Mahler: Symphony No. 5. New Philharmonia Orchestra, John Barbirolli. EMI Studio CDM 7 69186 2 (1988) barcode 077776918624:
Recorded 16-18 July 1969 at Watford Town Hall, London
For the subsequent reissues and editions, see after the review.
Barbirolli’s Mahler 5th: a historical perspective on an interpretation that stands out
Originally posted on Amazon.com, December 29, 2011
His dates (1899-1970) would make Barbirolli part of the 2nd generation of Mahler conductors – not those, like Mengelberg, Walter, Fried or Klemperer who were direct disciples of the composer, or like Stokowski who saw him conduct, but those who championed his works from the early 1920s onwards, like Scherchen, Rosbaud, Mitropoulos, Horenstein, Steinberg, Ormandy or Kletzki. But in fact the British conductor came fairly late to Mahler. According to the very informative Barbirolli-Mahler biography posted on the website of EMI classics, “Barbirolli was a late convert to the music of Gustav Mahler. He had first come across it in 1930 when the Fourth Symphony, as heard for the first time at somebody else’s rehearsal, struck him as being thin, certainly by comparison with Berlioz and Wagner. After some early excursions at the beginning of his career – such as in 1931, when he conducted the Kindertotenlieder for Elena Gerhardt at a Royal Philharmonic Society concert in London – Mahler scarcely even figured in his programmes until 1946, when he included Das Lied von der Erde in his third season with the Halle Orchestra. Then in 1952 his friend, the critic Neville Cardus, recalling that Sir Hamilton Harty had given England its first hearing of the Ninth Symphony during his reign as Hallé conductor (1920-33), urged Barbirolli to consider conducting it himself. It was, said Cardus, ‘the ideal work” for him. Two years later the thing happened: moreover, that first-ever performance by Barbirolli of a Mahler symphony opened the floodgates to a 16-year period in which he embraced them all save No.8“.
His first Mahler recording was in 1957, the 1st Symphony with the Hallé Orchestra on Pye, which didn’t make much of a splash back then (first reissued in the early days of the CD on PRT 8385 barcode 5011664838521). So Barbirolli’s real outburst on the music scene as a major Mahler conductor really came with the 9th Symphony he recorded for EMI with the Berlin Philharmonic Orchestra in 1964 (link will open new tab to my review). Other than the orchestral song cycles with Janet Baker (Kindertotlenlieder and Lieder eines fahrenden Gesellen in 1967, Rückert Lieder in 1969, now conveniently collated on EMI Great Recordings Of The Century 5 66981 2 barcode 724356698123 – HMV logo – or 5 66996 2 bc 724356699625 Angel logo), he recorded two more symphonies for EMI, the 6th in 1967 (first CD reissue in 1994, CZS 7 67816 2 barcode 077776781624), and this 5th in July 1969. The Barbirolli admirer may regret that EMI’s famous producer, Walter Legge, was so infatuated with Klemperer, because one can suppose that this is what prevented EMI from producing more Mahler recordings from Barbirolli. The 2nd, 4th, 7th and and Das Lied von der Erde were entrusted to Klemperer (who had to arm-wrestle his own 9th from them). As for the ever popular 1st, EMI seemed happy with the 1961 recording of Paul Kletzi it had in its catalog, despite its 24-bar cut in the finale, and replaced it only in 1971, with Giulini’s; and it is not until Tennstedt’s 1979 recording that they put a 3rd Symphony in their catalog. In fact, though Mahler’s star was rapidly rising in the 1960s and `70s and CBS, Decca and Philips had completed their Mahler cycles with Bernstein, Solti and Haitink, EMI simply didn’t seem interested in jumping into the bandwagon, and its only at the beginning of the 1980s, with Tennstedt’s, that they finally had their own (since then, they’ve added Bertini and Rattle).
Comparing Barbirolli and Klemperer seems quite appropriate, because there are many similarities in their conducting style, at least in their EMI years, the 1960s. Both conductors favored very deliberate tempos, compensating in orchestral mass and power the potential sluggishness and lack of tension of their pacing. What Barbirolli retained though which Klemperer in his later years had lost, was the ability to whip up the music to brisker speeds and intense passion. With Barbirolli, deliberation, when he adopted deliberate tempos, was clearly a choice; with Klemperer it wasn’t so clear… In that respect, Klemperer’s 7th Symphony (recorded in 1968) was Klemperian to the point of caricature and distortion, and Barbirolli’s 1964 9th was not AT ALL Klempererian – but Barbirolli’s extraordinary 6th was, to the hilt: Klemperer at his best, deliberation and implacable power: a steam-roller doesn’t move fast, but just try opposing a steam-roller.
I’m bringing up all these considerations because, despite being much lauded ever since it was published, Barbirolli’s 5th is a very problematical interpretation. In fact, some years ago (and before my Amazon reviewing time), I already did a comparative survey of Mahler’s 5th, and thought Barbirolli so wrong-headed that I sold it away – something I very rarely do. And now I know why: I’ve bought it back just for the sake of this new, and even more extensive survey, and only to be able to substantiate what I though was going to be a dismissive review. But now I’m not so sure any more…
Today when Peter Fülöp’s stupendous and indispensable Mahler discograpny lists, as of April 2010, 176 different versions (including Dvds), we tend to forget, but these were times of Mahlerian scarcity. When Barbirolli recorded his 5th, it was only the 7th recording ever published of the symphony, after those of Bruno Walter (1947, Columbia), Hermann Scherchen (Westminster 1953), Rudolf Schwarz (1958, Everest), Bernstein New York Philharmonic Columbia and Leinsdorf Boston Symphony RCA (both 1963), and Vaclav Neumann (1965, Leipzig Gewandhaus Orchestra, originally on the East-German label Eterna, made available in the US by Vanguard Classics – interestingly, on a 2-LP set with exactly the same coupling as Leinsdorf’s RCA set: Berg’s Three Excerpts from Wozzeck, but conducted by Herbert Kegel – see my review for the discographic details). The recordings of Solti (Decca), Haitink (Philips) and Kubelik (DG) were still one or two years away. Among the available versions, those of Walter, Leinsdorf and Neumann more or less followed the same interpretive model – brisk, playing more the military than the funeral, the exuberant joy than the burdened plight -, Schwarz was overall laid-back and easy-going, and Bernstein occupied a kind of middle ground – at least until his uniquely slow Adagietto and his exuberant finale. And Scherchen? Scherchen was, as always, the maverick. But ultimately, the interpretive options explored by these different versions were of a fairly limited compass, and there was scope for more, more extreme, more scraping against the edges rather than driving in the middle of the road.
And Barbirolli provided exactly that. The first movement offers Barbirolli at his most Klempererian, and Klemperer at his best. Very typical is the opening trumpet solo, and especially the upward quarter-note triplet at 0:19 leading to the orchestral outburst. Mahler instructs those triplets to be played “flüchtig”, fleetingly. Barbirolli does the opposite: he slows `em down, like playing those four notes on a giant foot keyboard like Tom Hanks in “Big”. Follows deliberation and mass. The brass eighth-notes triplets are very detached rather than tight, and powerful, conveying a sense of implacability and fate. The 1969 sonics afford details of great presence and vividness (try the big horn and trumpet surge at 6:29, measures 189-191, or the two huge horn brayings in the second movement at 9:31, measures 354-5), and they even pick up of few grunts from the conductor, but they also lack a touch of brilliance (and the remastering from 1998 on Great Recordings Of the Century improves only imperceptibly over the previous one from ten years earlier), and the timpani are disappointingly ill-defined to inaudible (especially regrettable in the second and third movements). Barbirolli’s march advances slowly, with an immense burden on its shoulders: Mahler’s indication to play “schwer” (which in fact appears only measure 262, at 8:20) has not escaped the conductor’s attention, and he extends it to all the movement’s marching. He reaches the fast section in 5:50. For reference, Walter was there in 4:41, Leinsdorf 4:54, Bernstein 5:09 and Schwarz 5:19. Not until Bernstein’s 1987 remake in Vienna for DG was the opening funeral march taken slower – although, interestingly, Boulez conducted it in a concert with the BBC, a year before Barbirolli’s recording but published only in 1990, in a very similar fashion (Nuova Era 2326).
Barbirolli also shows a great sense of architecture, eschewing the extreme of contrasts favored by Scherchen (and, later, by Karajan, in his 1973 recording on Deutsche Grammophon), and his fast section remains very controlled in tempo (the section is taken in 1:46, same as Bernstein in both his recordings, as opposed to Leinsdorf’s 1:32 or Scherchen’s hectic 1:17. Walter here was uniquely slow, at 1:56, followed by Schwarz and Boulez with the BBC, 1:53). In Barbirolli’s favor is also the fact that his trumpet is less glaring, more embedded in the orchestral texture than his counterparts from New York or Boston. Returns the slower pacing and comes the last section at 10:50, and Barbirolli’s expansive tempo develops a unique sense of despaired wistfulness, but thanks to the conductor’s beautiful control of rubato it never sounds dragged (for the anecdote, the New Philharmonia violins do not play the first upward surge as written, E-A-C, but E-C-luftpause-C; I have no idea on what authority Barbirolli instructed them to do that).
Deliberation and massive power still in the second movement. Walter and Leinsdorf had taken it with unique urgency and raging bite, and Bernstein had not quite emulated that. Schwarz was more deliberate but still biting and powerful. In a very similar approach Barbirolli is even more deliberate than Schwarz (first section taken in 1:35, to Schwarz’ 1:28, Bernstein’s 1:24, Walter’s 1:19 and Leinsdorf’s 1:18), and equally, or perhaps even slightly more biting and powerful (screaming brass!). The effect is less furious than with Walter and Leinsdorf, but more threatening. Where Barbirolli diverges from Schwarz is in his adoption of a markedly more spacious tempo even in the slower sections, with their reminiscences of the first movement; both here are in fact being very consistent with their first movement tempo choices. Needless to say, these moments under Barbirolli have great atmosphere, full of pent-up tension. On the other hand, Barbirolli doesn’t stage much contrast between slow and fast. There were moments where I wondered if it wasn’t starting to sound only sluggish and ponderous – I must have thought so some years ago when I decided to sell my CD. Now, frankly, I don’t know. I depends if you are taken or not by the power that emanates from it, and re-listening to it now, I was. But to be honest, listening to Karajan’s 1973 recording in direct comparison, the latter is better, in a similar approach: there is an added bite, added urgency, none of Barbirolli’s traces of plodding, the great chorale before the end (Barbirolli’s 12:11, Karajan’s 12:06) is more grandiose, abetted by the better sonics. For the anecdote, Barbirolli didn’t hold his record of the longest running second movement for very long (15:07). According to Peter Fülöp’s table of timing, the slowest ever in this movement was Wyn Morris in 1973, at more than 17 minutes. Wonder how that sounds like – no, in fact, I don’t think I want to know. Bernstein in Vienna in 1987, for all his Barbirolli-like deliberateness, was overall slightly faster, at 14:57. Abbado with Chicago in 1980 was exactly at the same timing as Barbirolli, but through radically different means: his fast was fast and his slow very slow, which doesn’t give the same sense of deliberation as Barbirolli.
Barbirolli’s Scherzo is more within the bounds of whatever tradition existed at the time – in fact, it is quite similar in approach to Bernstein’s and Leinsdorf’s (here straying from the Bruno Walter model), also the one adopted afterwards by Levine or Boulez: moderate in tempo, leisurely and good-natured in mood rather than with a feeling of urgency. The affecting little waltzes at 2:30 and 6:56 are phrased with an affecting simplicity – which is more you can say of Bernstein – and the moments of romantic nostalgia have great romantic nostalgia, thanks to Barbirolli’s expansive tempo and great instrumental character and warmth (as the passage at 7:50, measure 345, starting with a surge of violins). But the movement is not just about good-natured bonhomie or romantic nostalgia, and there are all those moments written “wild” by Mahler with their frenetic eighth-note string runs (3:54, 10:29, 12:56, 14:53), and if the performer doesn’t let the listener feel the heat of the desert wind on his skin, he’s missed it. I won’t say that Barbirolli is the most scorching reading there is, even back then – a brisker tempo does help, and Bernstein’s and Leinsdorf’s fine elasticity of pacing allowed them to do just that – but he keeps the heat level sufficiently high, and any potential feeling of sluggishness is dispelled by the great instrumental pungency, vividness and bite of every section of the orchestra. And Barbirolli is a supreme conductor: he finds all those little tricks of phrasing or balances that make it work, like giving a strong accent on the first note of the typical biting little theme first played by woodwinds at 0:46 (but here under Barbirolli it lacks a touch of sardonsim), typically when it is repeated by trumpet at 1:30 and upon each of its returns (especially when it is underpinned by clap-boards at 10:55), to the point almost of distorting its rhythm and changing its first double eighth-note into two sixteenth or a repeated-note apoggiatura. Or again, at 10:31, hear how his balances strongly favor the horn ff harmonic underpinning at the expense of the anguished violin melody, which, at his deliberate tempo, might not have carried enough tension. Listen also to his (unindicated) staccato phrasing from the horns then trombones at 13:03. The coda, at 17:10, almost comes as a shock – no more deliberate bonhomie, Barbirolli rockets into it.
Beautiful, affectionate Adagietto, quite “normal” in its choices of tempo – at 9:50, it is midway between the Mengelberg/Walter tradition (here also followed by Schwarz) and its 7 to 8 minutes’ timing, but slower than Bernstein (11:00), and very similar to the way Solti did it a year later, although Barbirolli’s less glaring sonics give it more an impression of softness and restraint, which doesn’t preclude moments of more passionate animation.
Timings never tell the complete story, but they do tell the beginning of a story, and a simple look at those of Barbirolli’s finale will point to the “problem”: 17:22. Bruno Walter took it in 14:08, Bernstein in 1963 in 13:43 and even Bernstein in 1987 played it in 14:57, Schwarz in 16:43. So you know this is going to be extreme, and it is. Not a sense of exuberant, elated joy (Mahler writes “Allegro giocoso” and “frisch”, lively) but a kind of clumsy, genial and good-natured bonhomie, pastoral, closer to the first movement of the 4th or the Ländler of the 9th, or even, why not, to Schubert’s Trout Quintet, or maybe Mr and Mrs taking the baby out in the stroller for a little countryside digestive outskirt after lunch (but naturally it rises to more intensity than that; they must have inadvertently entered a secret doorway to Oz and marvel at what they see). Not that the approach was unprecedented – Scherchen and Schwarz had adopted a similar tempo and Mitropoulos (in the 6-CD set Music & Arts CD-1021 barcode 017685102127 with Symphonies Nos. 1, 3, 5, 6, 8, 9, & Adagio 10) an even slightly slower one in concert in 1960, but around the 6 minute mark for Scherchen (measure 273, rehearsal figure 10) and the 9 minute mark for the two others (measure 423, rehearsal figure 17), with the build-up of dynamics and tension they imperceptibly accelerated; Barbirolli hardly budges until the coda, and even there, he goes for mass more than for speed. Again, Barbirolli didn’t hold his record for long: in a recording made in the studio for the Austrian Radio in February 1971, but released on CD only in 2000 on Berlin Classics, Hans Swarowsky took it in a jaw-dropping 18:44. Even if I didn’t want to know what that sounds like, I happen to have this version and I’ll soon review it.
Exceptionable the approach certainly is, and that finale was the last straw for me those few years ago. But now… I’ve grown more welcoming to the “alternative” approach, and as exceptionable as it is Barbirolli realizes his conception with great instrumental flavor and robustness. So forget what (you think) it should sound like, or what (you think) Mahler wanted it to sound like, and listen to Barbirolli for the unique atmosphere he has to offer. Certainly, this is unique, but be aware that this is not a “fair” rendition of the finale, but an extreme and outlandish one. Not the version to have if you have only one, but one certainly to have if you want to have a complete view of what ranges of music the notes of Mahler can yield.
And that is in fact a comment that I can extend as a conclusive one to the whole recording. This time, I’m not selling it back.
As I mentioned in the review, the 1998 remastering improves only imperceptibly over the one initially published on the budget collection Studio in 1988. Whatever sonic weaknesses (a lack of brilliance at times, that anyway is perceptible only when jumping to other and better recorded versions, and muffled to inaudible timpani most of the times) appear to be in the master tape and not solvable by remastering. So nobody needs to hesitate in buying (or not replacing) the old EMI studio from 1988 or Studio Plus from 1993.
And a post-script from the next day. Latest news: since I am moving on more or less chronologically in my comparative survey of Mahler’s 5th, I’ve just listened to Swarowsky’s version (see above). This is jawdropping. It is Barbirolli – but better, even more extreme deliberation, even more power and bite (I thought Barbirolli’s 2nd movement was biting!), even more vivid and present sonics (and I thought EMI’s sonics were fine!! They sound muffled in comparison). Fans of Barbirolli, don’t miss it. Fans of Mahler’s 5th, don’t miss it.
The subsequent reissues (Japanese releases listed only when audiophile editions – SMH-CD or SACD – not otherwise available in the West):
HMV Classics vol. 18 7 67622 2 (1992) 077776762227:
EMI Studio plus CDM 7 64749 2 (1993) 077776474922:
VSM 4 78974 2 (1993?) 724347897429:
EMI Great Recordings Of the Century (HMV logo) 5 66910 2 (1998) 724356691025:
EMI Great Recordings Of the Century (Angel logo) 5 66962 2 (1998) 724356696228:
Collected in 16-CD set EMI 4 57767 2 (2010) 5099945776724 “Sir John Barbirolli The Great EMI Recordings”:
(SACD) Esoteric (Japan) ESSE-90057 (2011) 4907034217717:
(SACD) EMI Japan TOGE-15018/19 (2012) 4988006897878:
EMI/Warner 4 33290 2 (2013) 5099943329021:
Collected in 5-CD set Warner Classics (2021) 0190295004286 with Barbirolli’s recordings of Symphonies Nos. 1, 6, 9 and Lieder):