Notes on Havergal Brian’s 6th symphony

Anno Schreier

Anno Schreier

Translated by Alan Marshall

Translator’s preface: German member Dr Josef Schreier, who contributed an article to NL120, produced a further article some months ago about Symphony 6. His son Anno, who has just taken up a guest studentship at the Royal College of Music, and whom we therefore welcome to London, conducted a technical analysis into the symphony as an input into the preparation of that article, and he later wrote it up into an article in its own right. It happens to make a very suitable introduction to his father’s article, which goes further into musical philosophy and abstractions than may be familiar to an English readership. Anno Schreier’s article is therefore offered first; and as usual the translator must bear in advance the odium for any misunderstandings of the original, and for any failures to be familiar with musicological terminology.—AM

As Martin O’Leary has expounded in his detailed analysis, Havergal Brian’s Symphony 6 is a one-movement symphony of three parts.

This subdivision is apparent even without knowing the score. After hearing it, and without necessarily being able to point to express separating points, one clearly remembers three parts; the expansive melody which is the principal element of the middle part, and the closing evocation of catastrophically overwhelming conflict, will linger especially in the mind.

What is remarkable is the relationship of these three sections to each other. They are not independent one from another, but the transitions are not seamless: there are clear caesuras, and despite their contrasting characters these sections are inseparably linked beneath the surface. In Brian’s case one would never look for such a sharp divide as in, for instance, the four movements of Schumann’s 4th Symphony, which are clearly separated despite fully worked-out transitions and numerous motivic references.

The manner in which this three-part structure registers on the ear—that is, how it becomes apparent that what is now sounding is actually a “new part”—has to do with how Brian allows his individual musical components to grow away from each other and how he combines them into larger units. Now this does not come about (at least not to any major extent) through motivic or thematic development, but more in the fashion of a growing and unfolding plant-life, something which conjures up associations with landscapes and may be viewed as typical of the English symphony (as begun by Vaughan Williams and nowadays taken up notably by Peter Maxwell Davies).

Of course, this is not to say that thematic development is entirely absent: for Brian especially the strict contrapuntal technique is very important. It is equally not the case that there are no violent contrasts or “shocking” surprises. Brian’s Symphony 6 is pre-eminently characterised by disparate elements, but instead of being played off against each other in a conflicting way, they are bound together as by a sort of underground stream.

Naturally this leads to implications from completely outside music; we have already been using terms like “landscape”, “underground” and “stream” which are borrowed from a pictorial vocabulary. What this means in the case of Brian’s Sixth, and the role played by that remarkable melody around which everything else seems to be grouped, are the subject of some comments in what follows here.

The first section (bars 1-57) 1 introduces disparate elements in an especially extreme manner. Three similarly-formed “openings” follow one another, having in common a motion in paired semi-quavers (initially on the cellos) and a tuba melody (in the first two openings). From the outset, any suggestion that anything is being settled or stated here is dispelled by an enormous dispersion in the shape and the sound (bare-sounding cellos moving around B, against a morose grace-note motif on high As from flute, piccolo and glockenspiel.

After only three bars the music breaks off, and not even the tuba melody can get it flowing because this melody, incapable of development, consists merely of reductions, rising in pitch and volume, of its descending opening motif and therefore leads straight to a breaking-off. The second opening breaks off in a similar way, and the third no longer has the tuba melody, but in its place a strangely angular crotchet motion, somehow caught in its own remarkably awkward jumps, on glockenspiel and pizzicato violins or basses. However, at the end of each “opening” a sustained brass chord, coloured by glissandi or figurations on the harp, provides temporary release for the accumulated energy.

As O’Leary writes, the Prologue does not establish a character but creates an atmosphere. The tuba melody misses the chance of becoming a theme because of a sort of vegetative over-determination—stolidly confined to reduction and sequencing as its development principle; the semi-quaver motion on the strings, while an element running through the whole, meanders oddly through the landscape. To elaborate O’Leary’s findings one might say that the landscape in which the subject will appear is now present, but the subject itself has yet to arrive.

The sole arresting moment is a trumpet fanfare which breaks into the second “opening” (bar 17, repeated in bar 21 a whole-tone higher). In this context it is an alien element, having almost the effect of incidental music—similar to the Leonore fanfare, ie indicating an event outside the music, possibly an actual event on stage; incidental music, nevertheless, to something which is not actually there yet. If there were a “scene” here it could at most be an “imaginary scene” in the Berlioz sense, but there is no sign of any sort of subject, or any action, which this could be incidental to.

The extent to which this fanfare might actually have been incidental music in the stage action for which the music was originally conceived, and then left as such in the symphony, is something that would bear investigation. What is clear though, is that in the symphony (especially at this point) its function is a puzzle or, perhaps, that it is precisely its puzzling nature which justifies it, ie by indicating something hitherto unknown. ’Leary places the opening of the second section in bar 57 with the appearance of a new motif on the cor anglais. But what begins here could equally be seen as belonging to the previous section 2, because very little of a concrete nature is formulated here either. The cor anglais motif wearies on eleven repetitions of a single note (here E flat) and a little decoration looping down from a higher note, runs aground in a descending line of crotchets on bassoon, thereafter does not make it beyond monotone repetitions, and finally settles on swelling E flat octaves (horns and tremolo violins), suddenly broken off by a hard orchestral stroke which is completely contrary, both harmonically (E) and in sound (flutes, oboes, horns, harp, timpani, pizzicato strings)—producing one of the most hard and startling contrasts of the whole symphony. The cor anglais motif, now in the deeper voices (bass clarinet, then bassoon and cellos/basses), is then subjected to a similar “development” as earlier the tuba melody, breaking it down steadily into smaller elements and leading naturally to a breaking off, as before. A further section picks up the semi-quaver motion, then the opening motif of the tuba melody is passed canonically to and fro between oboes/trumpets and trombones/tuba/cellos/basses. This passage culminates finally in the empty fifth E/B which is retained in the next section as “shadow” in the high and low strings.

In this next section there are renewed fanfare sounds having the effect of announcing something. Whatever is being announced, though, it can hardly be the unison flute melody, nor the rising crotchet motions in octaves supplied by the low strings and harp, for three side-drums cut across with march-like rhythms, which at this juncture point into a void much as does the opening fanfare.

It is rather the case that these crotchet motions provide a foretaste of the accompaniment to the main melody, already mentioned, which begins at bar 111 after a cadence, in trombones and tuba, of remarkably conventional effect which establishes, for the first time and out of nowhere, a clear tonic (E flat major).

The start of the melody derives much of its surprising effect from the fact that the expectation aroused by the fanfares is disappointed by the anticipation of the (later) accompaniment structure, combined with the confusion produced by the drum rhythms. Finally, the strange plagal cadence on the trombones does not just prepare the tonal ground for the melody—its unambiguously recognisable tonality makes that superfluous—but it also re-establishes the sense of announcement which the fanfares gave, though in a way which is far more unexpected and “underground”.

The melody itself is carried by unison violins (muted, on the G string). It grows from a descending series of notes, a transformation and dissection of the tuba melody from the inversion of which the accompanying crotchet motions have an upwards direction. The melody does not only provide the symphony’s first unambiguous tonality, maintained for a considerable time, it is also the first instance of a musical statement over an extended period of time—a melodic line undisturbed by any interruptions.

The character of the melody consists in a number of special qualities. While it is undoubtedly in E flat major, it generally avoids melodic figures implying clear harmonic progressions (for example, progressions via leading notes). Where such figures do occur, they are compensated for in the accompaniment (about which more later), so that the motion becomes soft (instead of being made “hard” by the dictates of harmonic relationships).

In the same way the melodic-rhythmic phrasing, at both the small level and the larger, defies any attempt to apply a consistent beat or even to break it down into regular periods. On even a superficial glance at the score one is struck by the continual changes between 3/4 and 4/4 (and the ties in and from bars 112-114).

Instead of being determined by the “given” stress pattern, the melody develops out of itself, as it were from one note (or group of notes) to the next. So we are dealing here with a purely horizontal development, a discursive one as it were. Its antithesis, a vertical one, would be where the melody itself plays a lesser role than the harmonies it implies. 3

One may therefore say that the melody has a “plant-like growth”, similar to a river which makes its seemingly unpredictable way through a landscape.

The melodic development is mostly impelled by free suspensions in “soft”, broad seconds. Even the opening line, descending from F to B flat in calm crotchets, can be regarded as a chain of suspensions, each resolving the last and finding a temporary repose in the low B flat. The melodic rhythm immediately doubles in speed, the melody leaps an octave to the higher B flat and falls again via F to E flat to leap immediately even higher to C until it drops an octave to rest on the lower C. Rising with irregular rhythm first to E flat again, then descending from A flat via G to F, the melody finally returns to the low B flat via a conventional triadic figure, ending its first “period”. 4

Now if we consider the first period quite divorced from its accompaniment and look for the melodic “centres of gravity”, so to speak, then the first substantial one is the B flat in bar 112, then (less substantial) the C in bar 113 and the A flat which is resolved by the F (bar 114), finally the last substantial point which is the closing B flat. It is striking that, apart from the dominant B flat, none of these central notes belongs to the E flat triad.

As the melody unfolds, the notes G and E flat are only briefly touched or appear as transitional notes (eg the triadic figure in bar 115). This development, which apart from the closing triadic figure is downright anti-triadic, imparts to the melody a pseudo-modal character (that is softer and more flowing than a pure, hard E flat tonality), so that, given the central role of B flat, one might speak not so much of E flat major as “mixolydian B flat”.

The modal coloration is substantially strengthened by the accompaniment. Brian sets rising crotchets (harp and low strings, three voices) against a melody in which falling figures dominate. These crotchets do indeed correspond (inverted) to the beginning of the melody, but with their rhythmic and melodic regularity they oppose the melody, with the phrasing often not coinciding with that of the melody.

But much more important are the harmonic factors introduced by the accompaniment. Because it too is conceived as a linear movement, it has harmonic structures which follow its own logic and are largely independent of the melody. These structures are of a decidedly modal nature, that is, there are few major or minor triads, rather mostly layers of fifths, fourths and diminished sevenths which, compared with conventional functional harmony, establish no tensions between the chords but exist in a kind of tension-free stream. Notable is that the accompaniment also deploys the mixolydian seventh D flat, so that one may speak of simultaneous E flat major (or mixolydian B flat) in the melody and mixolydian E flat in the accompaniment.

What is decisive about this is that it further obscures the harmonic sense of the melody, especially at those points like the end of its first period where that sense threatens to assert itself clearly (in the triadic figure G-E flat-B flat). If one were looking at the melody on its own, one would be expecting at this point an imperfect cadence in E flat major, that is, a progression from E flat major (tonic) to B flat major (dominant). Instead of the E flat major a fifth appears—similar indeed to E flat major—but the dominant is replaced with an entirely different D flat major sixth, to which the B flat in the melody adds a modal sixth.

In similar roaming fashion the melody goes on to reach still higher, the accompaniment brings yet stranger notes like E * (bar 117) or G flat (bar 120), transferring finally to trombones and to clarinets and bassoons.

From bar 125 the melody goes from the E flat major, disturbed only by the G flat in bar 120, into a sort of modal hybrid between E flat minor and G flat major, leading to an unexpected conclusion on the empty fifth C flat-G flat (bar 129), the incomplete subdominant of G flat major. This “roaming” leads, therefore, to a melodic path which is neither expected nor foreseeable at either the small level or the larger.

After an intermezzo with contrapuntal development of the cor anglais motif (see above), Brian repeats the whole melody in E flat major. Now carried by a larger orchestra (all woodwind, horns, all strings), and furnished with contrapuntal, sometimes canonic counterparts, the melody loses most of its earlier modal character because it is now being given relatively conventional harmonies (so far as that is possible with the unusual melodic twists.

Going back to the initial thoughts of this paper about the presence or absence of a subject, the melody might stand in fact as the expression of a subject which was missing in the first part of the symphony. But the subject expressed by the melody is not the one which the martial elements (fanfares, drum rhythms) might have led one to expect. What comes across here is a restrained, gentle subject (O’Leary, with an eye to the title of Brian’s planned opera, Deirdre of the Sorrows, sees it as careworn), and there is no trace of a Beethoven-esque per aspera ad astra principle, not of a changeable adventurous subject as in Berlioz’ Harold in Italy for example.

If one seeks to interpret the picture of battle, comprising the third part of the symphony, as the conflict of the subject with the world (in much the sense of Richard Strauss’ Ein Heldenleben), then it must be noted that the subject now formulated (the melody) is not confronted with any opposition because it does not occur at all in this part.

The motivic structure of the battle picture (along with the usual military allusions) is made up rather of the tuba motif from the beginning of the symphony and the cor anglais motif (though they are motifs related to the melody). The conflict represented by the battle picture reaches a negative close, as underlined by the funeral march episode between bars 189 and 198.

It seems a valid thought that the short cor anglais phrase (bars 193-196), which makes a resigned, downward-moving appearance in the midst of this funeral march, could be interpreted as being derived from the melody; consider the falling succession of notes with which both begin. That might indicate the “defeat” of the subject, and the closing gentle tutti passage (from bar 299), though accompanied below the surface by march rhythms on trumpet and side-drums, might be a kind of transfiguring resolution. The cor anglais motif, now on violas (bars 304-305) does indeed have the last word but can hardly assert itself against the rest of the orchestra.

The gentle nature of this final section recalls the melody and even contains something close to a quotation from i —cf the first oboe and first horn.

None of this alters the fact that the melody in the middle of the symphony appears as an anomalous block between ungentle disparate elements. It immediately gives up the connection it has established, and because there is scarcely any mediation between it and the disparate elements, the ultimately negative outcome of the battle cannot simply overcome the significance of the melody.

If we relate this to our initial view of the natural concept underlying this melody, then perhaps this indicates an eternal value hidden in nature.

  • Translator’s note: This is one of Brian’s ambiguous note-heads and looks rather like an E though D seems more likely.

© Anno Schreier 2003 NL168

translation © Alan Marshall 2003

  1. O’Leary calls this section a Prologue ↩︎

  2. O’Leary has presumably arrived at his subdivisions because the three “openings” belong together in a formal sense. ↩︎

  3. For example the variations theme in Schubert’s string quartet Death and the Maiden. ↩︎

  4. If one considers the first three notes of the melody (bar 111) as an extended upbeat, then the first period does in fact comprise 4 bars, albeit one of them a three-quarters bar. ↩︎

Newsletter, NL 168, 2003