‘(Braut helped a great deal to make these little tunes)’: Elgar's modest inscription on the four-hand piano arrangement of his Serenade for Strings was an affectionate acknowledgment to his wife, Alice, at the time of their third wedding anniversary in May 1892. The absence of an autograph full score of the complete work makes the first print by Breitkopf und Härtel, published the following year, along with well-prepared parts, into a principal source. Introducing his meticulous new edition of the Serenade, Christopher Hogwood draws on these early, authoritative printed sources, along with three extant manuscripts: an autograph full-score version of the first movement (now held at the Elgar Birthplace Museum); a draft of the third movement, on two folios, showing Elgar's revised ending, including his apparently last-minute decision to conclude the work by cyclic return to the opening movement (held at the British Library); and the autograph piano-duet score already mentioned (also held at the British Library), itself the basis for the Klavierbearbeitung published by Breitkopf in 1893 alongside the orchestral score.

In his Introduction, Hogwood notes the draft status of the first-movement manuscript (labelled ‘1st copy’), which is evident in details of rhythm and articulation. For example, the viola's siciliano dotted figure persists as an inner voice (in bar 4) in this version, but was omitted in the printed score; and there is a layer of pencilled slur annotations, ‘some carried through to the printed edition and others overlooked or discarded’ (‘Introduction’, iv). Given Elgar's habitual care in preserving articulation and dynamic indications of the original when making such transcriptions, the piano duet manuscript offers valuable corroboration and clarification where questions linger in the print score source. Other details in the duet score are of interpretive and expressive interest: for example, Elgar's Adagio marking for the tender second movement, which in the print becomes Larghetto. In the Critical Commentary (pp. 20–22), Hogwood usefully collates many variants. His edition ends with a three-quarter size facsimile image of the last page in the autograph piano score, showing the cancelled original ending in G major and expression marks that betray its transcription from a string version.

Another recent edition of the Serenade, that prepared by Julian Rushton for volume 24 of the Elgar Society's ongoing Complete Edition,Footnote ¹ reproduces part of the ‘1st copy’ manuscript in facsimile, along with a source not mentioned by Hogwood: an early version, laid out for string quartet, of parts of the first two movements. The slow movement (again marked Adagio) is notated in rather than its eventual form (in which Elgar halves all note values). Both Hogwood and Rushton mention an earlier work, Three Sketches for String Orchestra, played in Worcester on 7 May 1888 and usually considered a likely precursor to the Serenade, though the manuscript has not survived.

Rushton's prefatory descriptions of sources, as one might expect from a hardcover Gesamtausgabe, are more detailed than Hogwood's,Footnote ² though a comparison of the actual scores yields comparatively minute distinctions. For example, in bar 4 of the first-movement violin melody, Hogwood indicates the same two-slur division of the bar as Rushton, but he also suggests a whole-bar slur (dotted), to point up the articulation Elgar uses at the later return of the same music (bar 93). In bar 12, likewise, Hogwood prints a hairpin diminuendo missing in cello and contrabass, matching the phrasing of the upper voices (drawing on the piano-duet score), where Rushton, hewing more closely to the printed full score as copy text, does not correct the oversight. Reaching Elgar's turn to the deliciously loping E-major violin tune of the first movement, Hogwood prints dolce, a marking found only in the draft first-movement score; the initial up-bow mark (at bar 46) is from the same source; Hogwood also includes occasional string fingerings, collating both the draft and the Breitkopf parts.

In the Larghetto, similarly, Hogwood's edition clearly aims for unambiguous directions to performers, as when the first of the chordal interpolations in A minor (at bar 34, and returning at 41) is marked with bracketed pp dynamics (not included in Rushton's edition). The music four bars earlier is already pianissimo; one might object that Hogwood's dynamics are intrusive or redundant, until one notices that the second-violin responses (at bars 34 and 36) are louder – mf, then p cresc. The music's strongly antiphonal effect – an ecclesiastical semi-chorus, almost – is heightened considerably by the dynamic contrasts, which are lacking in all but the piano-duet autograph. One could argue that Hogwood's reading promotes a distinction Elgar made when transcribing for keyboard, one lacking in the printed orchestral score. Yet Hogwood's choice seems to offer a dramatic contrast that might be lost in a narrower reading of sources. The boundary between editorial consistency and interpretive intervention is by no means clear, surely, in music this rich in local nuance. Another case in point, also from the Larghetto: the long-limbed second melodic strain (beginning at bar 17, and running all the way to bar 32) is marked dolce by Hogwood (and Rushton) in accordance with the printed orchestral sources; but his Critical Commentary also notes the label più tranquillo in the published (but not in the autograph) piano score. In the absence of Elgar's autograph full score, we cannot tell if such expression markings originated with the composer himself or with a Breitkopf editor; still, it is good to know about such details, for they bear directly on the refined inner life of the Serenade.

Inscribing the piano-duet score with a reference to his wife's influence on ‘these little tunes’, Elgar may have been speaking more of his debt to his wife – Alice, surely, had helped ‘a great deal’ – than of the relatively modest dimensions of the Serenade. Questions of dynamics and articulation go hand in hand with music of considerable structural refinement. The sculpted elegance of the opening melody of the Allegretto piacevole is a source of its emotional richness, even while the surface appears to present relatively untroubled melodic utterances. If the opening segment (through rehearsal letter B) adds up to 32 bars, the internal phrase patterning is less than conventional. Two of those bars, in fact, are short bars, gently pushing the music forward, as at bar 10, where an apparently balanced parallel period phrase is suddenly short (three-and-a-half, not four bars), the hasty cadence serving as a springboard for a two-bar extension, itself dovetailing into the new skipping theme (at letter A, bar 13). That theme offers four-square regularity of phrase length (bars 14–21), but as a contrasting ‘middle eight’, it is succeeded by Elgar's sf return (bar 22) to the loose-limbed metric irregularity of his original first theme. A new phrase (bars 22–26) develops the earlier extension idea by stretching it to five bars, and is followed by a five-and-a-half bar closing phrase, reharmonizing the original head motive, then dying away in a modally inflected cadence. The balance here is emotional – a matter of earlier impetuosity answered by written-out languor – rather than strictly metrical.

In music with this degree of phrase-level give and take, every expressive detail counts, and one scrutinizes each of Elgar's score markings for clues as to meanings pictured within. Take the sudden freezing of the melodic flow in bar 25, a moment of local withdrawal where a sustained E–G minor third in the violins is held for two full bars and all other instruments drop away, in the kind of Elgarian gesture Diana McVeagh aptly characterizes as ‘half-shy, half-impulsive … a shrinking for an instant into a secret self’.Footnote ³ The violins are marked sf with an accent, but what of the succeeding dim. indication? Does it begin immediately, intensifying the sting of the short (>) accent, or does the dim. arrive just slightly later, or even a full bar later, with the return of a bass line, breaking the stillness and ushering in a cadence? Hogwood's score prints the dim. immediately after the sf; in Rushton's edition, its arrival is delayed and stretches over the bar line; the two possibilities are subtly different, but the difference has expressive meaning. Similar, possibly insoluble questions could be raised here, concerning for example the hairpin decrescendo sign accompanying the dramatic falling motion of the melody at bar 28: should the hairpin cover the whole bar (as Hogwood prints it), taking us down quickly to a concluding pp, or does it come subtly later in the melody (as in Rushton, but only in violin I itself, for the violin II part has an independently placed hairpin of shorter duration)? One could identify comparable subtle distinctions in the placement of dynamic swellings and lulls, comparing Hogwood's score with Rushton's, not least in the emotional heart of the Serenade, the Larghetto.Footnote ⁴

Elgar's habit of giving detailed expression markings might have originated in pedagogical necessity.Footnote ⁵ He first rehearsed the Serenade in Spring 1892, with his Worcester Ladies’ Orchestral Class, a group of amateurs, one of whom glibly reported: ‘He's always writing these things and trying them out on us’.Footnote ⁶ To what extent should reviewers attend to the strictly local notation of gestures realized anew with each performance? Hogwood's edition shows every sign of care with the expressive details, and beside standard dynamic and articulation marks, he singles out ‘better length and placement’ of hairpin crescendo/decrescendo signs as a specific feature of the composer's piano score. Hogwood's edition is relatively compact in visual layout (covering 18 pages, to Rushton's 26, for example), though I noticed very few problems of typesetting.Footnote ⁷ And it is surely in the local gesture, the characteristic way of doing things, melodically and rhythmically, that Elgar's ‘little tunes’ carry their distinctive expressive force. Hogwood's thoughtful introduction credits the Zürich Chamber Orchestra for help road-testing his new edition, and he notes Elgar's own relatively loose approach to tempo in the Serenade (as conductor in his very last recording session, in August 1933). This music lives and breathes in its characteristically ‘stringy’ (the composer's own word) effects – an adjective one might apply not merely to the timbral effect of sustained notes on the open G string, boldly opening the violins’ melody in the Larghetto, but also to questions of string texture more broadly: the gossamer-light violin/viola triplets over which the second-strain melody returns in the Larghetto (at bar 53), or the cheeky but delicate pizzicati in violin and contrabass, adorning the main theme's sordini return in the violas (bar 67).

The term ‘stringy’ might refer, equally, to those moments at which a string orchestra can evoke the passage of a human breath and the affective/expressive energy of a sigh, or those many signs of musical rhetoric that together signal the welling-up of feeling, which is where Elgar's meticulously notated hairpins come into view. Confidently brandishing the composer's own recording, and with a relatively continuous performance history of a century to look back on, we may assume we know how Elgar's music goes: how, precisely, he wanted those curving melodic shapes to move through the air. Yet in music of such poignant melodic detail, ‘gesture’, however meticulously inscribed on the score page, will always exceed and escape the outlines of its notational trace.Footnote ⁸

No doubt someone at Novello lived to regret their reviewer's short-sightedness in rejecting the Serenade for publication in 1892, though the work was judged ‘very good’ (‘We find however that this class of music is practically unsaleable’). Breitkopf und Härtel's decision to take the piece may well reflect (as Hogwood suggests in his introduction) the popularity of string ensemble works with late-nineteenth-century German audiences, and the championing of such scores by conductors including Weingartner and Richter. More than a century later, the appearance of Hogwood's valuable new Bärenreiter edition of the Serenade for Strings provides performers with the kind of inexpensive scholarly-critical text that may continue to stimulate performances of Elgar's music worldwide.