when a tune comes back to the tonic by scale degrees from the dominant we may not hear it as having come to an end. Instructive in this respect is ‘Baa-baa, Black Sheep’ (Ex. 2), which returns to the tonic and then, continuing an impulse that has not been concluded, goes back to the dominant to try again. And after that final note there is no conceivable way in which this tune could be prolonged: any attempt to prolong it will produce another musical individual, and not an expanded form of this one.
That last feature is in part definitive of the tune, as opposed to the theme. Bach’s fugal subjects in the 48 are melodious, often catchy, as in 17 of Book I or 7 of Book II, and they have very definite beginnings, rarely with upbeats, since their entry is integral to their dramatic and structural purpose. But they are open at the other end, can be continued ad lib., and also concluded should the need arise. Something similar can be said of the great themes of the classical repertoire, such as the theme of Beethoven’s 9th that I discuss in Chapter 12, or the equally great but far more seductive theme which opens Schubert’s A minor Quartet (the ‘Rosamunda’) (Ex. 3). This last theme is a model, since it is so near, in its lyrical melodiousness, to being a tune, and yet so far from it. The theme begins with a phrase that leads naturally to a semi-closure on the tonic. But this is at once repeated with an interpolated phrase, making an asymmetrical variant that now refuses to come to a close: gradually the theme moves on, throwing out phrases that invite (and in due course receive) separate elaboration, and ceasing only when the music has moved into another key, and the cessation can no longer be heard as a conclusion, but only as a pause in a new and strange landscape. By contrast with such themes a tune is boxed in at both ends, and it is because of this, indeed, that tunes are so easy to remember. Adorno’s lament over the ‘predigested’ nature of the melodies that figured in the popular songs of his day seems unpersuasive to us today, partly because we recognize what has been lost by a society that has no memory-bank of tunes. As I suggest in Chapter 14, modern Pop has become so dependent on the backing for its musical movement, that it cannot really be retained as melody, and the young person who wishes to sing his favourite songs cannot sing them at work or while walking, but must wait for Karaoke Night at the pub, when he has the benefit of an external musical movement. In a very real sense he has no properly internalised songs.
By contrast it is interesting to consider a musical culture in which the tune retained a central place – a culture that endured into my lifetime, and which can still be encountered here and there in Wales and in North America – the culture of the English Hymnal. During the 18th century the hymn became the symbol and triumph of the non-conformist churches, providing to the ordinary congregation the means to memorize holy words and theological doctrines, while rehearsing their membership through uplifting song. The Church of England meanwhile kept close to the metrical psalter – that peculiar attempt to retain the open-ended flow of plainsong, without the melodic seduction. The metrical psalter emphasizes the venerable and untouchable nature of the psalms, sung above and behind the service, as though in some antique realm where the dead still congregate within reach of the living. No wonder, however, that the people began to drift away.
It was not until 1769 that the first church hymn-book for general use was published in England – though the idea of a hymn book was far from new. In 1589 Palestrina had published his own hymnal – the Hymni totius anni – in which the old plainsong chants are measured out in bar-lines and developed polyphonically for the daily use of church choirs: but of course those gorgeous settings had fallen under the same puritan interdiction as plainsong itself. The English hymn-book expressed a late and reluctant recognition that music is not a distraction from the religious message but a way of communicating it – more, a way of memorising it and integrating that memory into social life. The Wesleys amplified the hymn book in ways which also served to legitimize their evangelizing mission, and to make the Wesleyan churches into a growing counter-culture to that of the Anglican norm. John Wesley was a poet as well as a preacher; like his brother Charles, he wrote religious verses comparable at their best to those of George Herbert and Isaac Watts: all four were repeatedly set to music in the nineteenth century. The two sons of Charles Wesley – Charles and Samuel – were among the foremost composers of their day, though Samuel’s conversion to Roman Catholicism led him to write masses, motets and glees rather than straightforward hymn tunes. Samuel’s illegitimate son and namesake, however, was a leading force in the choral movement which made the Cathedral choir into one of the glories of Victorian England. His anthems and services, with their romantic harmonies and poignant English melodies, epitomize the nostalgic vision of our country which lies at the heart of the Angican faith, while his hymn tunes remain among the most memorable in the hymnal.
Welseyanism was, in fact, intimately bound up from the beginning with our native tradition of choral music-making. Hymns passed freely from the non-conformist to the official church and back again, so that, by the middle of the nineteenth century the hymn was the living symbol of the English churches, and of the unity-in-diversity which made these churches into a single and singular social force. The Church revival brought back many of the old Latin hymns, along with the more dignified products of the German Reformation. Operatic dross was discarded in favour of solid, psalm-like, pilgrim-pushing melodies, and these were again the common property of the English churches.
Hymns Ancient and Modern first appeared in 1861. It has no liturgical authority, and was quickly followed by scores of non-conformist song-books, as well as Church Hymns and The Hymnal Companion to the Book of Common Prayer. The English Hymnal was published in 1906, edited by the young Ralph Vaughan Williams, who added to it some of the most popular of modern hymn-tunes, including Sine Nomine (‘For all the Saints’) and Down Ampney (‘Come Down O Love Divine’). The first of those tunes is remarkable for its descending up-beat, which can equally be heard as a down-beat, which pushes the melody forward to the glorious ‘Hallelujah’, which is such a fitting conclusion even though rhythmically and melodically a completely new departure. (Ex. 4.)
Look back over this wealth of musical and poetic material and you will be struck not only by its vitality, but also by its extraordinary continuity: except for periods of puritanical zeal, the hymn has grown organically with the church. The hymnal, like the Bible, has been common property of the Anglican and non-conformist churches. The Book of Common Prayer, by contrast, is a specifically Anglican text, although its language permeates the traditional non-conformist idiolect. The hymns have been shaped by the liturgy; but they in turn have shaped the services of the churches. They have provided the important moments of participation, in which God is praised and worshipped in song, and the drama of the service is momentarily interrupted by the chorus. They have replaced the psalms as vehicles of collective sentiment, and the psalms in their turn have become part of the liturgy. The hymns emancipated themselves from orthodoxy by virtue of their musical power, becoming the foundation of a popular musical culture centred on non-conformist worship. The Labour movement would not have been the peaceful thing it was, without the repertoire of hymns which gathered people in congregations, rather than in armies, behind the brass-bands of the collieries.
Fragments of plainsong have survived in the hymnal – ‘Pange Lingua’, ‘Veni Emmanuel’ – and several of Palestrina’s hymns are still with us. Bach, Handel, Mendelssohn and Haydn are all represented in the hymnal. But the most important input has come from English composers themselves, and especially from the romantics and proto-moderns. A whole school of Victorian composers arose from the Cathedral choir and its musical context, and the church choir, which was the foundation of musical education in this country until very recently, helped to spread their music far and wide across the country.
The very language of English romanticism, from Parry and Stanford through Elgar and Holst to Vaughan Williams, Ireland and Bax, is saturated with the idiom of the hymnal. Song and collective worship form the background to their music; and it is hardly surprising to discover that some of our most memorable hymns come from their pens. I have already mentioned two of Vaughan Williams’s contributions; and everybody knows Parry’s ‘Jerusalem’, adopted by the brass-band movement as its signature tune. Let me mention just one other lovely hymn, since it was written by a composer whose death I remember, and from whose works I learned to play the piano – John Ireland’s ‘Love Unknown’, with its heart-rending minor third, at the point where any lesser composer would have wanted a major third to swell the melody. (Ex. 5.)
As composers felt the influence, first of folk-song, and then of Debussy, the harmonisation of the hymns changed. Significant in this respect is the Oxford Book of Carols, produced in 1928 by Percy Dearmer, Ralph Vaughan Williams and Martin Shaw, as part of a conscious return to a more folk-inspired and corybantic musical tradition. Vaughan Williams’s harmonisations lean over backwards to avoid accidentals, delight in parallel fourths, and in general make no concessions to the diatonic scale when a modal harmony seems to be implied by the melodic line. (Witness his harmonisation of ‘Away in a Manger’, which succeeds in de-kitschifying a carol that has been responsible for more puking over the years than the excesses of Christmas dinner. Ex. 6.)
Hymns Ancient and Modern has more or less disappeared from our churches; the Anglican Hymn Book, from which many of the Victorian classics have been dropped, is itself being replaced by Mission Praise, which, while retaining many of the old hymns in their original settings, shows a marked preference for the happy-clappy over the solemn and the patriotic. Of course, the editors of Mission Praise are not to blame for this: the book is a record of musical decline, but not a cause of it. Nevertheless, it is a significant record. It is significant, for example, that, while the old four-part harmonisations are faithfully reproduced, they are embellished with chord sequences for strummed guitar. It is significant that the many new additions are essentially pop songs, written in the international idiom of the TV advert, and harmonised not by part-writing and voice-leading, but with chords above a walking bass. It is significant that many of them are kitsch and that most are desperately uninspired. (Take a look, for example, at no. 264: ‘When I feel the touch’, by Keri Jones and Dave Matthews, marked ‘worshipfully’.) Why is this? It seems that tunes, unlike themes, are vulnerable to the vagaries of popular taste, and cease to be memorable when a new musical vernacular takes hold of the popular imagination. Those solid borders become flabby and permeable, and those compelling sequences somehow lose their strength. The voice of a living community is no longer heard in the old tunes, but only the echo of a community that was, while the kitsch that has replaced them reminds us of the very thing that kitsch always lacks – the firm foundation in belief, that enables people to sing out strongly and rejoice in their God.
In the decline of the hymn tune, therefore, we witness the kitschification of mass culture – the phenomenon to which Adorno devoted his most acerbic criticism, and which he believed to be indicative of the final collapse not of the tune or the song but of tonality itself. I return to Adorno’s arguments in Chapter 13 below; but one thing should be said in conclusion, which is that the way back from kitsch is not to be discovered by turning away from tunes. For without them music ceases to be sung, ceases to be a part of everyday life, ceases to exert over our daily thoughts and actions the discipline that is its greatest gift to us. It is when the soul loses all sources of inner discipline that it lapses into kitsch; the way back from kitsch is by remaking the tune, as something that we can possess in common. The great question facing the Western composer today is this: can I write a tune which is a memorable and living individual, and which is neither banal, nor pedestrian nor kitsch? And, as we learn from Messiaen, Britten and Dutilleux, and from younger composers like Robin Holloway, David Matthews and John Corigliano, the question can still be answered with a yes.
 Theodor Adorno,