The composition draft of the War Requiem (GB—Lbl Add MS No 60609) comprises 102 folios of 24-stave manuscript paper, measuring 36.5 x 27 cm, bound in blue hoards with a leather spine and corners. It is dated on the final page: 'Aldeburgh Dec. 20th 1961'.
A few general words on the nature of Britten's draft and its related discarded leaves are necessary, although it is impossible in the context of the present discussion to explore these documents in exhaustive detail. The pattern of Britten's working methods hardly ever varied and the disciplined routine adopted in early adulthood was maintained, with only very few exceptions, with unfailing regularity. Britten's business-like timetable for the working day is well attested: two main periods of composition at his desk (not his piano), one in the morning, the other in the late afternoon/early evening, framed a long 'thinking' walk after lunch. He always mistrusted working at night, although scoring might be undertaken then if a dead line were fast approaching. By adhering to this rigorous schedule, Britten was usually able to judge the amount of time needed to complete a major composition with unnerving accuracy.
As with the vast majority of Britten's output, the War Requiem took shape on the manuscript paper as a through-composed short score draft written in pencil throughout, with the orchestral texture reduced onto two, three or occasionally four staves and the vocal lines occupying their own staves. At first glance the draft resembles something approaching a vocal score and was certainly used as a guide by Imogen Holst when she prepared the vocal score (under Britten's supervision) for Boosey & Hawkes. The instrumentation is indicated by verbal abbreviations—'str', 'trbn', 'ww'—at the time of composition, ready for instant retrieval when the moment came for the full score to be made. This simple technique was effective in allowing Britten to press on to the end of a work before making the full score, safe in the knowledge that the piece was in effect written; in the case of complex works, such as the War Requiem, it also allowed an assistant to follow behind the composer using the draft as a basis for the all-important vocal score from which the soloists and chorus would learn their music. Once a new piece was complete in draft form, the business of making the corresponding full score was largely a calligraphic labour; the vast majority of problems had already presented themselves and solutions been found.
Throughout his career Britten preferred to work in pencil when composing because of the freedom it offered for changes of mind; the use of pencil was a significantly liberating factor to the composer's creativity since anything that was committed to paper might be easily rubbed out and rewritten. Almost every leaf of the War Requiem draft shows evidence of the use of the eraser, demonstrating unequivocally how closely Britten tested his original ideas. By and large it is not very easy to read the rubbed-out notes, although occasionally one can discern the impression made by Britten's pencil and decipher something valuable.
Two other, related, methods of making changes to a work can be found on any of Britten's composition drafts. Rather than erasing passages, particularly if they were more than a few bars, Britten would cross-through a section which he wished to delete. Occasionally, if the deleted passage amounted to a full page, he might detach it from the bifolium and use the available blank verso elsewhere in the manuscript draft. If more than one page were rejected then Britten would almost always remove the offending passage from the main draft and place the particular folio(s) to one side. These discarded leaves form an important category of substantial earlier versions of the music.
Much of the foregoing holds true for Britten's methods while working on the War Requiem, and his work on the draft generated a quantity of discarded leaves. In two places in the work—the settings of Owen's 'Anthem for Doomed Youth' and 'Move him into the sun'—the number of discarded composition-draft pages is, even for Britten, unusually large and varied in scope, suggesting that the composer experienced greater difficulty in defining and refining his ideas than was customary. What is undoubtedly embodied in all the draft pages for the War Requiem, however, bears testimony to just how remarkably successful Britten was at planning out his music in advance of committing the actual notes to the page: in effect, the long, thinking walks were where much of the compositional thought processes sifted through and refined musical ideas. The composition draft and accompanying discarded pages (in the case of the War Requiem, the majority of the latter are bound with the main draft) offer, even at best, a relatively and necessarily limited view of the workings of the creative mind; they only represent the tangible by-products of Britten's skilful and inspired art in bridging the differing, often disparate, elements that make up the complex compositional process.
A discarded version (eighteen bars) of the opening of the work exists (fol 71r–71v), which approximates to the published version as far as seven bars after Fig 1. Many features of the final version of this memorable passage are present, including the funereal gong strokes, the tolling bells (F sharp–C) and the solemn, almost lumbering, orchestral march, though occasionally the detail of the pitches differs. Unlike the published score, this discarded sketch has uncomplicated homophonic choral entries in octaves (S+T; A+B) intoning the timeless words of comfort. Only at 'dona eis Domine' does Britten introduce overlapping vocal entries, shortly after which the sketch breaks off. This change of approach clearly set off in the composer's mind the possibility of introducing such an effect from the beginning of the movement, a detail which offers far greater atmosphere.
Fol 72r comprises what is probably the first notation of the boys' choir's entry at 'Te decet hymnus', a sketch that corresponds to the music between Figs 3 and 4. It is a true sketch page and never formed part of the main composition-draft sequence. The page contains two versions of the passage in question, the second a development of the first. Neither quite adheres to the vocal line as published, but both are demonstrably close to it. Both sketches incorporate a chordal accompaniment which points the way to the final draft, the second being the nearest. The second version has two sets of accompanying chords, the first of which is deleted. The second (later) sequence also includes a hint of the ostinato pitches (C–F sharp) played by the main orchestra in the final version (see fol 72r). This sketch page also contains an example of what proved to be a familiar Britten habit: the writing down of a full chromatic scale with the pitches crossed through as Britten used up the notes (or the chords they represent). The impersonal quality of the boys' choir is accentuated in 'Te decet' by the use of a sequence of chords each one of which is based on a different pitch. In effect, it is a subtle use of a twelve-note proposition in a wholly—and wholly characteristic—tonal context.
In spite of sketching the boys' first entry, Britten still evidently found dissatisfaction with his earlier thoughts on this passage: fol 69r–69v is a 24-bar deleted section corresponding to five bars after Fig 3 to seven bars after Fig 6. This leaf corresponds to fol 5r–5v in the main draft, and we note the twelve crossed-through bars on fol 4v which represents a still earlier continuation of this material.
Evidence of frequent use of an eraser on the pages containing the first Owen setting (‘What passing bells’, for tenor) suggests that Britten experienced some difficulties in drafting this setting. The struggle to find an appropriate expression for Owen's haunting words produced no less than five discarded attempts, not all of which are complete. These rejected versions may be found on fol 73r–77v. While a definitive chronological order for these leaves is difficult to establish, they may be summarized as follows:fol 73r–73v
A draft that corresponds to the passage between Fig 9 to six bars after Fig 11. The vocal line begins with repeated Cs familiar from the final version but with a changed rhythmic profile. The harp's 'bisbigliando' and the shapes of the woodwind flourishes are present; however, the mocking dotted figure in the strings is missing, with a consequent lack of energy to the setting. The 'anger of the guns' is depicted by side drum only, ie without the strings and bass drum which make such a memorable contribution in Britten's final version.
A further draft of the same passage which may be an earlier version of fol 73; the pair of leaves have many ideas in common. The vocal line begins in a similar fashion to the draft notated on fol 73, but quickly diverges from it. The 'anger of the guns' is here represented by a simple quaver figure on strings; no percussion is indicated.
A further draft corresponding to the previous leaves, though it is undoubtedly closer to fol 73, which may suggest that in fact fol 74 is the earliest. A version of the repeated string figuration at 'anger of the guns' is present. Fol 75v contains sketches for what finally became the string figure at Fig 9, material that was incorporated onto the version drafted on fol 77r–77v. Further sketches for this material were made on fol 76r.
Another attempt at the same passage in which the vocal line is clearly becoming closer to shapes recognizable from the published version. The harp writing diverges from the final version in this draft.
An extended draft corresponding to the passage between Figs 9 and 12. It is more extended and 'finished' than any of the previous attempts, and has most of the features of the final version (fol 7r–9v). Unless an intermediate attempt is now missing, this draft must surely have preceded the final version.
Fol. 70r (from the ninth stave down) shows an exact transcription of a sketch of the brass fanfares that open the movement, a sketch which most probably represents the first written form of this seminal material. As the sketch shows, Britten made two attempts at the passage; the final version combines elements from both. The return of the fanfares at Fig 22 (after 'Coget omnes ante thronum') caused Britten a moment's thought as evinced by a rejected draft of this section of the movement (fol 80r).
One leaf of the baritone's first Owen setting was rewritten (compare fol 81r–81v, the discarded draft, with fol 14r–14v), although the differences are minimal when compared to the difficulties Britten found with the tenor's first Owen setting. Elsewhere in the 'Dies irae' there are minimal (and relatively uninteresting) redrafted passages, sometimes of only a few bars' duration: for example, in the soprano's 'Liber scriptus' (her first entry), the chorus entries at 'Quid sum miser tunc dicturus' and in the tenor-baritone parodying duet, 'Out there, we've walked quite friendly up to Death'. A more substantial reconsideration took place at the 'Recordare' where the orchestral introduction (Figs 39–40) originally included a version of the 'Recordare' theme on brass, thereby making the theme's relationship to the opening fanfares particularly clear (fol 82r–82v).
The alteration of the 'Lacrimosa' and the tenor's 'Move him into the sun' required some redrafting before Britten was satisfied, in particular the Owen setting of which the many abandoned drafts are comparable to the level of rethinking that took place in the 'Requiem aeternam'. A deleted version of the close of this section exists within the main body of the composition draft in which Britten sets the words of the 'Pie Jesu' not as in the published score–ie as a reprise of the cadential figures first heard at the close of the 'Requiem aeternam'–but to the 'Lacrimosa' motif with a homophonic, almost Verdian parlando chanting of the words 'dona eis requiem' on the all-important pitches of C and G flat (=F sharp), which clearly relate the material to the central tritone axis of the work. Part of the original ending to the movement can be found on fol 29v (fourteenth stave onwards).
Among the discarded leaves can be found several attempts at the setting of 'Move him into the sun'; each begins in a similar manner, ie with the chorus and solo soprano's 'Judicandus homo reus' (four bars before Fig 56). The different attempts may be summarized as follows:fol 84r (from sixteenth stave)–84v
Most probably the first attempt. The vocal line and harmonies differ from the published version; in addition to the string tremolando (present in the finished work), Britten included echoes of the tenor line which he subsequently excluded. The draft breaks off around the words 'Always it woke him' (there is no text present in this draft.
In addition, fol 84v includes an attempt at the vocal line alone, which proves closer to the rhythmic profile of the final draft than the draft made on fol 84r–84v.
A substantial draft of the same passage as (1) above, in which the orchestral imitation of the vocal line is present but here more convincingly formed. This draft breaks off three bars after the chorus's interpolation of 'Qua resurget'.
Another substantial attempt at the same passage in which Britten moves closer to his final version (the orchestral imitation is now expunged). This (third) attempt reaches the end of the movement with a first version of the choral 'Pie Jesu ... Amen' echoing the tenor soloist. Britten may have abandoned this idea as it draws together the ritual grief of the Latin text and the song's real expression of grief (the soldier of Owen's poem is beyond consolation) in a way that was to be achieved more effectively in the 'Agnus Dei'. Britten's second attempt at the close of the 'Lacrimosa' (fol 29v) incorporates the soprano's 'weeping' motif.
The draft of the compact 'Offertorium' is freer of evidence of the use of the eraser than any of the previous sections of the work. This is, perhaps, to be expected as Britten is reworking previously composed material for a substantial part of the movement (ie the incorporation of material from Canticle II: Abraham and Isaac). Discarded drafts are extant for the opening of the movement (as far as four bars after Fig 61), for the baritone's entry (‘So Abram rose ... '), and for the boys' 'Hostias’. The latter (fol 89r) is a genuine sketch of this material, devoid of its relationship to the Owen setting.
A pair of discarded drafts of the opening of the movement survive. One, which is a mere five bars long (fol 90r), breaks off at the soprano's melisma; the basic material, however, is very similar to the published score. The other rejected draft is in private ownership and comprises a single leaf which takes the movement through to the first bars of the choral 'Pleni sunt coeli' (the leaf is a first draft of fol 41, the same passage in the main composition draft). The most interesting divergence from the published score concerns the setting of the words 'Pleni sunt coeli et terra gloria tua', which the divided chorus freely chant on approximate pitches; in his first draft Britten specified that the pitches were to be sung in the conventional way. Even in his subsequent draft (fol 41) it is possible to see that the chorus parts' noteheads have been crudely altered to accommodate this change of mind.
Further unsuccessful drafts from this movement include the transition from the 'Hosanna' to the 'Benedictus' (fol 91r–92v), and the close of the 'Benedictus' leading back into the reprise of the 'Hosanna' (fol 90v). In the latter Britten included further material for the soprano soloist.
No discarded leaves exist for this brief movement; nor are there any deleted passages within the composition draft itself.
An early draft of the opening of the movement (fol 93r) survives in which the choral entries do not follow the shape of the final version, although the percussion/string march is present virtually in its ultimate shape. In the published version of this passage the choral line (built around semitones) relates more clearly to earlier symmetries within the piece. Britten made several attempts at the soprano's 'Tremens' passage, the earliest having a waltz-like orchestral accompaniment (fol 94r–94v). The reprise of the 'Dies irae, dies illa' was also subjected to some reconsideration, whereas the final, long Owen setting ('Strange Meeting') appears to have troubled Britten very little.
War Requiem, Op. 66 © Copyright 1961 by Boosey & Hawkes Music Publishers Ltd. Reproduced by kind permission of Boosey & Hawkes Music Publishers Ltd.
Photographs of British Library manuscript Add MS 60609 courtesy of the British Library, whose Britten manuscripts are digitised at www.bl.uk.
Text from 'The War Requiem in progress' by Philip Reed in Britten: War Requiem by Mervyn Cooke. Cambridge University Press, 1996. Fig. references are to the published War Requiem score, which you can browse online at Boosey & Hawkes' online scores service (requires simple and free registration).