Harald Sæverud
(b. Nordnes/Bergen, 17 April 1897 – d. Bergen, 27 March 1992)
String Quartet No. 3, Op. 55
(1978)
I Allegro moderato, cantabile (p. 1)
II Andante con moto (p. 19)
III Allegro deciso (p. 35)
Harald Sæverud was born as the son of a respected and modestly wealthy business man and a devout mother. When Sæverud was 12 years old, disaster hit the household: his father, with his business partners, was found guilty of tax evasion and went bankrupt. He was sent to jail for three months. It was at this time that young Harald began to write music, perhaps as an inner escape from grim reality. His first formal studies took place at the Bergen Conservatoire where his main teacher was the pianist and composer Borghild Holmsen (1865-1938). By the time he was 17 Sæverud was working on his first symphony, often skipping school in order to do so.
Between 1920 and 1922 Sæverud studied at the Berlin Hochschule. While there a wealthy friend hired the Berlin Philharmonic for the first performance of “Overtura Apassionata”.
Later in life Sæverud would claim that he learned nothing in Berlin, and that his only teachers were Haydn, Mozart and Beethoven. However, letters from the period show that he did in fact learn a lot in Berlin, where he studied with Friedrich Ernst Koch (1862 – 1927).
Upon returning to Norway in 1922 he slowly built a reputation as one of the country’s most promising young composers, making ends meet as a music critic and by giving piano lessons. He received unexpected encouragement from Carl Nielsen (1865–1931), who wrote Sæverud a letter expressing his great enthusiasm for his “Five Capricci” for piano, op. 1.
Harmonien (today known as Bergen Philharmonic Orchestra) would eventually become Sæverud’s main expressive outlet. The ensemble went on to premiere many of his orchestral works.
In 1934 Sæverud married Marie Hvoslef (1900–1982), an American of Norwegian ancestry, daughter of a very wealthy ship magnate who had died in 1926. Marrying Marie Hvoslef enabled Sæverud to devote himself to composition full-time. It also facilitated the building of Siljustøl, the impressive mansion in the outskirts of Bergen which would become their home. Siljustøl, with its wild nature, would, from 1939 until Sæverud’s death, serve as his main source of inspiration, together with his family.
The completion of Siljustøl was timely. The Nazi occupation of Norway during WW2 started in April 1940. Sæverud was one of the few Norwegian artists with an unequivocal anti-Nazi stance throughout the war. WW2 turned out to be an especially fertile period for Sæverud. The anger caused by the Nazi invasion brought out a strong rush of creativity in him. This difficult time induced him to compose three symphonies (Nrs. 5, 6 and 7), as well as several other orchestral works and piano pieces. The Ballad of Revolt (“Kjempevise-slåtten”), by far his most famous work, was written out of rage against the occupation.
By the end of WW2 Sæverud had become an established composer. The Ballad of Revolt made him into a kind of national hero in a country that had just begun to recover from five years of foreign invasion.
Shortly after the war, the radical actor and theatre director Hans Jacob Nilsen (1897–1957) asked Sæverud to write new music for a stage production of Ibsen’s Peer Gynt. Sæverud hesitated at first but, on rereading the play, he decided he had indeed something to contribute. His Peer Gynt score stands among his most powerful creations.
In the 1950s the Koussevitzky foundation commissioned his Violin Concerto, op 37. The State of Minnesota commissioned a symphony (his eighth, the “Minnesota Symphony”, op. 40) to celebrate the 100th anniversary of the state in 1958.
Although the new wave of modernism that swept Norway in the 1960s relegated Sæverud to a less prominent position (at least in the eyes of the young generation), he continued to be held in high regard by the musical establishment and the public. He was made an honorary member of Harmonien, and was commissioned to write a work for the celebration of the 900th anniversary of the city of Bergen (“Fanfare og Hymne”, op. 48). He also wrote the opening work for the official opening of Bergen’s Grieghallen, in 1978 (“Overtura Monumentale”, op. 53). In 1977 he received the highest public honour available in Norway: the medal of Commander of the Royal Norwegian Order of St Olav. In 1986 he was made the Festival Composer at the 34th Bergen International Festival, for which he composed his last orchestral work: a suite for orchestra and choir based on Ibsen’s “Keiser og Galilæer” (Emperor and Galilean).
Although the final years of his life were marked by illness and predictable physical weakness, he remained alert and engaged with the world around him to the very end. Between 1986 and 1989 he wrote a number of small piano pieces and a Sonatina for viola and piano.
He was given a state funeral.
Sæverud is, arguably, Norway’s pre-eminent symphonic composer. With his nine symphonies, five concertos and numerous orchestral works, including ballet and stage music, he created a distinctive body of work, wide in scope and imagination. His music for piano stands as a truly original and unique contribution to the literature.
He began writing in a style redolent of late romanticism. For roughly a decade he experimented with atonality before finding his mature voice around the beginning of the Second World War. His music is characterized by an intense focus on the present moment and is strongly influenced by Western Norwegian nature, with its abrupt contrasts both in landscape and weather. He was always mostly interested in musical lines and the way they interact with one another, and was a master in the exacting art of two-part writing. He always demanded full characterization of every note and every phrase, often stating that “every note is a personality”.
Harald Sæverud’s third string quartet
Best known for his orchestral and piano music, Harald Sæverud did not approach the chamber music medium until late in life. He had written the 20 small violin duets op. 32 in the 1950s, and these proved to be a treasure trove for his subsequent work, as they provided ground material for several large compositions. However, Sæverud wrote his first proper chamber work (the first string quartet “Serenades of the two rivals”, op. 49) at the age of 73. The second and third string quartets followed in, respectively, 1975 and 1978. In 1982 and 1983 he composed two woodwind quintets and his last completed composition, the Sonatina for viola and piano, came in 1989.
The third string quartet was commissioned by the Bergen Chamber Music Society and completed on Christmas Eve 1978. It received its first performance at one of the Society’s concerts on May 9th 1979. The performers were the members of the Harmonien String Quartet at the time: Rolf Sandvik and Terje Norbeck Johannessen (violins), Jean Tørud (viola) and Kerstin Widengård (cello). For the premiere Sæverud gave each of the movements titles. The first was “søkende” (searching). The second “pastoralt hvilende” (pastoral resting). The third “det siste ord” (the last word). Although these titles are helpful in understanding the intentions behind the music, Sæverud decided to remove them later.
The first movement has, indeed, a very inquisitive character, with a very prominent viola part. The opening presents small motifs that become the seeds of which the entire movement grows. After being presented several times in embryonic form, the first proper theme appears at rehearsal nr. (4). This theme had earlier appeared in the orchestral work “Sonata Giubilata”, op. 47. An important 21-note motif is introduced by the first violin at nr. (5). It undergoes a great deal of variation and development until the previous theme reappears 5 bars after nr. (9). A new theme (which had been hinted at in the 18th bar of the movement) comes gradually into view, first appearing in full bloom five bars after nr. (12). This is the melody of the 11th violin duet (obviously a favourite of Sæverud’s, since he used it in several of his larger compositions). The ending is characteristically outlandish, culminating in a bizarre repetition of the notes B flat–A by the viola. This movement seems to epitomize Sæverud’s often stated intention of being a kind of caretaker for the musical material; allowing it to bear fruit in the most natural way instead of forcing it to behave according to his wishes. The form of the movement is evolutional, and one can easily picture oneself a large tree where all the branches, leaves, flowers and fruits are the result of the original seed, each of them containing that seed’s realized potential in its unique, unrepeatable way.
The second movement opens with the same interval that closed the first one: a minor second; downwards in the viola, upwards in the first violin. This striking opening acts as a swift stage shift. The restless, searching character that had permeated the first movement is put on hold and we are invited into a nocturnal soundscape, full of mystery and intoxicating beauty. We first hear an angular five-note motif followed by an ascending fifth that foreshadows a later theme. The main theme in the movement is an unusually (for Sæverud) extended melody (10 bars) in sicilienne rhythm presented by the first violin, with minimal involvement from the other instruments. The small arabesque at nr. (2) offers a new point of view on the music, as if momentarily stopping time on its tracks, before inviting the cello to deliver the main theme (now slightly transformed) in its entirety (nr. (3)). A longer episode follows this second exposition. Natural processes of transformation and growth are already taking place when, after a sudden pause, the second theme appears in unison in first violin and viola at nr. (5). This melody had been the viola serenade in Sæverud’s String Quartet no. 1, “Serenades of the two Rivals”. Unlike the first theme, which was left to express itself freely, this new theme (with its opening ascending fifth) undergoes development as soon as it is stated. The tempo gradually increases and the texture becomes denser, with the staccato articulation gaining the upper hand. It appears that the summer night idyll and the pastoral rest have been compromised by the capricious Western Norwegian weather. When the music eventually calms down, the main theme reappears (nr. (8)), very quietly and now in canon (a major seventh and one octave apart) between the viola and the first violin. The second violin holds the two strands together. The “bad weather music” returns at nr. (9), finally receding at nr. (10) with the return of the lovely first violin arabesque. From here to the end of the movement, the music winds down, gathering the different threads from the movement into a peaceful but unresolved ending. This movement certainly ranks among Sæverud’s loveliest and most imaginative statements.
The abrupt opening of the third movement unceremoniously breaks the spell created by the preceding music. This is typical of Sæverud, who, in daily life, was never fond of lingering on one mood and would consciously change it when he felt it had lasted long enough. The first half of the movement had originally been intended as a piano piece for the fifth book of Tunes and Dances from Siljustøl (Slåtter og Stev fra Siljustøl), op. 26, which never materialized. (Sæverud does have an op. 26, but it contains no music; for a man who was fond of practical jokes, this, whether intended or not, is one of his very best ones). He had even gone as far as orchestrating the piece, but neither version was ever published. As with several other earlier pieces, Sæverud obviously rediscovered it later in life and decided it had enough substance to use it. This final movement is a passacaglia based on the five-measure ground presented by the viola from 5 bars before nr. (1). This provides the basis for a long series of continuous variations that exhibit a seemingly inexhaustible imagination. The movement is divided into two sections. When the music reaches a breaking point, it suddenly stops and the cello restates the passacaglia bass, quietly and in half tempo (nr. (13)). A new set of variations ensues, now incorporating the inverted version of the theme, as well as presenting it at double and even quadruple tempo. An emotional climax is reached at nr. (19) after which there is a new build-up. At (22) the music takes a few moments to catch its breath before plunging headlong into a typically boisterous, almost cataclysmic ending. The conclusion, with its unexpected change of tempo, and the long final chord following a pregnant pause, takes a sideways glance at the magic of the second movement. It may be “the last word”, but one is inclined to think that Sæverud grants it to nature, rather than to himself.
It is remarkable that, although the compositional methods used are thoroughly classical, Sæverud makes the music sound, for lack of a better word, “modern” and definitely “contemporary”. Contemporary in the sense that it is relevant to our time, and in that it uses a language that is very much of our time, even if the dialect is distinctly Western Norwegian.
Sources:
WRITTEN:
- Original manuscript score from the National Library (earlier Norwegian Music Information
Center) – (NMI - SC)
- Original individual parts written by a copyist from NMI, with fingerings, bowings and
interpretational indications in the handwriting of the players who premiered the work
(NMI – vln. 1/vln. 2/va./vc.)
- Letter from NMI to Harald Sæverud, dated August 4th, 1986 (NMI – 8.86)
- Earlier original manuscript of 1st movement, Bergen Public Library – (BPL – 1)
- Copy of earlier original manuscript of 2nd and 3rd movements, Bergen Public Library –
(BPL – 2, 3)
- Original Sæverud manuscript of viola and cello parts, Bergen Public Library –
(BPL, va./ vc.)
ORAL:
January 26th 2018 interview with Rolf Sandvik, who premiered the work in 1979 after rigorous preparation with the composer.
EXPERIENTIAL:
- My own intense work with Harald Sæverud between 1987 and 1990.
- Years of preparation, numerous performances and a CD recording of the work with the
Hansa Quartet in the 1990s.
- Rehearsals, performances and making of a tutorial video in the spring of 2011 with a group
of students from the Grieg Academy.
- Rehearsals and performances during the academic year 2017-2018 with a group of students
from the Grieg Academy.
- Discussions with the research group “(Un-)settling Sites and Styles”, based at the Grieg
Academy in Bergen, Norway.
Recordings:
- Hommage á Harald Sæverud. 1986. Den Norske Strykekvartett. Booklet text by Lorentz
Reitan. VNP 0086-9
- Harald Sæverud String Quartets. Hansa Quartet. 1996. PSC 1141
General Remarks About Sæverud’s Notation
When discussing Harald Sæverud’s music one usually gets on to the subject of his highly precise notation rather quickly. I share the belief of many who knew him well that the main reason why his music is not performed as often as it deserves to be is precisely the great amount of markings (articulations, tempo, character and dynamics) that greet the potential performer when opening one of Sæverud’s scores. It is easy to become discouraged when one is not able to find many notes that do not have some kind of marking. It would appear that Sæverud gives very little freedom to those who interpret his music. This remains a point of contention among Sæverud enthusiasts: would it not be better to simplify the visual aspect of the scores in order to make them more appetizing to prospective players and conductors? I clearly see the point of that rhetorical question. On the other hand, the scores of many great composers (particularly from the Twentieth Century) have a strong visual imprint. One can, for instance, open a score by Ligeti or Lutosławski and immediately know it is their music before even focusing on a single detail (incidentally, both of the mentioned composers often used notation that was far from “user friendly”, yet their music is widely performed). The same is the case with Sæverud. If one were to simplify his notation, the scores would not look like they were his anymore. Furthermore, they would offer little resistance. The concept of resistance was paramount for Sæverud. He wanted his musicians to struggle. The late Jan Henrik Kayser (1933-2016), a pianist who worked closely with Sæverud for several decades, is heard on a 1994 Norwegian Television documentary (https://tv.nrk.no/serie/festspillene-i-bergen-tv/fkum11000194/02-06-1994) saying that Sæverud, in a speech given on his 50th birthday, thanked specially those who had given him resistance, for they increased his strength. In the same program, Kayser also recalls how, on one occasion, Sæverud told him to disregard all the markings on the page and play the music the way he felt it.
Sæverud always said that Haydn, Mozart and Beethoven were his ideals. He considered himself not a neo-classical composer but a classical composer.
I early on came to see Sæverud’s detailed notation as a way to look at the classics through a magnifying glass. He wanted his music to flow like Mozart’s, but felt compelled to give precise directions about every parameter of the performance. This was probably the result of hearing poor performances of his music where inexperienced players would commit common errors such as slowing down at the end of a phrase or swallowing a weaker note in a musical line. In order to prevent such mistakes he would write “quasi accelerando” or simply “accel.” (to prevent slowing down – see 2 bars before (13) in the first movement and 1 bar before (11) in the second movement of this quartet) or use a dash or an accent on the weak note(s) of the phrase (see 1st and 2nd violins in the second movement, at rehearsal number (5)). The problem with such directions is that, if one simply follows them, without understanding the reason behind them, the music becomes stiff and mannered, as well as devoid of the natural flow that was so important for Sæverud.
Of all of the markings that Sæverud used one in particular was the cause of much misunderstanding. It is what he called “marcato espressivo”: <>
Rolf Sandvik recalls that Sæverud told him and his quartet colleagues in rehearsals that Mendelssohn, Brahms and Schumann used it. And so they did, as well as Bruch, Saint-Saëns, Dvořák, Schoenberg or Martin, to name a few prominent composers. The fact that several of these composers used the <> articulation in their piano music is a clear indication that, in spite of the way it looks, <> is not meant to be a swell.
In the inner cover of the manuscript of his Fifth Symphony (kept at the Bergen Public Library) Sæverud wrote: “<>: is played sf vibrato, it is not stranger than that”.
It is interesting that Sæverud mentions sf and <> in the same sentence. In my experience of working with him it became clear that both these markings concern a level of expressive intensity that applies to the entire length of the note they accompany. Rolf Sandvik remembers a conversation with one of the former chief conductors of the Bergen Philharmonic Orchestra, Carl Garaguly (1900-1982), whom Sæverud respected highly, in which Garaguly explained that sf is not an accent (like >) but a “heightening of intensity”. As such, it must apply to the entire note. In cases where it appears on the first note of a legato slur, it is understood that this “heightening of intensity” is to be kept through the entire slur. Where it concerns <>, Sæverud makes a differentiation. We can see this clearly in the first movement of the Third String Quartet. Beginning 8 bars after rehearsal number (7) we see, in 1st violin, viola and cello, the use of a wider version of <> over two slurred notes (a falling figure). In this case it does certainly look like a swell. However, 7 bars before (8) <> appears only on the first note of the slur on the second beat of the 1st violin, answered in the same way by the cello in the next bar. However, these two last slurs contain an ascending figure. After months of deliberation and discussions with my colleagues, I am now convinced that both these markings are versions of the same expressive articulation: where it stands over two notes it holds for both; where it stands on one it holds for that note alone.
Historically, we find an example of the use of <> over two identical chords in the second movement of Beethoven’s piano sonata op. 7:
Sæverud himself used the marking in this way (but in much closer detail) in his “Romanza”, op. 23 for violin and orchestra:
In the Third String Quartet there are two more instances of the use of <> over, in this case, a syncopated slur across the bar line. They occur in the 3rd movement, in the five measures before (5), in the 1st violin. Here also, <> applies to the entire note (understood as a quarter note - rather than two eighth-notes - played across the bar line). The same applies to the viola part in bars 2-5 after (21).
An interesting discrepancy occurs in the second movement. In the first violin, 3 after (1), Sæverud gives the first note of the bar (tied to the second beat) a sf as well as an extended <> (as discussed above). When the exact same melody is restated by the ‘cello at 3 after (3), there is only a regular <> above the beginning of the note. It is quite possible that Sæverud meant for both these to be executed in the same way, but the absence of the sf in the cello raises a question that is not answered in any of the sources.
It is worth mentioning that Sæverud even invents one articulation to suit his purposes. At rehearsal number (2) in the second movement (and later at (10)) the peculiar marking used in the 1st violin is to be played “sonoramente” and with vibrato. This articulation is a pertinent example that shows what a great pity it would be to remove Sæverud’s markings in order to simplify his scores.
Notes on the Musical Text
1st. movement
- 4-5 after (2): the pulse relationship quarter-note = dotted quarter-note appears here. It is understood that this relationship holds true for the entire movement. In the case of the 5/8 measures, the first three eighth-notes of the 5/8 correspond to 3 eighth-notes in 6/8 or an eighth-note triplet in common time.
- (3): The eighth-note rest originally missing on the third beat of the bar in the viola part was added by Sæverud in blue pen on BPL, va.
- 8 after (2): in the viola part, the dotted lines have been extended to the end of the bar, indicating that “tranq.” only applies to that measure.
- (4): it is understood that all the repeated E’s in the second violin part are to be played as harmonics for six bars. Throughout the quartet, Sæverud only repeats the harmonic sign (0) on tied notes if they happen to be artificial harmonics.
- 2 before (5): <> (in parenthesis) added to first note in the first violin.
- 1 before (5)-3 after (5): trills extended for the entire value on the notes in second violin part.
- 3 after (6): accent (>) in parenthesis added to fourth note in ‘cello part, to correspond to the first violin part, with which it forms a canon, one beat apart. The reason for the parenthesis is that, although it seems logical to give the accent to that note, it happens at the same time as a sf in the first violin part. It may well be the case that Sæverud, who had generally a very sharp eye for detail, considered it superfluous to have an accent and a sf simultaneously in two instruments, as it would not be possible to hear the difference.
One may then legitimately ask why he does precisely that (sf in the cello and accent in the first violin) on the first note of 4 after (6). In this case, however, the 1st violin note is part of a two-note motif that kicks off the important 21-note ostinato row that begins on the upbeat to (5). The first two notes of that row are consistently accented, regardless of the instrument on which it appears.
- 6 before (7): 32nd-note quintuplet to be played in separate bows, as notated.
- (8) – Tempo I is to be understood as quarter-note = 80, since poco più mosso (3 before (9)) has the metronome marking of dotted quarter-note = 96 (the same as (2)).
- 5-6 after (8): in viola part, BPL – 1 contains no dynamic indications. NMI – SC shows a hairpin crescendo in both bars, followed by a question mark. This edition gives those two hairpins in parentheses.
- 6-5 before (9): the ‘cello part has been beamed in the same way as the two following bars (not so in NMI – SC).
- 3 after (9): BPL – 1 has no sf on the first violin’s first note. NMI – SC shows a sf, added by the composer. The latter has been kept in the present edition.
- 4 after (9): all sources have no tenuto dash (-) on the second violin’s third last note, unlike the corresponding note in the three preceding bars. This seems to be justified by the fact that in this particular bar the second violin changes pitch on the two final eight-notes, while in the three preceding bars it stays on the same pitch. The tenuto dash used in those three bars is consistent with Sæverud’s wish for clarity: he makes sure that the eighth-note played after the two ricochet sixteenth-notes is heard clearly.
- 5 after (9): none of the sources gives a dynamic indication for the first violin. Although it has mf at (9), I consider it prudent to add the same dynamic in parenthesis four bars later, since the viola enters in f in that bar after having been silent for 11 bars.
- 1 before (10): f added to second violin part (necessary, but missing in all sources).
- (10): only the first violin has accents (>) on the eighth-note pitches E-C#. In this edition accents have been added to the same pitches in the second violin and ‘cello parts, as I see no musical reason why they should be absent.
- 7 before (11): p added to first violin part (to correspond with second violin)
- 7-6 before (12): in NMI – SC, Sæverud puts a question mark on the first three notes played by the first violin. The doubt concerns the notation of the eighth-note triplet across the bar line. BPL – 1 has triplets, without any comment or question mark, while NMI – vln. 1 has two sixteenth-notes and an eighth-note, with the “3” crossed out. This edition presents the triplet notated across the bar lines, as it seems obvious that this is what the music demands.
- (12): It seems obvious that the ‘cello carries the main voice for four measures. The first violin alternates between f (first, third and fourth bars) and p (second bar), while the second violin begins in f and stays p for the second to fourth bars. This is precisely the way it appears in all sources but is mentioned here, just in case.
- 2 before (13): the “accel.” indicates forward motion for the three beats it covers, but does not have an effect on the tempo at (13).
- (13): For four bars the first violin and ‘cello play the same intervals in contrary motion in straight eighth-notes. In the first bar both instruments have light staccato dots (.). These disappear mysteriously from the ‘cello part from the second bar onwards. The same is the case with the sf on the first note of the second bar and the accent (>) on the first note of the third bar. These apparent inconsistencies are common to all sources. While it may appear illogical to have two different articulations for two instruments simultaneously playing virtually identical material in the same rhythm, the great distance between the registers (over three octaves at its largest) actually makes it possible to hear the difference of articulation. The one factor that makes it impossible to be 100% sure that this was indeed Sæverud’s intention (as opposed to a common human mistake) is the fact that the articulation is identical for both instruments in the first bar. Rolf Sandvik is of the opinion that Sæverud, who was meticulous and fastidious to a fault, would not simply forget such an important detail. While we established, with our research group, that the effect of playing two different articulations is both plausible and interesting, in this edition I have decided to put the staccato dots in the ‘cello part in brackets, while the sf in the second bar and the accent (>) in the third bar appear in parentheses. Interestingly, Sæverud does give a sf to the fourth note of the ‘cello in the third bar (corresponding exactly with the first violin part).
NMI – vln. 1 shows that this issue must have been discussed in rehearsals. We see, in Rolf Sandvik’s hand: “staccato?” and, underneath “spicc”. This seems to suggest that the violin part is meant to be played off the string. Unfortunately, no indication as to the desired type of bow stroke appears in NMI – vc.
- 2 before (14): “(nat.)” added to viola part (to cancel the previous “estremamente sul ponticello”)
- 4 before (14): ff is given in parenthesis in the second violin part.
In the same bar, Sæverud gives the ‘cello a tenuto dash (-) on its fourth note. The corresponding note in the same figure in the next bar has a staccato () and accent (>).
In spite of Sæverud’s penchant for variation in every aspect of his music, there it is truly no musical justification for the aforementioned tenuto dash. This edition thus gives the tenuto dash in parenthesis and the staccato and accent (without parenthesis), as the preferred articulation for that particular note.
- The odd “anti-solo” that the viola plays at the very end of the movement (from the second bar of (9) to seven bars from the end) can either be played in first position or in third position. Playing it in first position requires a very large hand and would be uncomfortable, regardless. In all of the performances in which I have been involved, the violist has played it in third position, that is: playing the A as an open string.
2nd. movement
NMI – vln. 2 has, in the hand of Terje N. Johannessen’s, the words “sløret summarnatt” (mild summer night) written at the start of this movement. Johannessen’s part is, incidentally, the only one that gives Sæverud’s original titles for each of the movements.
- BPL – 2 has “con sordino” written on all staves.
In NMI – SC “con sordino” is written above the top staff and “Sord.” appears in Sæverud’s handwriting (obviously from a later date) to the left of the first score system, indicating that all instruments should play with mute. In this edition “con sordino” has been added to all four instruments, as in BPL - 2.
- 1 before (1): the “rit. molto” only applies to that bar. The main tempo is reinstated from nr. (1)
- 6 after (1) – and later 3 before (2): NMI – SC shows that Sæverud was unsure about the rhythm of the first violin (marking it with a cross). He may have considered changing it to the sicilienne rhythm that prevails in the rest of the melody. The same doubt is shown 5 before (4) in the cello solo and in all the parts in the section between (8) and (9). As this doubt is not shown in any of the individual parts (NMI – vln. 1 etc.) and Sæverud shows no definite decision on the matter, the original rhythm has been kept in this edition.
- (2) – in (NMI - SC) one can see that, in the first violin part, Sæverud crossed out the PP and wrote mf with pencil. This change is not reflected in the separate part (NMI – vln. 1) and so pp has been kept for this edition. Although the viola is indicated as the soloist here, it goes without saying that the first violin part is very prominent, playing one of the most memorable figures in the entire work. Sæverud writes “vibrato” in blue ink (next to the special articulation in the first violin) in NMI – SC.
- (2) and the following bar: In the second violin part, in NMI – SC (but not in BPL – 2) Sæverud specifically asks for the harmonic that produces the pitch C’’ to be played as a major third (fourth overtone) harmonic. The same pitch can much more easily be produced by the traditional fourth (third overtone) harmonic. Taking into consideration Sæverud’s very sharp ear for details, I have chosen to keep his original notation, which gives a more ethereal effect than the traditional fourth-harmonic, though it is undoubtedly harder to control.
- 5 after (2): p added (in parenthesis) to the second violin part.
- 3 before (3): p added (in parenthesis) to the first violin part.
- 1 before (3): the first two notes in the cello are meant to form a 16th-note triplet, although this is not explicit in any source. The triplet sign (“3”) has been added in this edition.
In the same measure NMI – SC shows f written in pencil for the viola. This has been kept in this edition.
- (4): In the viola part, the last four notes are obviously meant to be played in treble clef (missing in both NMI – SC and BPL - 1). In this edition they have been kept in the alto clef, thus avoiding yet another clef change in the following bar.
- 2 after (4): f added, in parenthesis, to all parts.
- 4 after (4): the obviously missing bass clef for the cello part has been added.
- 6 after (7): the obviously missing bass clef for the cello part has been added.
- 4 before (8): ff added to second violin part in parenthesis.
- (9): Sæverud, who is known for his love of the unusual and the outlandish, had originally intended the second bar of (9) to be compressed into the fourth beat of the first bar. It appears so in BPL – 2. Sæverud clearly realized that the resulting micro-rhythms were unplayable and thus thought better of it and made that originally intended fourth beat into a 3/4 measure. It appears in this way in this edition.
- 2 after (9): NMI – vln. 2 has a comma (,) at the end of this bar. This appears to be a practical way of ensuring good ensemble in the following 5/8. However, since taking a breath there would stop the sense of great momentum accumulated up to that point, the comma has not been used in this edition.
- 3 after (9): although it is not specified in any source, it seems logical the pulse relationship quarter-note = dotted quarter-note applies here (as it did in the first movement). It has, thus, been added in parenthesis in this edition.
- 6 before (10): tenuto marks (-) added to the last two notes of the second violin, to correspond with the first violin.
- (10): no technical specification is given for the performance of the C harmonic in the second violin part. It is understood that Sæverud wished for it to be played in the same way as previously (2). No indication has been added in this edition.
- 1 before (11): all sources contain the same unfortunate mistake: a bass clef is given to the cello part (superfluous, since it has been in bass clef for the previous 33 bars), which is obviously meant to be a treble clef. This has been corrected in this edition.
The “accel.” at the end of the bar is only meant as a direction for the last two eighth-notes in the bar to move forward, without affecting the tempo of the following bar.
- 6 before (12): the obviously missing alto clef in the viola part is restored in this edition.
3rd. movement
It is patently obvious that the opening scale is physically impossible to play at the given metronome marking. This opening has a characteristic element of surprise and shock, dear to Sæverud. The players are, thus, encouraged to play the run as quickly as possible and with the sense of an unexpected burst of energy.
- Bars 3 and 4: BPL – 3 has a diminuendo hairpin and no accel. NMI – SC has an added accel. and the curious superposition of a crescendo hairpin underneath the diminuendo one. Both these changes have been kept in this edition. It makes sense that Sæverud may have wanted the energy of the pizzicato cluster to carry on through the following two bars of silence, hence the accel. and the crescendo hairpin. The odd visual effect of the two simultaneous hairpins is in keeping with Sæverud’s love of the unusual and therefore it has been kept for this edition. The performers are encouraged to make of it as they see fit, taking the above information into consideration. All four individual parts (NMI – vln. 1 etc.) have an unequivocal cresc. hairpin and no accel. We must, thus, conclude that this must have been Sæverud’s preferred mode of execution for those two bars. The “accel.” appears in parenthesis in this edition.
- 5-1 before (6): interestingly, NMI – SC shows, in very faint pencil, a descending chromatic line in the cello, starting from G and down to E flat, in quarter notes on the second beat of each bar. Sæverud obviously decided against using it.
- (6)-1 after (6): NMI – SC shows a sf in pencil for the second note of the first violin. As this is not repeated in the following bars, it appears in parentheses in this edition.
- 5 before (8): no dynamic indication is given for the first violin. While it is conceivable that Sæverud may have wished the harmonics to begin loudly and then fade away, this edition gives p in parenthesis. It is understood that the second violin carries the main voice, while the viola and ‘cello play a new variation of the passacaglia theme. It is thus left to the discretion of the performers just how loudly the first violin note needs to be played.
- (9): I believe the three repeated sixteenth-notes in the viola for five bars are meant to be played on the string. This intuitive remark is supported by Rolf Sandvik’s comment on the fact that “Sæverud wanted Sound” on everything, going on to mimic the way Sæverud always sang detached musical lines: “voh-voh-voh”. This was wholly in keeping with my own memories of working with Sæverud. A down-bow in parenthesis has been added in this edition on the first bar of (9), followed by “sim.” in the following bar.
- 4 after (11): missing flat added to the second note in the viola part.
- In the five bars before (12) the awkward rhythmic notation in the cello has been kept. Although the second beat of the bar is not visually represented, the notation gives a sense of displacement, as if the cello is lagging one sixteenth-note behind the other instruments, which is, undoubtedly, Sæverud’s intention here.
- 4 before (12): accents (>) added to violin parts, to correspond with previous and following bars.
- (12): ff added in NMI – SC in the cello part has been kept in this edition. For some reason, there is a pencilled question mark to the left of the cello staff. Perhaps Sæverud considered allowing the cello to surface gradually as the other instruments become softer, but he seems to have discarded the idea.
- (14): NMI – SC shows that Sæverud considered using the mute for the first violin here, but he crossed it out later.
- (15): the word “dynamikk” appears in very faint pencil on NMI – SC. Sæverud was, perhaps, unsure of the dynamics for the violins. The given mp has been kept here.
- (16): the same is the case here. The given mf has been kept.
- 4 before (17): in the second violin “arco” is crossed out in NMI – SC. The second violin has been playing arco up to this point, thus making the indication unnecessary. However “(arco)” has been added to the second violin in the previous bar, since the first violin and viola are playing pizz. at that point.
- 2 after (17): NMI – SC and BPL – 3 have “senza sord.” written on top of the first violin staff. This is superfluous, since Sæverud decided against using the mute at nr. (14) (see above). The indication has, therefore, been removed in this edition.
- 5 before (19): on the second beat of the bar “(arco)” is written above the first violin staff and “pizz.” above the second violin staff. This has been kept for this edition.
- 4 after (21): I believe (for the same reason explained in relation to the articulation at (9)) the repeated sixteenth-notes in the upper strings are to be played on the string (staccato or with a strongly articulated portato) for five consecutive bars. Interestingly, though, NMI – vln. 1 has an up-bow for the first three sixteenth-notes. This suggests the possibility of an off-the-string or, in any case, lighter stroke for that section.
- (22)-2 before (23): the “con grazia” indication seems to suggest that the two repeated sixteenth-notes in the second violin and viola parts are meant to be played with a graceful ricochet stroke. The exception is the first violin part, 5 before (23): here the sixteenth-notes need to be played spiccato, since they are followed by a longer legato stroke. NMI – vln. 2 gives the interesting bowing suggestion down-up-up for the two sixteenth-notes/one eighth-note figure.
- 5 before (23): “p con grazia” added in parenthesis to first violin part, corresponding to the indication previously given to the second violin and viola.
- 1 before (23): mf and crescendo hairpin added to second violin part in parenthesis, since no indication is given in either NMI – SC or BPL – 3. The “arco” indication in the following bar suggests that the second violin is expected to play “pizz.” one bar before (23). The “pizz.” indication has been added in the present edition.
- 5-3 before (25): in NMI – SC, sf is added in blue ink to the third note in the violins. Interestingly, this is not reflected in NMI – vln. 2.
- The final chord of the piece is a chapter in itself. Sæverud had to struggle sometimes in order to find the best ending for his compositions. The Third String Quartet is one such work. NMI - SC shows an ending Sæverud discarded, followed by the one he considered to be appropriate. Both versions end with the same chord: B minor in first inversion. However, the final ending was pasted onto the manuscript in a separate piece of paper.
Here the viola note (the root: B) lacks the alto clef in front of it. Since the viola has been playing in treble clef for a number of measures prior, it looks like its last note is an A in treble clef (one ledger line above the staff). The copyist who copied the parts that can be purchased from the Norwegian National Library simply copied what he saw, without looking at the discarded ending. The result of this clerical mistake has been that the two available commercial recordings of the work end with a D major chord, with the viola’s fifth on top. I should, of course, have noticed this error when I worked on the piece with the Hansa Quartet in the nineties but, alas, it was not until 2011, when I revisited the work with a group of Grieg Academy students that this became apparent to me.
Very recently (May 2018) I was given access to diverse Sæverud manuscripts kept at the Bergen Public Library.
In a letter to Sæverud, dated August 4th 1986 (NMI – 8.86), a Norwegian Music Information official writes:
“Please find attached the score and parts of your third string quartet. As you see, we have written out new parts for the first and second violins. This is because it took time for us to get hold of your parts, since Norske Strykekvartett was busy preparing the work for the Festival record. Our copyist did not manage to copy the rest at that time, so he is finishing them this fall. It can seem somewhat superfluous to write them out, now that your material has returned to us, but we think it is tidier to have a set of parts written out by the same copyist. We hope you agree”
The “Festival record” referred to is the one mentioned above (VNP 0086-9)
No record of Sæverud’s reaction to this letter is available. However, I was dismayed to discover in Sæverud’s handwritten copy of the viola part (BPL, va.) that he had appended an alto clef, in blue ink, before the final note in the viola. Whatever doubt may have remained about the nature of that final chord, is now forever dispelled.
In gratitude
Many people must be thanked in connection to this publication. They all have contributed significantly to my understanding of Sæverud, the man and his music, at different times in the past thirty-two years.
Einar Røttingen, my good friend and colleague, who introduced me to Harald Sæverud’s music and to the composer himself in 1986, thus setting my musical life on the course it has followed since.
Harald Sæverud himself, who taught me much more than can be set into a few simple words.
Sveinung and Tormod Sæverud and Ketil Hvoslef (the composer’s three sons), Alma Sørbø Sæverud and the late Inger Bergitte Sæverud (the respective spouses of Sveinung and Ketil), and Trond Sæverud and Line Hvoslef (two of the composer’s grandchildren), who knew the composer better than anyone.
Lorentz Reitan, for his superb biography of Sæverud (“Harald Sæverud: Mannen Musikken Mytene” – 1997. Aschehoug & co. Oslo)
My colleagues from the Hansa Quartet, Anne Helga Martinsen, Helga Steen and Walter Heim, for countless hours of very hard work in the 1990s and their commitment to bringing this very complex music to life. Our 1996 recording of the three string quartets remains a reference and the only one to date.
Martina Hevrová, Karolina Vik Hege and Lala Murshudli (students at the Grieg Academy in the academic year 2010-2011), who performed the third quartet with me twice in the spring of 2011 and made a video recording of it at Siljustøl.
Jorunn Eckhoff Færden (from the Bergen Public Library), who has given me access to Sæverud’s manuscripts.
Kari Margrethe Sabro, from the Norwegian National Library, for giving me access to the original score and NMI set of parts with annotations from the original players.
Vladimíra Ščigulinská, Oddhild Nyberg and Carmen Bóveda, students at the Grieg Academy in the academic year 2017-2018. They have worked tirelessly with me on this music for nine months.
My colleagues from the research group (Un-)settling Sites and Styles, based at the Grieg Academy: Signe Bakke, John Ehde, Arnulf Mattes, Liv Elise Nordskog, Einar Røttingen, Njål Sparbo, Hilde Haraldsen Sveen, Torleif Torgersen and Knut Vaage. The insights they have provided throughout the process of relearning and researching this work during the 2017-2018 academic year have been invaluable.
Sæverud’s Third String Quartet is a masterpiece that deserves to stand proudly side by side with other great string quartets from the Twentieth Century. It is thus with great joy and pride that we present here the first edition of this work.
Ricardo Odriozola, July 3rd 2018 – 11:47