Robert Schumann is probably the most romantic of all German composers – his biography alone could stand as a perfect Romantic German novel. So when he says that his Liederkreis on poems by Joseph von Eichendorff, op. 39, is the most Romantic music he ever composed, we already know that we are going to hear about moonlight, castles, ghostly apparitions and, of course, the Rhine. Today we’re listening to one of these songs, Auf einer Burg (In a castle), which opens the second part of the cycle.
Auf einer Burg is a strange poem, the meaning of which no one could really explain. It presents two apparently disconnected images: first, an old knight turned into stone in his watch, forgotten in a corner of a castle for hundreds of years where the outside world only exists in the form of the rustle of the forest. If this were a movie, the camera would face the lattice and slowly travel to the forest outside. Birds sing undisturbed, the place is empty, for everyone has gone to the valley, where a wedding is taking place by the Rhine. There is music and animation – and (not “but”) the beautiful bride, she weeps.
We could joke that, if you fed a computer program to generate a German Romantic poem, the result would be something very close to Auf einer burg. But Eichendorff was no robot, and there is an almost non-verbal truth behind the verses in this poem. In a superior level, we have the remnant of an old world, a knight, whose memory is all but forgotten. But he still exists. In the real world below, everything is transitory – songbird, a party, a beautiful bride. And she weeps because she knows it. All colorful, happy things are meant to disappear. Eternity is grey and made of stone. It is a world entirely apart and there is nothing there for the young bride whose childhood is about to disappear forever.
I can think of few other composers who would have chosen this text for a Lied*. It is a song about non-communication, about concepts we will never grasp because they are older, larger, deeper than what we are in our daily lives with our daily problems. Schumann chooses a very particular point of view for this song – the bride who understands for the first time that nothing she knows is permanent. Schumann’s music depicts the insight. It is not a concept – it is a glimpse of something not fully understood. The only thing the young bride can understand is that there are two separated levels. And that is the musical cell of the song – all phrases in it involve up and down. The first two phrases have the same profile – we have a descending interval, then roughly the notes in between the lower and the upper note of the interval and finally we go down halfway. The next phrases ping-pong between registers: 5-1-1-1-1-1-6-6-6-2-2-2-2-2-7-7-7-3-3-3-3-3 (in a d major scale, where 1 is d), until we have a downward half-scale with repeated notes and a harmonically unresolved conclusion. The second stanza has the same structure, only the final bar has a slight alteration that makes it end in a half-cadence, what makes the listener believe that the song is not over yet (right where the bride weeps). Also, the song has the atmosphere of a church hymn with organ-like pedal, chorale-style in old-fashioned harmony, almost modal. The motive in the second and in the 23rd bars (what I called “roughly the notes in between”) reappears now and then in the piano part, as an echo. I feel tempted to call that the very image of the insight that made the bride perceive this other dimension, but that’s pushing it too hard. Anyway, it’s there and it propels the piece forward in a faintly Bachian way.
So here we are – Schumann concocted this fascinating structure that makes the listeners feel like they are hearing an echo of a forgotten world, but they cannot come to any conclusion: the first part of the song seems to go nowhere and the second basically stops midair. And the big question is – how would you build an interpretation as a singer for a Lied as mysterious and philosophical as this one? The answer is : you don’t. All singers who tried to sing it as something very profound fail. Most of them in this group are not native speakers. They start with a knowing attitude, as if they were going to let you know something important. But we can’t forget – the song does not tell you anything, there is no important message here. It just SHOWS you (it doesn’t EXPLAIN anything) that there is something out there you won’t fully understand. That is why the singers who just surrender to Schumann’s creation without trying to superimpose anything are those who come closer to the mark. This involves a very strong discipline – some singers feel that Schumann’s phrases here are too square and try to soften them by connecting the dots (with portamento) and that’s a no-no. Most sing it as precisely as possible and end up sounding as if they were reading the telephone book with a very intimate voice. If I had to be honest, in terms of technique, style, expression and vocal efficiency, Matthias Goerne (provided you can put up with the slightly nasal tone) doesn’t do anything wrong here, you could use his performance as a tutorial. He doesn’t even do the funny breathing pause after “oben” – which sounds awkward to my ears – or emphasizes too much low notes on the ping-pong leaps, what sounds abrupt. And yet it is not my favorite recording.
I am not a great enthusiast of Dietrich Fischer-Dieskau’s DGG Lieder recordings, most of them made in the phase of his career where the voice had lost its baritonal warmth and his legato was largely gone. But his early EMI recordings are treasurable. Those days he joined the best of two worlds: textual intelligence with perfect cantabile, dark resonance with angelic clarity. It was indeed the stuff of legends. And, even with the bizarre pause after “oben”, the emphasis on low notes in the ping-pong intervals and almost too much crooning, DFD’s recording with Hertha Klust is the one that brings this song to its full potential. The idea of singing it from a trance-like perspective is the best interpretation strategy for a song that speaks of things we see but not fully understand. It is almost as if he were hypnotized and just sharing the images in his mind with the listener. And the voice is to die for, glowing in sensuous velvetiness. The absolute clarity of diction and the occasional highlighting of one word or another never interrupt the flow of melody. He sings it expressively, but without trying to express anything in particular – and that is why this is so special. The very sound of his voice in the last word of the song weinet (weeps) is hushed, almost airy, but it is so well supported on a steady vibrato that seems that could go on forever.
Hertha Kluster’s piano seems to be a part of Dietrich Fischer-Dieskau’s voice – one hardly feels that the piano is a percussion instrument here. She plays it as an organ, the attack almost imperceptible. As much as the singer, she does not try to add anything to the song – she operates in a sound palette of demi-tintes, entirely open in interpretation. But I would be lying if I didn’t acknowledge that the recorded sound itself is part of the appeal. This track does not sound as something made by a recording engineer in a professional studio, but rather as an open reel tape made at home in an empty apartment. In my imagination, it feels like a memento of a special day of music making when these two people felt that they had to record their own insight about things that not words and maybe also not even music can explain.
*Schoek, for instance, did set Auf einer Burg to music. It has many similarities with the Schumann setting, although the atmosphere is very different.
I HEAR VOICES 
Riveting song. What a voice!
Good to hear from you, Roberto! I’m glad you like it.
DFD probably recorded a bit too much for his own good, his ubiquity worked against him to a degree. Which is odd because while he was certainly a very busy singer he was actually comparatively harder to track down in person. I don’t actually spend much time listening to him but I certainly love this recording you’ve posted quite a bit and his Winterreise in person was quite an experience, even if it the singing became progressively less the more he recorded it. But you are quite right, the EMI recordings are the idea ones for his leider, there was a golden hue to the timbre that is captured wonderfully and while I never found him to be spontaneous as such he wasn’t “precious” at that point. Later on when I saw him in recital it tended to feel like I was getting a lecture in a somewhat hectoring manner.
I also think because his Leider pretty much defines him on record, it’s easy to overlook how legitimately wonderful he was in opera. There were more limitations in that area and again the returns were diminishing to a noticeable degree from the 1970s on. But I think his Barak and Mandryka were heartbreaking as characterizations and he’s in such healthy beautiful voice as recorded in those roles it’s easy for me to overlook that both were probably marginally too heavy for him. And I actually think his German Verdi is superb and the Kubliek Rigoletto is at least interesting and that whole recording worth knowing.
Hello, Peter!
Yes, he recorded too much and became a bit self-indulgent at a certain point, what is understandable – reviewers worshipped him and one tends to loose a bit context in circumstances like that.
The Munich Mandryka and Barak are certainly some of his best performances. The Wolfram in the Konwitschny recording is exceptional and I even like the Konwitschny Holländer (no idea of how he would have sounded live). I’m not crazy about the Amfortas in the Solti studio, but the Rheingold Wotan for Karajan is kind of fun. The Telramund for Kempe is a bit clear of voice, but the whole obsessive-compulsive thing makes sense with the Ortrud of Christa Ludwig and I find that they live up to the reputation there. His Mozart was good but I wouldn’t say unforgettable – the Count Almaviva was a good role for him, but I have some difficulty in seeing a Papageno in him.
As for the Italian opera recordings, yes, the Kubelik Rigoletto is interesting for a change, but the Falstaff is an acquired taste I’ve never acquired. I don’t like the Rodrigo in the Solti studio, but the live in German at the Deutsche Oper (as seen on video) is a different story. Although I am not a great fan of opera in translation, for some reason, Verdi in German is worth the detour. And some of the casts are very tempting too.
I love the Wolfram at Bayreuth (well both of them I guess) quite a bit, certainly I think it ranks with the best. And yes I think the Wotan for Karajan is fun, though you probably need to be inclined to like him in general to really enjoy it. I like the Telramund as well, it’s basically an impossible role and I’m sure close miking helped but he’s certainly characterful and light or not more attractive than a considerable amount of the recorded competition. He and Ludwig certainly have chemistry and are well matched. I don’t think he did either that or the Dutchman live but I could be wrong (I haven’t heard that Dutchman recording in a while). I don’t like the Solti Parsifal much in general I have to say, and I like Solti more than most. But given the names it’s a pretty underwhelming go at the opera and DFD very much gives a “DFD performance”. His Sachs is bad honestly and that recording is the pits.
The Rigoletto aside I don’t think his actual Italian Verdi is worth bothering with much (he and Mazaal almost seem like they are deliberately trying to sabotage Lorengar on that Traviata). But I think he’s a divine German Rodrigo on video and the Germont in German with Gueden is IMO a lovely account of that scene and very much a performance worth checking out. I love a bunch of German Verdi performances, there was a strong Verdi tradition in that country and some of those performances are very potent. Metternich and Modl going at Macbeth is certainly something to hear even if as usual the sounds she emits often require a sixth sense to parse (she’s actually pretty good though and healthy sounding). And Reining in Trovatore is heavenly.
DFD in Mozart I can’t abide saving, yet again, the occasional live German performances. I guess Italian just wasn’t a functional language for him most of the time. The German Don has its moments even if it’s misconceived as an interpretation. But the commercial recordings are a might camp for my taste and his technical limitations in operas like Cosi and Figaro are made way too apparent. Granted most of that generation had similar issues but over a mic his are a little too audible. Live he just played those operas for sheer buffoonery, easily the dumbest count I’ve ever seen. It was terrible because he did not have a light touch to play the role that way so watching try so hard at playing the role as an idiot could be painful when his voice also wouldn’t really back him up.
– I was trying to keep track of the DFD’s recordings as Wolfram in Tannhäuser. Apparently there are three: 1954 (Keilberth), 1955 (Cluytens) and 1961 (Sawallisch). I don’t think I have ever listened to the Keilberth. But, yes, both the 1955 and 1961 are reference performances.
– I notice that Solti is a name increasingly in disfavour. I guess that if we go with individual performances, he might rarely be the first choice, but considering the side of his repertoire (really – it ranges from Bach to Schoenberg, Bartok…) it is amazing that the level was relatively high. But the Parsifal, yes, there is something missing there.
– The Jochum Meistersinger is a very strange affair indeed. I won’t even try to understand it 🙂
– I don’t know the Traviata with Gueden. I was a bit wary of hearing her as Violetta, but I’m open to suggestions: I’ll give it a try.
– I believe DFD had pretty decent Italian, above average for German singers those days. He didn’t have the sense of Italian phrasing, I’d agree with that. I see what you say about his being the dumbest Count Almaviva and I think it is a valid approach and he did it as part of the concept, as we can even see in the Ponnelle film. The mannerisms too fit into a certain way of understanding the role. But, as I said, I find it all forgettable in comparison to other singers in these roles.
Yeah I have no problem with his pronunciation of the language, it’s just like you said the phrasing seems to be at odds with his singing. There’s a stiffness and lethargy to it that just gets tiring for me and seems to expose some flexibility issues. His Count was certainly all of a piece as an interpretation. I just missed a sense of danger and menace other singers have brought to the role and I just think he came across as too intelligent that there was a sort of condescension and “this is how I see the role” attitude that I couldn’t get on board with. It’s not to be sneezed at, it’s just hardly a favorite of mine.
As with DFD, I’m not sure I would want to hear Gueden do that particular role in Italian (although I think she fared better than he did in terms of smooth phrasing and ease) but as someone who generally likes her work quite a bit even taking her “precious-ness” into account, I think she’s quite fine there. It was a comparatively later role for her so the tone is a bit rounder and she’s a bit older than the average Violetta so she has a certain faded worldliness and repose I happen to like for that role and which is rare (more so now than back then, but still). I guess you’d have to be somewhat partial to her in general to make it really worth tracking down, but I think she’s quite good.
With maybe one exception, there is no specific rep I would necessarily classify Solti as being definitive or anything. And I can see where some of the criticism is valid. An opera like Parsifal is not one I would necessarily expect him to excel with. But I personally can’t get on board with the way the tide turned against him. Again, maybe like DFD, it had to do with his ubiquity on records in relation to some other singers who only got appreciated later on when pirates and older recordings came back on the market. But I think he had his strengths and generally was fairly smart about things like casting.
– I would say that most Counts have problems with flexibility, especially in the stretta of Vedrò mentr’io sospiro. Even in recordings I’ve rarely heard it really well sung. So DFD would be in the majority here 🙂 But I agree he does see the role in some sort of elementary comedy approach.
– Solti seems to have been a trouper and Decca explored the non-problematic nature of their relationship to use him in all kinds of repertoire based in the belief that listeners didn’t really care for the conductor, but for the CASTS. In a certain way, Solti’s approach was a non-approach – he just tried to be faithful to the score and his concern was basically to make it sound precise and dramatic at the same time, a bit like the poorman’s Toscanini. I have to say that the non-approach could sound refreshing in stuff like Das Rheingold, when you basically can hear everything and it just flows without any philosophical momentum (i.e, slow tempi, heavy orchestral sound and smudged phrasing). But when we speak of Parsifal, the the 3rd act of Tristan and Isolde… then it could be testing. And here it’s a wild guess – I think singers liked to work with him because he was “nice” (i.e., when he wanted) and let them be, rather than try to make the whole cast produce a coherent and unified vision. This could be problematic, especially when the composer had singers in recitative-like phrases and the orchestra does not have a forward-moving rhythmic frame. Then it can become really pointless. All that said, I still find that he is someone who should be evaluated in a “lifetime achievement” category.
Incidentally I just saw your post about the Steber Arabella. I’m so glad you enjoyed it. Yes, it’s maybe my favorite performance of hers and lord knows she left many treasure able souvenirs. But her singing there (LIVE!) is just unbelievable and her generosity of spirit and vivaciousness really work wonders for the role. At her best she was remarkable. And unusually for a met broadcast the whole thing is top notch. Even taking personal preferences into account, it’s easily one of the most evenly well cast and played across the board.
And I thank you again for recommending it to me! I’ve even ordered the Andromeda release to put on my shelf. It’s really the strongest cast in the discography – and she is vocally stupendous. It’s a pity she didn’t try Ariadne – or did she?
I don’t believe so. It didn’t make it to the met until after she was basically gone and though her European career resulted in some indelible live recordings, like many Americans from the Johnson met era her appearances there were relatively few. The only Strauss she did onstage was the Arabella (just at the met, and maybe just that run) and Sophie and the Marschallin. The latter unfortunately doesn’t lay congenially for her, although when it does (the trio and conclusion of act one) she sounds divine. It’s pristine singing but just not a role designed to showcase her best assets, temperamentally as well. The later Lieder recording is too late sadly but again the upper third of her voice still has its sheen and there’s enough soul to her singing that it’s maybe worth a cursory listen.
Kubelik Rigoletto would have been a dream recording if it wasn’t for DfD. Hermann Prey comes to mind as German baritone who sang Italian just fine. DfD was gross. His Scarpia in Maazel’s recording. Argh… He was not subtle. I could not get into the character that he was portraying. I sensed that he wa worrying too much about DfD to think about the character.
Roberto, I find DFD an exotic piece of casting (I guess everybody does) that makes it hard for this recording to be a “reference” recording, but still he makes interesting things there and his interaction with Renata Scotto is something of scientific curiosity. Hermann Prey on the other hand was almost idiomatic in Italian and, if his voice was irresistibly velvet, it was maybe too velvety for Verdi (and not as large as one would with in that repertoire). And he knew that – he said it often in interviews. Personally, I prefer Prey to DFD more often than the other way round, but curiously not in Auf einer Burg. And almost never in Verdi (when DFD was young and sang with the full color of his voice).
Actually, there are TWO Maazel Toscas – both recorded in Rome. The Italian Tosca has Nilsson (IMO, one of her best Italian opera recordings) and Corelli, whereas the German Tosca has Silja and James King. I would have to listen to it again (and I’d rather not 🙂 ) to be sure, but I have the impression that DFD is marginally better in the German recording. In any case, I don’t know if I’d say that he was “out of character”. I would rather say he got the character wrongly. He plays him as “evil and loving it mwahahahaha”, and that’s not really the story there. I mean, Scarpia is clearly a psycho, but the point of psychopaths is that they are not doing all those bad things because they are choosing between doing a good or a bad thing – they just do it.