by Ricardo Odriozola

In the Bergen Philharmonic recordings, the physical closeness of the recorded sound seems very appropriate to the Sæverud ethos. This physicality is only matched, if not surpassed, by the 1971 recording of the Peer Gynt suites Op. 28 with the Oslo Philharmonic, conducted by Miltiades Caridis (a conductor with a close affinity for Sæverud's music).


In the 1970 recording of Sinfonia Dolorosa, Op. 19, the Sæverud spirit shone through, despite a recorded sound that favoured an overall sound mass over instrumental detail and clarity. A prime downside is the occasional predominance of the first trumpet in the sound spectrum, at times to the detriment of other important lines. Another is the lack of clarity in some of Sæverud's articulations. Although close listening reveals that many of these are being observed, they tend to get lost in the larger texture. The long gradual accelerando from rehearsal no. 29-34 is plodding and heavy, failing to create the intended build-up of tension.


The 1978 recording of Salme Op. 27 is somewhat more satisfactory with regards to clarity. This work contains for the most part leaner chamber-like textures which allow instrumental detail and articulation to be more clearly perceived. Gro Sandvik, the former principal flutist of the Bergen Philharmonic once related to RO how, during the recording of the fifth variation, with its high, soaring flute solo marked ff espr. molto, sempre tenuto, the composer stood behind the conductor, shaking his fist at her throughout the solo, urging her to play for her life. Judging from the result, she definitely did.


This is one documented example of the composer's presence having a palpable effect on the final result. Unfortunately, that magical moment is spoiled by what comes right after: the first trumpet again barges rudely into the sound spectrum with disproportionately loud accents (actually marked sfpp). What was meant as a lively and light awakening after the transcendental flute solo is rendered banal by the failure to balance the trumpets, timpani, and violins into a cohesive sonic whole. It is not even rhythmically together. Was this a mistake made at the editing table or was it the least unsatisfactory of the available takes? Did the composer have any say in the matter? This sixth variation is supposed to pave the way for the euphoric, excited seventh variation. Instead, because of its shoddy realization in this recording, the final, seventh variation fails to deliver its ecstatic message. There is another odd moment in the recording. Towards the very end of the fugue, from the upbeat to the third measure after reh. No. 36, there is sudden drop in the pulse, heralded by the ff entrance of the horns. This is not marked in the score; there is an allargando poco a poco only in the two very last measures of the fugue, 14 measures after the mentioned spot. It is such a dramatic and memorable moment in the recording that it seems obvious the sudden drop in tempo was made consciously, most likely according to the composer's wishes. Having worked with Sæverud, it seems unthinkable that he would have left such an important detail up to chance.


The 1985 recordings with the Royal Philharmonic and Per Dreier were, no doubt, a momentous occasion for Sæverud. It was the first time a major orchestra outside of Norway was making a professional commercial recording of his music. He is reported to have said to the orchestra after the sessions were over, "I am a melancholy man, but today you have made me very happy."


Considering that he was present at those sessions, one is surprised to find the tempo in Galdreslåtten Op. 20 to be considerably slower than what is indicated in the score. The tempo in the printed score is 8th note = 168-176. In the RPO recording the piece begins at ca. 150, going down to ca. 132 from the passacaglia section and never quite recovering beyond ca. 144 in the final segment. Were these differences in tempo a wish of the composer? Or was the generally slower pulse a consequence of the space where the music was recorded? In either case, it results in a not very exciting performance of Galdreslåtten, one of Sæverud's most immediately engaging orchestral pieces. The ubiquitous dotted rhythm, so characteristic of Sæverud, is not well served by a "comfortable" pulse. RO participated in the recording of the same work with the Bergen Philharmonic in 1996. The conductor, Dmitry Kitayenko, insisted on sticking to the given tempo (168). Although the pulse slackens ever so slightly towards the end of the passacaglia, it gets back on track for the rest of the piece. The result is riveting and more in keeping with Sæverud's temperament than the RPO recording, despite the fact that the editor managed to chop off two beats from the fifth bar before reh. no. 35!


The incredibly charming third movement of the Symphony No. 9, Op.45 suffers from the same lack of propulsive momentum and light-footed gracefulness in the RPO recording. By comparison, Alexander Dmitriev's 1997 recording with the Stavanger Symphony Orchestra is quicker and lighter in this waltz movement. However, neither Dreier nor Dmitriev take the Doppio movimento indication at reh. No.3 literally. Had they done so, the main section of the movement would play at quarter note = ca. 180, perhaps bringing the bird-like weightlessness of this delightful movement to life.


The reminiscences from Arne-Peter Rognan, producer during these sessions (see below), help shed light on the recording situation and its results. These RPO recordings with Per Dreier are of very high professional standard: both the playing and the recorded sound are superb. The comments above are not meant to detract from the importance of that album containing the first professional recordings of two of Sæverud’s symphonic works.


Arne-Peter Rognan reminisces about the Royal Philharmonic sessions in 1985:


You know, the London orchestras were veritable music factories at that time… It was normal to make recordings in venues with good acoustics around London. We had mobile recording gear of highest quality ...with three complete sets of gear that were carted around for sessions every day…
The orchestra gets the music in advance… It is pure sight-reading. BUT THEY EAT NOTES!
This costs money ... so they make themselves marketable by being able to produce a lot of music in a short time. Therefore, they have one musician who keeps an eye on how many minutes of music they complete per session ... so-called needle time ... never over 25 min per 3-hour-session without extra pay.
Such a recording is very demanding for all involved. Everything goes incredibly quickly.
… At that time, we went right onto two tracks ... that is, no mixing afterwards, so we seldom got more than ca 10-15 min for setting the microphones and finding the balance before the red light and the first take. Very stressful! …As producer, I had familiarized myself with the score and made myself a picture of what balance problems we might have to solve…
For this recording we had two sessions, as far as I remember. That is 6 hours. [...]
Harald Sæverud was very much present. He was a very colourful man, but his hearing was already then becoming worse, so we had to give him a headset ... and thus he was largely in his own world.
He was very concerned about the rhythmical aspect ... the feeling in the phrases ...  which is easy to understand when one studies his incredibly detailed notation.  Sometimes one needs to read the lines more than once to understand his semantics.  Therefore, it demands a lot from the musicians to extract the style from the orchestra parts like we had to.
But the musicians were very pleasant and were very inspired by having the composer present. And he certainly was (present)! He gave messages through the talk-back in his peculiar Bergen English, loud and demanding. I had my hands full trying to “translate” his comments into English orchestral language ...
Harald was in excellent humour and had a great time. [...] We managed it, although such a recording can appear to be chaotic to an outsider because one builds a splicing list in the score that shows which are the best takes for the different sections and phrases.
The cooperation between the conductor and the producer is incredibly important ...  actually, they are the only ones in the recording who know what the result will be. In many ways we can say that creating an organic expression from what we splice together is difficult. We are in danger of making a “Frankenstein” where every part is technically perfect, but which expresses illogical phrases ... this must be avoided.


(Email from Arne-Peter Rognan, 19. May 2020)


Trond Sæverud made two recordings of his grandfather’s Violin Concerto, the first with Sønderborg Symfoniorkester, conducted by Karsten Andersen in 1993, the second with Stavanger Symphony Orchestra and conductor Ole Kristian Ruud in 1997.


Despite the short period between these two recordings, they are vastly different. The Danish 1993 recording is part of a CD containing all of Sæverud's published violin music, including the unpublished early work Elegy, originally for violin and piano (here in its 1986 version for solo violin). In this recording the performers adhere largely to the tempos of Sæverud's revised edition (Musik-Huset, 1982). The presto sections in the first movement are not sufficiently quick however, detracting from the intended excitement. The same can be said of the last movement, with a relatively lacklustre tempo not fully conveying the music’s intended vitality. Although Trond Sæverud's playing is superb throughout, one gets the feeling that he feels somewhat constrained.


The 1997 recording with the Stavanger Symphony Orchestra is to RO more satisfying. After consultation with Ketil Hvoslef (composer, father of the violinist and son of Sæverud senior) Trond Sæverud decided to revert to the tempo relationships of the original version in the first movement, resulting in a much more cohesive flow in the music. The presto sections are much livelier and faster than in the 1993 recording. Another important consequence of Trond’s discussions with Ketil Hvoslef is the excision of the second to fifth measures after reh. no. 26 in the first movement. These are four typically eccentric, quirky, and outlandish Sæverud measures. Although wholly in keeping with Sæverud's focus on the intensity of the moment (perhaps to the detriment of organic flow), they were deemed unnecessary and inappropriate by both Trond S. and Hvoslef. Sæverud was known for making drastic changes to his own music. His son and grandson were his closest musical associates. If anyone could be entitled to make a post-mortem change in the music, it would certainly be them. And yet...


A far more perplexing change is the removal of the winds in the first two measures of reh. No. 13 and of the entire orchestra in the four subsequent measures. This is not the case in the 1993 recording, and it seems one step too far in interfering with the composer's work. Although Sæverud did express to RO his wish to thin out the orchestration of his early Cello Concerto Op. 7, the mentioned segment in the Violin Concerto works well with the original orchestration, with its flute trills and clarinet arabesques. There are other questionable minor adjustments to the orchestration.


The last movement in this recording provides the euphoric conclusion to the work that the 1993 recording failed to deliver. One also senses that the soloist feels much more at ease in the latter recording.  


RO recorded the Sæverud string quartets with Hansa Quartet in 1996. Listening to one's own recordings is very much like looking at old photographs of oneself: "Did I really wear those clothes? Was I ever so young?" After Sæverud's death it took RO five years to be able to fully move on and become independent from his influence, which had been enormous during the period they had known each other and worked together. The Hansa Quartet recording was made during a time when the "Sæverud sheen" had not yet worn off. RO never had the chance to discuss the string quartets with Sæverud. RO joined the Hansa Quartet in 1989 with the understanding that they would play the Sæverud String Quartets. All three works were performed many times, but no. 2 was the one with the least "road time" by the time of recording. Listening to the CD at the time of this writing (May 2020), after the period of intense research, rehearsals, and performances of (Un-)settling Sites and Styles, is a mixed experience. Have these "Frankenstein" renditions of the works (in Arne-Peter Rognan’s words) withstood the test of time?


To RO the Sæverud spirit seems to shine through, albeit more brightly at some places than others. The third quartet stands out. One has to admire the sheer abandon and audacity which the violist Helga Steen displays in that very demanding and often awkward part. The tempos are lively and engaging. The second quartet is somewhat less convincing. The tempos, particularly in the first movement, are not quick enough. This movement moves in many directions, making sense only in the last movement, like a good detective novel (which Sæverud loved). It demands close attention in order to stay on track with the musical train of thought. The slightly slack general tempo does not aid that purpose. The second movement lacks somewhat in poetry. The third movement, which pushes the players to the limits of possibility, has great momentum and some seriously incandescent dotted pizzicatos in the cello. The emotionally complex final movement is likewise rendered in a very satisfying way. The return of the lyrical theme from the first movement, accompanied by the first violin's "primal screams", is particularly effective. As of today, this 1996 recording by Hansa Quartet is the only reference recording of Sæverud’s string quartets. An earlier recording of the third quartet (VNP 0086-9), coinciding with Sæverud's 1986 residency as festival composer at the Bergen International Festival, contains a number of serious inaccuracies.


Lacking a performance practice of Sæverud's music on the concert stage, recordings are the best way we have to become familiar with his music. Do these provide us with a reliable representation of the music? Does the fact that the performers knew Sæverud well guarantee a listening experience that is wholly in keeping with the composer's view of the world? Performers who have enjoyed the privilege of working closely with a composer generally regard it as an obligation and a privilege to hand on that knowledge to new generations. But one might as well accept that, in “Chinese whispers” fashion, the communication of our supposed "authentic" understanding of the composer's intentions is going to be gradually transformed from one performer to the next.


Recordings are compromised artefacts, as Arne-Peter Rognan's reminiscences would indicate. They represent the best result that could be achieved by those players in that place and that time. Even in exceptional cases, they are only representations, not the real thing (just as a photograph of the Sistine Chapel ceiling is not the ceiling of the Sistine Chapel).


Insofar as there exists a performance tradition of Sæverud's music, as established on record by J.H. Kayser, K. Andersen, M. Caridis, Trond Sæverud, Hansa Quartet and E. Røttingen, it can be said to be dormant at this time. There are only a handful of performers worldwide with a commitment to performing Sæverud's music. In the event that major soloists and orchestral conductors should begin to take more interest in Sæverud's music, the aforementioned books, articles and CD texts provide essential information for new performances.


In October 2019 the Oslo String Quartet performed Sæverud's 3rd string quartet using RO's new edition. It was a superb and fresh rendition, surprising even its editor.

S C R O L L   D O W N

Foto: Bente Elisabeth Finserås