Top of page

Collection The Moldenhauer Archives - The Rosaleen Moldenhauer Memorial

Arnold Schoenberg's Adagio for Strings and Harp

This complete, single-movement, lovely little piece in A-flat, for an unusual combination of solo violin, small string ensemble and harp--twelve players in all--immediately brings to mind the slow movement of Brahms's F-Minor Piano Quintet (whose tempo indication, by the way, is Andante, un poco Adagio), though the resemblances are subtle rather than blatant. Perhaps the connection is implied through an initial melodic span of thirds--tonic to third in Schoenberg; third to fifth in Brahms--and stepwise motion, up and down within that span, which is gradually expanded. Perhaps it is the indirect rhythmic kinship between them; or maybe it is just the registrally central, reiterative middle C in both melodies that suggests some "subliminal" recollection working in Schoenberg's mind, for surely his familiarity with the chamber music of Brahms must have gone back to the time when he composed this Adagio.

Adagio, for string orchestra and harp, [ca. 1895]
Reproduced by permission of Lawrence Schoenberg

Fifty measures long, straightforward in design, the piece unfolds by gradual registral expansion and a modicum of harmonic development abetted by the consecutive appearances of the violin's solo line and the harp's arpeggiations, each underlining both the dynamics' increase and a growing textural density, and all of it subsiding at the close--though the violin, appropriately, has the final ascent to its previously established highest A-flat. While the piece would be memorable for its eloquent tunefulness, its smoothly drawn lines, and the skill shown in its instrumental writing, as characteristic of a very remarkably accomplished young composer, that composer's unmistakable identity is made manifest by the manuscript itself, for not only is the handwriting immediately recognizable, but a number of features on those three pages add up to a view of the twenty-three or twenty-four-year-old Schoenberg for whom the massive achievement of Verklärte Nacht, composed no more than a year or two later, in some ways seems more plausible than it did before we knew of the existence of the Adagio. The very fact of a string ensemble of more than four parts suggests technical resources, necessities, and availabilities that his several experiences of the time in composing for string quartet had largely excluded, and that therefore lay bare a convincing connection to the future String Sextet, op. 4. One need but consider the prevalence of octave duplication of melodic strands--as in measures 29-35 (Vl.I and Vcl., with Vl. solo adding the highest octave in measure 35); measures 25-28 (Vl. solo and Vl. II); and measures 35-36 (Vl. solo, Vl.I., Vla.--also octave tripled).

The manuscript remained unknown and unmentioned until very recently, when its existence in the Moldenhauer Archive was disclosed. The only printed notice about it appeared in a one-page article in the Österreichische MusikZeitschrift1 along with a reproduction of the first page and an altogether preposterous assumption that the year of composition was 1904 (i.e., after Verklärte Nacht, much of Gurrelieder, Pelleas und Melisande, and several of the Lieder, op. 6!).

Enhancing the delight in reading and playing through the score, there are a few tantalizing crossings-out, penciled additions, and changes in this manuscript that call for some scrutiny.2 To begin with, the tempo indication, which had first been Andante (in a calligraphic writing similar to the one used for the instrumental accolade), was crossed out and substituted by the more "informally" written Adagio. Thereby, however, the rather crowded indication for Adagio in measure 43 becomes redundant; it can be understood only as an afterthought that preceded the change from Andante at the outset: for the forty-two measures before Adagio, the tempo had been Andante, and the change at measure 43 had been taken in terms of a slower peroration. One of several indications that the little work was not performed at the time of its composition is, surely, that this second Adagio was not crossed out. (An insinuation that this second Adagio might mean a return to the original tempo, after some uncalled-for and noisome heating and speeding-up--say, from letter B, measure 17, on--would seem unjustified by the music itself. Moreover, in the event of such change, the designations a tempo or Tempo 1° would have been more appropriate.) Yet, on the other hand, there is some evidence on these pages that a performance might have been planned: the penciled numbers between the instruments of the initial accolade may or may not be in Schoenberg's hand, but they certainly specify 2 + 2 for the divided violins, 3 (= 2 + 1) for the divided violas, one each for the divided cellos and a single bass. (It is conjecturable that these specifications stem from the performance in 1984 mentioned in the Österreichische MusikZeitschrift, but it is improbable for me to envision the original manuscript being marked up by Moldenhauer--rather than a Xerox copy of it. Further, the penciled senza sord. At letter B, over the entering solo violin-­corroborating Schoenberg's distinctly recognizable, handwritten diminutive sine (or ohne?) sord beneath--is in a characteristically German hand, though slanted leftward, as Schoenberg's is not, and has a final d unlike his more habitually Latinate lettering. Finally, the most tell-tale pencil marking on this page is the large diminuendo hairpin intended for the bass's staff, but mistakenly placed between both silent harp staves--an instance of notational slovenliness to which Schoenberg would be far less prone than to the inadvertent omission of parallel diminuendo hairpins for the other string parts in the same place. The dynamic on the downbeat of the measure at B being piano for everybody, and having been forte in measure 15, Schoenberg obviously intended that diminuendo for the strings.) Likewise, the four rehearsal letters (in pencil) may have been written by Schoenberg, but their placement, either superimposed on a staff already occupied by notational symbols (A and B) or cutting into notation--C cuts into a slur; D virtually obliterates the high F at the beginning of measure 37--would belie Schoenberg's scrupulous, customary neatness.

On page 2, one measure after letter C, the arco indication is missing in the bass part; its unintended omission, coupled with a mistakenly notated A-flat eighth note at the end of measure 30, both support my assumption that the piece was never performed in Schoenberg's presence. He would undoubtedly have made these necessary corrections in the score used at the time. In measure 34, the last sixteenth note is confirmed by the almost vertical as (meaning the pitch A-flat) written alongside. (The question, "Why there, if not elsewhere?" could be answered in terms of the very neatness already cited: because that final A-flat barely misses hitting the lowest staff line of the harp part notated directly above.) The erasure--with razor blade or knife, as was necessary for pen-and-ink copying--in measure 28 left a blank which should have been filled with eighth rests, and the subsequent E-flats in the cellos and bass were mistakenly made into dotted quarters because the erased notes were evidently of that value.

Everywhere is Schoenberg's notational care, nevertheless, markedly in evidence. The conscientious fingerings given the cellos in measures 19-20 are practically "autobiographical"; the unis for the last eighth note on page 2 is made unnecessary by the double-stemmed notation, but he wants to leave no possible doubt about it; the before the last D-flat in measure 36 is made superfluous by the eighth rest beneath this upward-stemmed note, but, again, he wants to be sure it is unmistakable, just as the ensuing divisi at letter D is included by way of insurance.

The harp part, which begins at letter B by octave-doubling the bass as it enters at the same place, comes into its own at the forte restatement of the melody (measure 29), arpeggiating harmonic changes in groups of thirty-second sextolets, septolets, or octolets, with occasional melodic duplication (mainly for the right hand) in measures 33-36. The boundary notes of these arpeggiations are of interest; they embody Schoenberg's concern for making the harp's top line rise very gradually, as well as exemplifying his diminution technique in having that top line imitate the melody (at one eighth-note's remove) in the measures preceding the aforementioned duplication, from measure 33 on. Seeing the accented F on the first quarter's afterbeat of measure 31 makes one wonder at his not having placed a (possibly accented) C on the last eighth, rather than the repeated F. He could have done so easily enough by making the immediately preceding septolet end on A-flat rather than C, and he consequently missed an opportunity for a cross-referential F-C, in anticipation of the (accented) B-flat-F in measure 31.

On the third page of this manuscript, several corrections and emendations confirm this very young composer's linear bent and certify his proven deftness in writing tonal music. The evident erasures at the end of measure 39 and at the downbeat of measure 40 show that the last eighth notes in Vl.I and II had been D and F (respectively), and that the ensuing downbeat would then have read C and A-flat in those parts. By changing these segments to B-flat-A-flat (Vl.I) and B-flat-C (Vl.II), Schoenberg "reserves" the higher C's slot for the harp, on the downbeat of measure 41, and thereupon allows the violin solo's two high Cs during the second quarter of that measure, as well as the syncopated one in measure 42, to make the two-octave downshift to the last occurrence of the melody in measure 43 both smoother and more meaningful. The same holds for the emendation in the harp part at the final eighth of measure 39. Originally, the upward arpeggio had ended on high D (duplicating Vl.I before it was changed), but Schoenberg realized that the occurrence of B-flat as the upper boundary of the arpeggios in measures 37-38 made a longer-range link between that B-flat and the subsequent high C of measure 41's downbeat possible. (Leaving out that C within the last eighth of measure 40 was crucial; had it been left in, the arpeggio would have ended on a senseless high E-flat, thus obliterating the connection to the repeated B-flats in measures 37-38, and spoiling the aforementioned pitch-class link to the melody's "familiar" middle C beginning.)

The correction that most eloquently proves Schoenberg's mastery is the elimination of the composition's third-last measure. Had that measure remained in place, the "rhetoric" of the ending would have become uncharacteristically redundant--even orotund. Considering the solo violin's rising line--without, for the moment, taking any of the other instrumental parts into account--it is clear that Schoenberg wanted each rising segment to begin on pitch-class B-flat: the first below middle C, the next one(s) an octave higher. The fact of the second rise starting on the second sixteenth of the measure (as did the first), would have conditioned the third (eliminated) rise to start at the same metrical point, but then the second rise would merely have been duplicated, inasmuch as a rise yet a further octave higher would have been absurd. Such a repetition would have been pointless, vacuous, and odious to the composer. The alternative was, then, to delay the rise's third entrance, by allowing for an eighth rest at the beginning of the measure; it could in that way have contributed to a natural sense of slowing toward the close. But we can discern from the crossed-out measure that this rise would, in this metrical garb, have entailed either of two intolerable features. As we see in measures 47-48, the arpeggiated rises eliminated none of their four constituent elements of this harmony, but in the crossed-out measure there is no A-flat. While that might contribute to the slowdown (by dividing the rise into two three-note segments before reaching its goal, rather than four-note ones reaching across barlines), the skip between F-flat and B-flat is jolting, especially in a final trajectory toward the high A-flat. On the other hand, the inclusion of the missing A-flat within this rise would have required a skip from D-flat, at the end of that measure, to the high A-flat, which would have been even more awkward. The alternatives would have improved nothing: neither beginning on D-flat (rather than the "accustomed" B-flat) nor beginning after a sixteenth rest would (for the reasons given) have been satisfactory. If we note, further, that the higher octave location of the harp chord in the crossed-out measure would have been senseless without some precedent, and that that, exactly, is provided by the palpable erasure in measure 47 (especially of the top staff!), with its subsequent emendation--obviously undertaken after the crossing-out of the offending measure--we understand the lack of pertinence in shifting the harp's register at all, for there is no reason for any such shift between the chord in measure 43 (bottom staff) and those in the two final measures. Lastly, a rhythmic-metric inconsistency was avoided: the isolated eighth-note articulation of the harmony on the last eighth of the eliminated measure would have had no metric connection to the preceding triple articulation of the same harmony in measures 47­48; nor would the syncopations in both violin parts have had any direct rhythmic bearing on the single-eighth-note chord, and the connection between that and the iterated tonic harmony of the last two measures would have been arbitrary, at best.

By canceling what would have been the third-last measure, then, Schoenberg made for a smoothly rising, consistent solo violin part, and he also made sure that the motivically syncopated harmonic articulations in measures 47-48 found a consistent consequence in the triply articulated, nonsyncopated tonic harmonies, metrically placed so as to give prominence to the violin's arrival on its high A-flat and designed to close this sweet piece, with all due gentleness, on a sonority very similar to the one with which it opened.

-- Claudio Spies


  1. Günther Weiss, "Fünfzig Unbekannte Takte," op. cit., 48/3-4 (March-April 1993). [Return to text]
  2. I am much obliged to Jon Newsom for confirming and identifying markings in pencil (rather than pen and ink), and for corroborating that these were made with one sort of pencil only; i.e., not with differently colored crayons. [Return to text]