Bruckner Symphonies: An Introduction and Review of Selected Recordings
(Updated version of April 2009)
by Patrick Waller and John Quinn 

This article was originally published in 2005. We have now updated it to include the recordings we have heard for the first time over the past four years and to add links to all the relevant reviews on MusicWeb International published by the Spring of 2009. As further Bruckner symphony reviews are published they will be added to the links which are provided in the Masterwork index here.

As one would expect, in some cases our views and recommendations have changed a little over time. We would like to thank readers who contacted us about the article or posted on the Bulletin board for their comments. Further feedback and discussion via MusicWeb International’s forum would, of course, be welcomed. 

Introduction

The Austrian composer Anton Bruckner (1824-1896) struggled to gain recognition in his lifetime but is now universally acknowledged as a major symphonist. His music is quite original and has a characteristic style, which led to the suggestion that "he did not compose nine symphonies but one symphony nine times". In fact, he wrote eleven symphonies in all and each has much individuality. Together they represent a series of increasing power and profundity akin to a great journey.

Our purposes here are three-fold. First, to provide a brief introduction to each of the works; secondly, to consider the merits of the recordings which either or both of us have heard and finally to make some recommendations for collectors who are coming afresh to the music of this composer. It should be borne in mind that our combined experience of Bruckner recordings is far from exhaustive and that the choices made are necessarily personal ones.

Our perspective is that of listeners rather than musicologists (which we are not) but it is impossible to write about these symphonies without addressing in some way the question of editions. During his lifetime Bruckner’s symphonies were living creations and were influenced by (a) his desire for them be heard, even if cut, and (b) the suggestions and tamperings of others, notably the Schalk brothers. This led to numerous revisions and the existence of various editions. During the century after his death a lot of dust settled, some editions were discredited and definitive editions were published under the auspices of the International Bruckner Society. There remain five symphonies (Nos. 1-4 and 8) for which there is effectively more than one version and for which a listener will have good reason to question which edition is being played. We will deal with major issues in relation to the versions of each work as we go along, although there are many points of detail that are beyond the scope defined above. We will make no attempt to argue which editions are "best". It is perfectly possible to regard each separate version as a valid work in its own right and, indeed, admirers of this composer will surely want to hear all of them.

In addition to the nine numbered symphonies there are two forerunners, which are now usually known as Nos. 00 and 0. The latter was mostly composed between the 1st and 2nd symphonies but, in view of its number, we shall consider it before the first symphony. All Bruckner’s symphonies have four movements but the 9th was left complete only as far as the end of the third movement. There are extensive sketches for the finale and realizations exist which have been recorded but these are not considered below. With the exception of the (relatively rarely recorded) first version of the 2nd, as far as the 7th symphony, the slow movement is placed second and scherzo third. In the 8th and 9th this order is reversed.

In compiling this article, we have found John F. Berky's online discography invaluable. Disc numbers, dates of recordings, timings for individual movements and detailed information about editions can all be accessed on or through this site, regardless of whether or not a disc is currently available.

Symphony No. 00 (Study Symphony)

This work dates from 1863 when Bruckner was aged 39 and is also known as the Studiensymphonie since it was written whilst he was taking lessons in orchestral composition from Otto Kitzler. In F minor, it takes about 40-45 minutes to perform if the exposition repeats in the first and last movements are included (they are omitted by Tintner – see below). Classical in structure and perhaps most influenced by Schumann (whereas Schubert was a bigger influence on later works), this is the least individual of the symphonies but it is possible to hear some things that foreshadow his later style. Bruckner never intended it to be more than an exercise and, given the problem he had getting his works performed, would doubtless have been amazed if he knew that anyone would be interested in it 140 years later. Since it does not provide the kind of powerful and uplifting experience that is the hallmark of this composer, it is primarily something to listen to for interest but it could also be a useful teaser for friends who think they are good at identifying composers.

Recordings of the Study Symphony

No need to detain ourselves for long here. The discography lists only a few recordings to date and one should suffice. Georg Tintner’s is the most obvious choice since it is well played and recorded, and at bargain price on Naxos. This also has a useful coupling that is otherwise going to be hard to find – the 1878 Volksfest finale from the 4th Symphony - an interim version between the original of 1874 and the final version now normally played (see review).

Review Index

Symphony No. 0 (Nullte)

This symphony is in D minor, a key shared by his third and ninth symphonies, and could be looked upon as a precursor of those later works. It was started in late 1863 (i.e. shortly after No 00) but then set aside until after the Symphony No 1 had been composed and performed. Bruckner revised the score (mainly, it seems, in the middle movements) in 1869 and the original version does not seem to have survived. Subsequently the score was shown to the conductor Otto Dessof in the hope of a performance, he is reputed to have looked at the opening and said "but where is the main theme?" Bruckner was devastated and effectively withdrew the symphony but fortunately he did not destroy the score. This is an attractive work of some substance (even if the main theme of the first movement is presented very tentatively) and the slow movement is particularly fine.

Recordings of the Nullte

Quite a few recordings of the Nullte have been made but it is never going to be a money-spinner for a record company, and they come and go quickly. Tintner is again the most obvious choice in a double-disc bargain price coupling with the (rarely recorded) first version of the 8th Symphony on Naxos (see review).

Chailly’s version with the Berlin Radio Symphony Orchestra (Decca) and Haitink’s generally swifter reading on Philips are also recommendable.

Review Index

Symphony No. 1

Bruckner’s first numbered Symphony is in C minor. It was composed in Linz in 1865-6 but extensively revised in Vienna in 1890-1, i.e. after the composition of the 8th symphony. The first performance was conducted by Bruckner in 1868 and did not make much impact. However, it is a highly original work, often more forthright than most of his later symphonies. The earlier version is much more frequently played than the revision, most conductors perhaps feeling that the changes to the scoring (there are few important structural revisions) made in Bruckner’s old age do not fit with the work of a much younger man.

Recordings of the First

Linz version

Haitink’s recording of Haas’s edition of the First Symphony with the Amsterdam Concertgebouw Orchestra is taut and brings out the originality of this music. Now available in a Philips Duo set, the reissue has the benefit of being coupled not only with Haitink’s earlier (1965) reading of the Ninth, with the Concertgebouw, but also with a blazing account of the Te Deum, set down with the Vienna Philharmonic in 1988.

However, to hear what are truly Bruckner’s first thoughts one should turn to Tintner’s recording of the edition recently prepared by William Carragan on Naxos. Most of the differences are in the finale and have much less impact than the revisions made in Vienna (see review).

Vienna version

The extensive revision of this symphony, which Bruckner made in Vienna, has been recorded quite rarely. Both Chailly and Wand have set down decent performances in acceptable sound. Chailly’s version is slower in all four movements and lasts for 54 minutes – six minutes longer than Wand.

Review Index

Symphony No. 2

Bruckner’s Second Symphony is in the same key as the first (C minor) but is less dark in feeling and represents a big step towards his later style. The use of pauses between some paragraphs was very unusual at the time of the first performance in 1873 (under Bruckner) and led to the nickname Pausensinfonie. This doesn’t seem to have stuck, presumably because it could equally be applied to later works. There are several editions of this work, which was originally completed in 1872. In the first version the scherzo was placed second but Bruckner revised the work in 1876-7, reversing this order and making various changes and cuts, particularly to the finale. In 1892 he made some further revisions but, to date, that edition has not been recorded. In practical terms, most recordings are of the 1877 version, of which there are Haas and Nowak editions. In this respect, the situation is similar to that pertaining to the Eighth Symphony: Haas reinserted some cut passages from the original back into the revision whereas Nowak omits them.

Recordings of the Second

1872 version

In 2005 we said that choice was easy here, partly because there are so few recordings and partly because Tintner’s Naxos reading with the National Symphony Orchestra of Ireland is wonderful in every way (see review). We remain impressed by Tintner’s performance, however, in 2008 Simone Young’s reading with the Philharmoniker Hamburger was issued (see review), and JQ marginally prefers the new recording although it is more expensive. 

1877 version

Haitink’s version (Haas) from 1969 is one of the highlights of his complete set and it is perverse that it seems to be the only one of the numbered nine in this series not to have been reissued as part of the Philips Duo series.

Wand’s 1981 reading with the Cologne Radio Symphony Orchestra is also of the Haas edition but is marginally less convincing.

Providing you are not wedded to Haas, Giulini’s reading of the Nowak edition with the Vienna Symphony Orchestra from 1974 (on Testament) is a clear first choice amongst 1877 editions. The sound is satisfactory and Giulini’s affinity with this composer shines brightly throughout the work. Indeed he seems to have had a particular feeling for this symphony, it being the only one of the early works that he recorded. There are various recorded versions of numbers 7-9 but we can find no evidence that he set down any of the others.

Review Index

Symphony No. 3

The composition of the Third Symphony perhaps caused Bruckner more grief than any other. In particular the first performance, conducted by the composer in 1873 was a complete disaster and the beginning of the serious criticism he received in the press from Eduard Hanslick. On the plus side, however, Wagner accepted the dedication of the work and it is sometimes known as the ‘Wagner Symphony’. Although some musicologists regard this work as rather weak, from a listener’s perspective it is splendid and probably the first of this composer’s symphonies than can be regarded as standard repertoire. Amongst many marvellous moments (including, for example, the mysterious opening and trumpet theme), the juxtaposition of polka and chorale as the second theme of the finale is unforgettable. Apparently Bruckner once explained that the idea for this came from passing a music hall whilst the coffin containing a famous architect was laid out in a nearby building.

In terms of versions, there were two major revisions to the original and in each case the work became shorter in an attempt to solve structural problems. The final version is most often played and recorded despite there being little doubt of the influence of the Schalk brothers.

Recordings of the Third

1873 version

Tintner’s version is well played by the Royal Scottish National Orchestra and an obvious choice at bargain price although some may feel that his tempi are too slow, particularly in a first movement which lasts just over half an hour. The whole symphony here takes nearly 78 minutes in comparison to Inbal’s 65 minutes for the same edition - this is an alternative to Tintner but we have not heard it.

1877-8 version

Bernard Haitink’s Amsterdam recording dates from 1963 and was the first of his complete series. It is a very convincing interpretation, which is slightly tauter than his 1988 remake with the Vienna Philharmonic Orchestra. There are some slight differences in the editions used by Haitink – the later version uses Nowak’s 1981 edition which includes a brief coda to the scherzo. The sound provided for Haitink was more than acceptable in 1963 and quite spectacular in 1988 (see review).

A live recording by Jascha Horenstein appeared on BBC Legends in 2007. This is a performance given at the Cheltenham Festival in 1963 when he conducted the BBC Northern Symphony Orchestra (nowadays the BBC Philharmonic). As with several other BBC Legends releases featuring this conductor, one is impressed by the integrity and dignity of the music making. It may not be a completely flawless performance but it’s very persuasive, although the scherzo is, perhaps, a trifle steady in tempo.

1888-9 version

Barbirolli’s 1964 performance is somewhat hobbled by the recorded sound which conveys little, if any, atmosphere and is constricted and rather shrill at climaxes. In the first movement Barbirolli doesn’t really give the music enough time to breathe. Happily, matters improve thereafter but overall this is somewhat disappointing (see review).

Skrowaczewski conducts the Saarbrucken Radio Symphony Orchestra in a very forthright and well-played reading from 1996. This perhaps misses some of the grandeur that Haitink brings to the outer movements and there is an odd tempo for the second theme of the finale. This is taken substantially more slowly than usual and doesn’t really come off. The sound is excellent and, if it can still be found at budget price, this would be a good choice for the final version.

Celibidache’s Munich Third (see review) is very broad in the opening movement but not otherwise particularly controversial and the playing is wonderful.

Finally, it is worth mentioning a recording of the version for piano duet, mostly arranged by a very youthful Mahler (see review). If you are sceptical just have a listen to the scherzo and you might be surprised! 

Review Index

Symphony No. 4 (Romantic)

The first symphony Bruckner wrote in a major key (E flat) was subtitled the Romantic and it was then followed by three more in major keys. Although its genesis was almost as problematic as the third and, despite arguments that he never quite solved the structure of the finale, this has become his most popular work. The opening tremolando and horn calls are particularly striking, as is the "hunting" scherzo. The original version contained a completely different scherzo and the finale went through two major revisions before reaching the form in which is it now almost invariably heard in 1881. Differences between the Haas and Nowak editions of this version are minimal. There are also later editions dating from around 1888 but the revisions were probably not Bruckner’s work. Newcomers to this composer would do well to start here by acquiring a recording of the 1881 version of this symphony or, better still, hearing it live.

Recordings of the Fourth

1874 version

Tintner did not record the first version of the Fourth Symphony but Inbal’s Frankfurt reading of 1982 is very well played and recorded. At bargain price it is an obvious choice. This is a disc which will surprise those who do not know the first version and it is well worth hearing. As with the Second, Simone Young opts for Bruckner’s first thoughts and JQ continues to be impressed (see review) although this is at full-price.

1881 version

There is a huge choice of recordings of this version of the Fourth Symphony. During the 1960s and 1970s Walter, Klemperer, Jochum, Haitink, Böhm and Karajan all made recordings in sound which is at least acceptable. Klemperer’s is the most idiosyncratic (see review). His is also the least "romantic", adopting fast tempi for each movement except the scherzo (which is slower than normal). Jochum’s is also a very personal approach but his 1965 reading with the Berlin Philharmonic Orchestra is more successful than his Dresden remake and conveys a great feeling of mystery when needed. Bruno Walter was undoubtedly a great Bruckner conductor (but under-recorded as such) and his 1960 recording with the Columbia Symphony Orchestra has fine sound for the period. Haitink’s Amsterdam version, which dates from 1965, is perfectly satisfactory but eclipsed in every respect by his Vienna reading of 1985 (see review).

Karajan’s DG recording of 1975 has excellent playing from the Berlin Philharmonic Orchestra (see review) but sound is not ideally clear. Preferable in almost every respect, and a plausible first choice, is Karl Böhm’s 1973 Vienna recording for Decca – this has a rightness about it which shines throughout.

For a different experience to any other version, Celibidache’s 1988 live reading from Munich is well worth hearing. It is very slow, particularly in the finale (at 79 minutes in total it only just squeezes on a single disc – 65 minutes is about par), but the concentration is extraordinary and there is a real sense of occasion.

Klaus Tennstedt recorded the symphony for EMI with the Berlin Philharmonic in 1981 (see review). So far as we know this, and the Eighth, were the only commercial recordings he made of Bruckner, which is a shame. In the Fourth, aided by superbly rich and sonorous playing from the BPO, he conveys the majesty and, where appropriate, the mystery of the work very convincingly. While this may not displace Böhm or Wand, this is still an impressive version.

Günter Wand’s 1998 recording is also with the BPO. It’s a superb achievement, an exalted performance that is majestic and unhurried. The Berlin orchestra plays magnificently. This is one of the finest readings of this symphony ever committed to disc (see review). There’s a later recording by Wand in which he conducts the NDR-Sinfonieorchester. This is taken from live performances in Hamburg in October 2001 and comes in a two-disc set from BMG-RCA, coupled with Schubert’s Fifth symphony. It possesses all the integrity of Wand’s Berlin account but does not surpass it. 

Giuseppe Sinopoli’s Bruckner recordings are surprisingly objective given his generally interventional approach to music. His Fourth is extremely well played (see review) but the approach probably suits this work less well than in the Fifth symphony (see below).

Simon Rattle’s recording with the Berlin Philharmonic has many good things but also some lapses and JQ regards this as “work in progress” (see review).

Of historical interest is a recording made in The Hague by Willem van Otterloo in 1953 (see review). This is a pretty impressive traversal but it only seems to be available in a 13CD box, which also contains Bruckner’s early Overture and works by many other composers. 

Review Index

Symphony No. 5

Whatever reservations musicologists may have about the preceding works, no one seems to doubt that from the Fifth onwards Bruckner’s symphonies are masterpieces. One could even argue that the Fifth is the finest and it seems very sad that Bruckner never heard it (the first performance was given in 1894 in Graz but Bruckner was too ill to attend). The opening movement is both massive and bold. The adagio begins with a memorable theme for the oboe playing two against three for the string accompaniment, and also contains a striking second subject for the cellos and basses. At the beginning of the finale Bruckner briefly recapitulates the preceding movements à la Beethoven’s Ninth and, later in this movement, the entry and integration of a magnificent brass chorale is one of his most stunning achievements. Although there are Haas and Nowak editions of the Fifth they are essentially identical. There is just one real issue of editions here and that is the existence of a now discredited version made by Franz Schalk in which the magnificent finale is cut and re-orchestrated. This version has been recorded a few times notably by Knappertsbusch and quite recently by Leon Botstein. Our tolerance of different versions does not extend this far and, unless you want to hear them for interest, we suggest that these recordings should be avoided.

Recordings of the Fifth

Klemperer’s 1967 reading with the New Philharmonia Orchestra is slightly disappointing. He takes a massive approach but, overall, it doesn’t quite hang together and this is not as fine as his Sixth (see below).

Haitink’s 1971 reading from Amsterdam is excellent but, as for the Fourth Symphony, it is not as fine as the Vienna remake from 1988. Haitink’s later versions of these two symphonies are available coupled together at bargain price (see review).

Amongst more recent recordings, Giuseppe Sinopoli’s rendition from Dresden in 1999 stands out. Stated to be a live recording, and with excellent sound, there is no evidence of an audience but this reading is superbly concentrated and lacking in idiosyncrasy (perhaps unusually for Sinopoli).

Horenstein’s is an epic, live reading, captured at the Henry Wood Promenade concerts in 1971. It’s now available on BBC Legends (though we have not heard that particular transfer). The Fifth is, arguably, the most intellectually rigorous of the canon and a successful performance needs a seasoned Brucknerian at the helm. Horenstein is just the man for the job. He plays the second movement at a nice flowing tempo and his command and control of the vast finale compels admiration. This is a very fine recording indeed and, in the opinion of JQ (who has not heard the Sinopoli version), a first choice (see reviews by John Phillips and Tony Duggan).

Wand’s Cologne version from 1974 is, it seems, his very first Bruckner recording (see review). He displays a mastery of the score, especially in handling transitions. Enjoyment of the performance is limited slightly by a rather enthusiastic brass section, whose playing is somewhat shrill at times. Wand’s control is especially evident in the massive finale. This is a good performance but his 1989 account with the NDR Sinfonieorchester is even finer, not least because the playing is better. There’s not a great deal to choose between the versions in interpretative terms but the later performance also benefits from a warmer, more pleasing recording (see reviews by John Quinn and Terry Barfoot). There are also Wand performances with the Munich Philharmonic Orchestra from 1995 (see Terry Barfoot’s positive review) and Berlin Philharmonic Orchestra from 1996. PW has downloaded the latter from AmazonUK (for a little over £3) and greatly admires this performance.

As so often in his Naxos cycle, Tintner is courageous in the breadth of some of his tempi. His 1996 version takes some 4 minutes longer than Horenstein’s. The studio-made recording doesn’t have quite the same electricity as Horenstein’s but, like the rest of his cycle, the reading is deeply considered and completely idiomatic. It is excellent value and unlikely to disappoint (see review).

Review Index

Symphony No. 6

The Sixth is the shortest and lightest of the later works and Bruckner thought it was his finest. Sadly, today is probably heard and recorded less often than the others. The composer did hear the middle movements (the adagio is marvellous) but the whole work was not performed until 1899, when Mahler conducted. There are no important issues relating to editions here – Haas and Nowak are essentially the same.

Recordings of the Sixth

Klemperer’s 1964 reading is generally highly thought of, a view we share. It has a rugged quality but is less idiosyncratic than some of his other Bruckner recordings. The sound is pretty decent.

Haitink’s first recording is underrated and, in our view on a par with Klemperer (see review). He also made an excellent recording in Dresden in 2003 (see review) which is an ideal choice for those looking for modern sound at less than full price.

To Günter Wand’s 1976 reading with the Cologne Radio Symphony Orchestra we can only give a qualified recommendation. The first movement seems rushed - see review for more details. His 1996 NDR Sinfonieorchester version recorded in Lübeck cathedral and issued on DVD is much more compelling (see review).

Celibidache’s Munich Sixth (see review) is very well played and recorded and his tempi are not particularly slow except in the slow movement.

Sir Colin Davis has recorded the work live with the London Symphony Orchestra. We have rather differing views of this recording. For PW, in terms of performance this is generally excellent, only a scherzo which drags preventing this from being a potential bargain choice. JQ, however, finds the rather suffocating, close acoustic of the Barbican Centre (often an issue in the LSO Live series) results in the brass being too dominant. Certainly neither of us would rank it ahead of either Klemperer’s rather gaunt reading or Haitink’s warmer vision.

Tintner’s recording is, uniquely in his cycle, with the New Zealand Symphony Orchestra. It was the first to be recorded and dates from 1995. The outer movements of this work are very hard to pace. Tintner seems to judge both movements very well. He also conveys nobility in the adagio and his tempo for the scherzo is infinitely preferable to Colin Davis’s trudge. The New Zealand orchestra may not be quite in the LSO league but they play very well and are recorded in a much more sympathetic acoustic. This seems to be a clear bargain choice.

Review Index

Symphony No. 7

The Seventh Symphony gave Bruckner his first taste of success when it was first performed under Nikisch in 1884, although his critic Hanslick was not appeased. It is a marvellous work with many contrasts, containing some of the sunniest and darkest of the composer’s music in the opening two movements followed by a fantastic scherzo. The last movement lacks the massive proportions of the Eighth but nevertheless brings a satisfying conclusion. The biggest issue relating to editions seems to relate to the inclusion of a cymbal clash at the climax of the adagio. This was originally suggested by Nikisch and added to the score but Bruckner seems to have been uncertain about it (although there is no doubt about the inclusion of the clash at the analogous point of the Eighth symphony). The Haas edition does not include the cymbal clash (whereas Nowak does) but some conductors, such as Karajan, have used Haas and then added the clash. It is a momentary event in a symphony lasting between 60 and 70 minutes and, for us, is not a major consideration in choosing a recording.

Recordings of the Seventh

Jochum’s 1939 performance with the Vienna Philharmonic Orchestra turns out to be something of a curate’s egg. Its particular distinction lies in an extremely spacious reading of the great Adagio. Sadly, the rest of the symphony does not come off so well and the recorded sound, from Telefunken originals, requires some tolerance (see review).

Recently a live recording by Bruno Walter and the New York Philharmonic has appeared on Testament. This preserves a concert performance given in Carnegie Hall in December 1954. The pacing of this performance is surprisingly urgent: it lasts a mere 55:48. It has to be said that there are times when one wishes the music had been given a little more space but set against that is the electricity of the performance and, whilst it would not be a library choice, this recording is well worth hearing.

Haitink made two recording with the Amsterdam Concertgebouw. His 1966 reading (which was included in the complete set) is generally quicker than in 1978, especially in the first movement. Only the earlier version seems to be currently available (on a Philips Duo – see review) but this is a pity since the 1978 traversal is preferable and one of the finest versions we have heard. Fortunately Haitink has also recorded the work quite recently with Chicago Symphony Orchestra, a performance which has comparable stature (see review) and the sound is excellent.

Karajan’s 1989 reading with the Vienna Philharmonic Orchestra was his last recording. There are many splendid things, including the recorded sound, but there is also something a bit detached about the performance and the very end does not quite hang together. This is not really on a par with his last recording of the Eighth and his Berlin reading from 1970-1 on EMI is preferable as a performance (see review).

Georg Tintner’s 1997 version with the Royal Scottish National Orchestra is amongst the best available and an obvious bargain choice.

Carlo Maria Giulini recorded the Seventh in 1986 for DG with the Vienna Philharmonic. We haven’t heard that account but there is a 1982 reading with the Philharmonia on BBC Legends (see review). This is described as being of the Haas edition but this may be incorrect since Giulini apparently used Nowak editions for his DG recordings of all the three last symphonies. The warm lyricism of this symphony suits Giulini very well. A sense of line is always a hallmark of his conducting and this is very apparent here, making us regret there is no recording by him of the Fourth, so far as we know. The adagio is profound with a grave nobility that is impressive. The climax, with cymbal clash, is finely achieved after which the coda glows nobly. There is an appealing freshness in Giulini’s handling of the finale though a touch more breadth would have been welcome in the concluding peroration where there are also a few small brass "fluffs".

No such blemishes are heard in Günter Wand’s 1999 reading, which was taken from concert performances with the peerless Berlin Philharmonic. This is a most impressive recording, characterized by long phrases, each one given just the right amount of time to breathe. The adagio is patrician and elevated. The climax of that movement, in which the cymbal is eschewed, is marvellously built after which the coda is profoundly satisfying. Wand plays the finale splendidly, pacing it with wisdom. The movement is crowned by a peroration of golden splendour. With excellent recorded sound this seems to be a clear first choice. Günter Wand’s reading 1980 from Cologne is well played and a very authoritative interpretation, which can be recommended confidently. The re-mastered sound is very good (see review).

James Loughran’s Aarhus recording of 2005 is a bit plain spun and not recommendable in the face of the competition from Wand and Haitink. 

Finally, for the adventurous, there is a very well-recorded reading of the chamber version of this work, made by Hanns Eisler, Erwin Stein and Karl Rankl in 1921, in an attractive performance given by Thomas Christian ensemble on MDG (see review). 

Review Index

Symphony No. 8

The Eighth Symphony is in C minor with the adagio in D flat placed third. The first two movements both last about 15 minutes, the adagio is immensely long-breathed and takes about 25 minutes and the finale takes at least 20 minutes. This is the only symphony in which he used a harp, to great effect, notably in the trio of the scherzo and in the adagio. The work was first performed in Vienna under Richter in 1892 and was an immediate success.

Bruckner started work on the 8th in 1884 and it was complete in 1887. He sent the score to one of his admirers, the conductor Hermann Levi and was distraught at his less than enthusiastic response. Nevertheless he then spent three years revising the symphony in many ways. Most importantly, he omitted his traditional "blaze up" coda of the first movement and introduced a quiet ending. He also completely replaced the trio in the second movement. The second version of 1890 is almost invariably played in preference to the first version but there is a complicating factor of editions.

In 1939 Robert Haas produced a new edition of the 8th symphony. He believed that a few of the revisions made between 1887 and 1890 were disadvantageous to the structure and/or were the specific suggestions of others, notably Joseph Schalk. He therefore added some passages from the first edition, which had been excised back into the second version. In 1955 Leopold Nowak published the second version of the 8th as Bruckner left it. In 1972, Nowak also published the first version of the score. Haas’s edition is probably still the most widely played and certainly the most recorded version of the work.

See also Patrick Waller's more detailed article on this symphony, which was updated in December 2008.

Recordings of the Eighth

1st version of 1887

Tintner’s recording is one of only a few that have yet been made using Nowak’s edition of the first version of 1887. Tintner adopts rather slow tempi throughout and the adagio lasts for over half an hour. The playing of the National Symphony Orchestra of Ireland and recorded sound are fine (see review). This is clearly preferable to Fedoseyev’s episodic approach in a recording from Moscow dating from 1999 (see review).

Revised version of 1890

Haas edition

Bernard Haitink’s first recording of the 8th was made in 1969 and is, at just under 74 minutes, the fastest reading we have heard. It has both vigour and structural coherence, and is very well played and recorded. His later recording with the Vienna Philharmonic runs for over 83 minutes and is a grander conception in particularly good sound (see review). More recently, there are excellent live Haitink recordings available from Dresden in 2002 (see review) and Amsterdam in 2005 (see review), the latter available on SACD.

Karajan made three studio versions of the 8th (in 1957, 1975 and 1988) plus a version made in 1944 in amazingly good sound, of which the first movement has not survived. The best way to acquire the 1944 torso is as a download from Amazon UK where is can be found for a little over £2. There is also a live version recorded in Vienna in 1957 (see review) but the studio version of that year is probably the better bet (see review). Comparison of the timings of his two 1957 recordings suggests that Karajan exhibited more urgency when performing live. Each movement was slower in the studio recording and overall the difference amounted to more than six and a half minutes. Otherwise, Karajan’s conception did not vary dramatically over the years. The 1988 reading with Vienna Philharmonic is clearly the most desirable. It was one of Karajan’s last recordings and was first issued around the time of his death in 1989. The music making has the quality of affecting you in a way that can’t easily be described. In places there is an "earthiness" that is not normally part of Karajan’s make-up; he had a special affection for the work and surely knew that this would be his last statement about it. The playing and sound are marvellous throughout. There is also a Sony DVD made at the same time, which is not an identical performance but certainly on the same level – this is at bargain price and coupled with an earlier performance of the Ninth (see below).

Boulez made his recording with the Vienna Philharmonic live at St. Florian (where the young Bruckner was a chorister, later returned as organist and is buried) on the 100th anniversary of his death in 1996. It seems that he had not previously recorded any Bruckner nor conducted this symphony. He brings surprising freshness and great coherence to the work, and his reading is highly recommendable on CD or DVD video.

Günter Wand’s 1993 account with the NDR Sinfonieorchester is a live performance and it is very fine. Wand’s final Bruckner recording was made live with the Berlin Philharmonic Orchestra in 2001. It was one of his last recordings and a fitting memorial. Wand and Karajan had little in common but their last readings are not dramatically dissimilar in overall approach. Wand’s tempi are slightly broader and he brings rather more humility and humanity (see review). No one buying the earlier version would be disappointed but his exceptional Berlin reading is finer in terms of playing and recording quality, and that is the Wand version to go for.

Sir Reginald Goodall’s reading comes from a 1969 Henry Wood Promenade concert and is issued by BBC Legends (see review). The first movement is extremely broad. Indeed, at 18’15" it is the longest we know, apart from Celibidache in Munich - he breaks the twenty minute barrier! Unfortunately, in this case the breadth is just too great and the BBC Symphony Orchestra doesn’t sound wholly comfortable. The adagio is intense and concentrated and here Goodall’s basic tempo is more "conventional". The finale is also taken broadly but the pacing is rather more convincing than was the case in the first movement. It’s interesting to hear a great Wagnerian in Bruckner but the reading of the first movement prevents a general recommendation.

No such reservations about Jascha Horenstein’s 1970 recording, which also comes from the Proms (see review). It’s also on BBC Legends (by our reckoning that label has issued at least four recordings of this symphony). This performance displays similar virtues to those that distinguish Horenstein’s Fifth (see above). Above all, structural command and blazing commitment are evident. The adagio is particularly searching and the finale majestic. This is a rather special performance and an attractive proposition for collectors, especially as it’s coupled in a 2 CD set with a fine account of the Ninth (see below).

Barbirolli’s version is also a bit special. Yet again it’s on BBC Legends (see review). This is a concert performance from May 1970. This was to be Barbirolli’s last London concert, for he died just a few weeks later. It’s a reading of astonishing intensity and passion. It lasts a "mere" 74 minutes but the music never sounds unduly rushed, despite the urgency of Barbirolli’s vision. The Hallé Orchestra plays with tremendous conviction and the few slips can be readily forgiven. This is red-blooded Bruckner and certainly not a "safe" library choice but it’s a performance that demands attention and respect.

An historical recording worth mentioning – and listening to – is the one made in Amsterdam by Eduard van Beinum in 1955 which can be downloaded cheaply from Classicsonline. A live performance from the same forces and period exists courtesy of the acetates having been rescued from a flea market (see review). Unfortunately, Carl Schuricht’s Stuttgart reading on Hänssler from the previous year is a fine interpretation let down by the playing.

Nowak edition

Horenstein’s studio recording was made for Vox with the Pro Musica Orchestra in 1955. The sound is mediocre for the period and the orchestra not in the same class as the Berlin or Vienna Philharmonics. And yet there is no doubt that this is a great interpretation. Horenstein’s conception of the work is seamless and tempi are perfectly judged. Although he does not linger in the first movement, an essential sense of mystery is retained. His adagio is perfectly poised, simply ethereal at the close, and the finale is very fine.

Otto Klemperer is notorious for having made quite major cuts to the finale in the EMI recording he made late in his career. However, a live 1957 recording from Cologne is uncontroversial and worth an airing (see review).

Karl Böhm’s Vienna 1976 recording of the 8th is quite middle-of-the-road in terms of tempi. The playing is wonderful and the phrasing has the same natural eloquence that marks out his classic version of the 4th symphony. This performance squeezes onto a single mid-price CD and would be a good mid-price choice for anyone wanting just one version of this work.

Giulini’s 1983 live version with the Philharmonia Orchestra is a wholly dedicated and masterly performance by a great conductor at the height of his powers (see review). A year later Giulini made a glowing studio version with the Vienna Philharmonic Orchestra which has been generally well-received (see review). At the moment it is available as a CD-R reprint from ArkivMusic or can be downloaded from Universal Classics & Jazz. Giulini also conducted the Eighth for the inaugural concert of the World Philharmonic Orchestra in Stockholm in 1985. This has been issued on DVD (see review) but the sound quality is not great and this is probably not the best way to experience Giulini’s Bruckner.

A live 1981 performance in which Klaus Tennstedt conducts the LPO has recently been issued on the orchestra’s own LPO live label. It has been very well received in other quarters. We have yet to hear it but hope to comment on it at our next revision. 

Of specialist interest only is a 1972 live version made by the Boston Symphony Orchestra under William Steinberg during the third and last season of his brief tenure (1969-72) as the orchestra’s Chief Conductor. The edition used is stated in the documentation to be Nowak but with some (unspecified) amendments by Steinberg. We note, however, that the discography cited above lists it as being the 1892 Lienau edition. This is a good, well paced and powerful reading, well played by the Bostonians, and as this under-rated conductor didn’t record the Eighth commercially it is good to have this available. Unfortunately, however, it’s only very expensively available at present in the BSO’s Symphony Hall Centennial Celebration boxed set, available direct from the BSO. 

Edition of 1892 (Ed. Lienau)

Knappertsbusch’s 1955 Munich recording is so quick (only just over 70 minutes in total) that he rushes the symphony off its feet (see review). The playing is scrappy and untidy and this is a version that ill serves the generally high reputation of the conductor. The only thing that seems to be going for it is the potential interest value of the edition used, i.e. the one was used in the very first performance of the symphony, although it should be recognized that has now been essentially superseded by those of Haas and Nowak.

Organ transcription

Organist Lionel Rogg has made and recorded on BIS a transcription for his instrument. This is a swift reading coming in at under 70 minutes but the fast tempi seem to suit the instrument. This is a curiosity but it might appeal to those who like different perspectives on great music.

Review Index

Symphony No. 9

Bruckner dedicated his Ninth Symphony to God and worked on it for several years before his death in 1896. Unfortunately he was initially distracted by what seem to be unnecessary revisions of earlier symphonies, in particular the First. The Ninth is complete as far the end of the third movement adagio. Although it is clear that Bruckner intended a finale on the scale of the Eighth Symphony, sketches show that it was a long way from complete. As for Schubert’s Unfinished Symphony, a performance as far as the end of the slow movement somehow seems complete and it is apposite that the composer’s work should end with an adagio. Realizing that he wasn’t going to complete the finale, Bruckner contemplated the use of his Te Deum as the finale. This would certainly have involved some transposition; its key is C whereas the symphony is in D minor (a key it shares with the Nullte and third symphonies). The opening movement is a massive inspiration which unfolds naturally in long paragraphs. The scherzo is full of demons but has a greatly contrasting trio. The adagio is hard to describe and seems to be not from this world. It opens with a huge and impassioned upward leap on the strings and ends with a soft brass chorale, which brings a great feeling of repose. There are no important issues relating to editions used in recordings.

In recent years there have been various completions of finale, notably by William Carragan. There are several recordings of them but no front rank Bruckner conductor has yet recorded a four-movement version. If you are interested in hearing a completion, it would make logical sense just to download the last movement (see section on Downloading below).

Recordings of the Ninth

Bruno Walter’s 1959 recording is very fine in every way and a possible top choice despite its age.

Haitink made two recordings in Amsterdam, in 1965 and 1981. The former is taut and logical, the latter more ethereal with generally slightly slower tempi and top-notch sound.

Karajan’s 1975 reading from Berlin is amongst his finest Bruckner recordings but, as with the Fourth Symphony, the sound lacks a bit of impact, even in comparison to Walter’s recording. Preferable is a 1985 live recording from Berlin which is available cheaply on a Sony DVD and come in harness with an Eighth from Vienna made at the same time as his last, and greatest, studio recording of the work (see above).

Günter Wand’s finest reading of this work was also made in Berlin and dates from 1998. This is excellent but perhaps not as fine as his last version of the Eighth (see reviews by John Quinn and Patrick Waller). Wand’s 1993 version is a live performance made with the NDR Sinfonieorchester. The orchestra plays well for him though, unsurprisingly, they can’t quite match the Berlin Philharmonic’s tonal lustre. This version has all the command of structure and balance that one has come to expect from this very fine Brucknerian. Once again, Wand’s final, Berlin thoughts are preferable but this NDR version is certainly not a second best; it’s a useful alternative.

Barbirolli’s recording, captured live at the Henry Wood Promenade concerts in 1966 is on BBC Legends, yoked with his reading of Mahler’s Seventh (see review). Like his reading of the Eighth (see above) this has great urgency. However, the performance is not of quite the same stature. In particular there is a worrying tendency to accelerate into climaxes in the first movement. It must be said that the recording quality leaves something to be desired. The strings sound rather thin and the brass tend to be too dominant. Also, at least as captured by the microphones, the internal balance of the orchestra is not always ideal. Barbirolli’s approach to the adagio has typical nobility and sincerity but realistically the appeal of this performance is limited.

By contrast, Horenstein’s 1970 Proms performance is a reading of undoubted stature (see review). The orchestra (the BBC Symphony) plays much better for him than do the Hallé for Barbirolli. It’s noteworthy too that Horenstein takes 61 minutes for the work against Barbirolli’s 53 minutes. The first movement is grand and spacious while the adagio glows. This latter movement is put across with total conviction by Horenstein and his players. This is a masterly reading by a great Brucknerian and the coupling of this Ninth and the Eighth discussed above is a mandatory purchase for Bruckner collectors.

Giulini’s 1976 EMI recording with the Chicago Symphony Orchestra is currently available only in a highly recommendable box of 4 CDs of his recordings with that orchestra. This is a most impressive reading, distinguished above all by Giulini’s care for the musical line. The concluding adagio has all the spiritual elevation that one would expect from this great conductor. He is supported by fabulous playing from the CSO and this is a most desirable version. Giulini’s 1988 studio recording with Vienna Philharmonic Orchestra is also wonderful and there is a live Giulini recording with the Stuttgart Radio Symphony Orchestra from 1996, which is available on DVD and highly recommendable (see review).

Tintner's recording is a fitting culmination to his cycle (though it was not actually the last of the symphonies that he recorded). There's a massive integrity to Tintner's Bruckner in general and this Ninth is no exception. His performance, which is well played by the Royal Scottish National Orchestra, is perhaps a bit more rugged than the visions of, say, Horenstein or Giulini but it is wholly convincing.

Leonard Bernstein hardly recorded any Bruckner at all and a priori it might be difficult to see this as a likely combination. However, a performance he gave in the last year of his life in Vienna is very rewarding and is available on DVD (see review). 

BBC Legends has complemented its recordings by Reginald Goodall of the Seventh and Eighth with a 1974 traversal of the Ninth. We have yet to hear this version but hope to comment on it at our next revision.

Review Index

Recordings of complete and partial cycles

A complete cycle potentially has advantages in terms of cost and consistency of approach but is also likely to have "ups and downs". Interpretatively, Bruckner’s music can be approached in many different ways and that is without even considering questions relating to the different versions of the works. For these reasons, one complete cycle will not be enough for serious admirers of this composer and some supplementation is likely to be required. Even the definition of "complete" is debatable in Bruckner! But we will take a conservative view that inclusion of at least Nos. 1-9 represents a complete cycle.

Eugen Jochum’s Dresden cycle for EMI was made between 1975 and 1980 and includes Nos. 1-9, each on a single disc and in each case representing Bruckner’s final thoughts on the works. Packaged in a slimline box, this is one of the cheapest ways of acquiring the works and Jochum has a considerable reputation as a Brucknerian. There is also an earlier Jochum cycle on DG of which we have yet heard little, although it is in JQ’s pending tray. The EMI sound is not ideal, being balanced rather forwardly and Jochum’s interpretations are interventionist. Whilst there are many wonderful moments, his free approach to tempi will put this out of court for some listeners. In particular, crescendo and accelerando frequently go hand-in-hand, adding excitement but ultimately being counterintuitive to the flow of the music. Perhaps the worst example is on the approach to the climax of the adagio of the 8th Symphony. This is a cycle that many will want to hear but it is not recommendable as a primary choice.

Georg Tintner’s cycle also includes Symphonies 00 and 0 and is very economical (see review of the whole set). In each case Tintner goes for Bruckner’s first thoughts, except in the Fourth Symphony (and No. 0 for which they are no longer extant) where he uses the final version of 1881. Between us we have heard all these readings individually and this is a most recommendable cycle. Some of Tintner’s tempi are rather slow (for example the first movement of the Third and the adagio of the Eighth) but there are many insights and marvellous control of structures. Tintner’s reading of the Second Symphony is particularly special. Various orchestras and locations are involved but the playing and sound are generally excellent. A cycle which does not include the 1890 version of the Eighth will surely require supplementation and it is slightly odd to have all the first versions except the Fourth. Adding Inbal’s recording of the 1874 edition of that symphony plus Haitink’s Philips Duo of the Third and Eighth would be a relatively inexpensive way of covering all of the essential ground in terms of editions (again there are major differences between the 1877 version of the Third used by Haitink and the original score). With these supplements, this would be our recommendation for a starter collection of Bruckner symphonies and (including the supplements) it should currently cost no more than £50.

Bernard Haitink’s Amsterdam Concertgebouw recordings were made between 1963 and 1972. He subsequently successfully re-made the Seventh and Ninth with the same orchestra and then re-recorded Nos. 3, 4, 5 and 8 with the Vienna Philharmonic Orchestra. The complete cycle of the first recordings has been issued more than once on CD. The VPO recordings and some of the Amsterdam recordings are available as Philips Duos (sadly, Haitink’s version of No. 2 is not included). The original cycle is a major achievement with straightforward interpretations and good sound. Haitink’s readings of Symphonies Nos. 1, 2 and 6 are particularly impressive and, re-listening to the 1963 recording of the Third, PW was surprised to find himself wondering if it wasn’t just as fine as the marvellous 1988 recording with the VPO. Tempi are often relatively fast but it could be argued that the general tendency to play Bruckner more slowly nowadays is not necessarily beneficial. No. 0 is included and Haitink uses the Haas editions. This remains a good general recommendation for a complete cycle.

Stanislaw Skrowaczewski has recorded all the symphonies (including Nos. 00 and 0) with the Saarbrucken Radio Symphony Orchestra. The individual symphonies were initially issued on Arte Nova at super-budget price but unfortunately they have become more expensive when collected together. This set has been reviewed by Terry Barfoot and PW. In general Skrowaczewski opts for Bruckner’s last thoughts and his performances are cogent indeed. We hope that this box will be reissued more economically and with better presentation, in which case it would clearly merit a positive recommendation.

Daniel Barenboim has recorded two complete cycles – in Chicago and Berlin. The former seems to have been completely superseded whilst the latter has many positive features (see review). His reading of the Second is the one significant disappointment. This set can usually be found for about £20 or less and is certainly very good value. 

Two other great Bruckner conductors have also made complete cycles which are currently available – Wand and Karajan. JQ has recently reviewed the Karajan set (see review) when it appeared in a budget slimline box. In both cases these include Nos. 1-9 only and are largely based on the Haas editions. Other similarities are that neither conductor had much, if any, concert experience of the first three symphonies and that they subsequently made some finer recordings of the later symphonies. Some of the recordings by both these conductors are amongst the finest ever made but Wand’s complete cycle might well be passed over and PW is less enthusiastic about the Karajan cycle than JQ, having some reservations about the general clarity of the sound based on hearing the original LPs, and the Fourth and Ninth on CD.

There are other complete cycles by Chailly and Solti. Except for Chailly’s readings of Nos. 0 and 1, we have not heard these but reviews we have seen suggest that they are unlikely to be superior to those discussed above. 

One partial set which is historically important consists of Symphonies Nos. 4-9 minus the first movement of the Sixth in live performances given between 1942 and 1951 under the baton of Wilhelm Furtwängler (see review). This set is also available more cheaply but with inferior sound on Andromeda (see review). Furtwängler’s only studio recording of this composer dated from 1942 and consisted merely of the slow movement of the Seventh (see review). In general, these live performances with the Vienna (Nos. 4 & 8) or Berlin Philharmonic Orchestras (Nos. 5, 6, 7 & 9) are high-voltage Bruckner, both interventional and inspirational. Everyone interested in the composer should hear them but they are hardly mainstream recommendations.

Another partial set which demands attention – although it could hardly be more different from Furtwängler’s – is Celibidache’s Munich recordings of Nos. 3-9 (see review). As well as the fairly expensive 12 CD set, it is possible to obtain the individual works one way or another if one is prepared to download (from AmazonUK or Passionato) or order an ArkivMusic CD-R reprint. Celibidache was averse to making recordings but there are also extant readings of Nos. 7-9 from Stuttgart (see review), which are less extreme. If you want to experience Celibidache’s Bruckner it seems more logical to go for the “Full Monty” from Munich. The 104 minute Eighth is the broadest of them all and it is surprising how one can and does adapt to such stasis. Individual reactions will vary enormously but the playing, concentration and sound is magnificent. The Fourth is perhaps the most recommendable and the place to start but once you have, you’ll surely want to hear the rest. Applause before and after the music and audience noise between movements  is generally left in to enhance the live experience.

Downloading Bruckner 

It seems appropriate to add here a short section considering some issues about downloading Bruckner symphonies, based on PW’s experience to date. Beyond saying that he doesn’t believe that it should be a major deterrent, the controversial issue of sound quality will be ignored but, whatever one’s views on that, there are pros and cons. The length of Bruckner’s symphonies means that it can take a while a download one and at the higher end of mp3 bit rates (which seems to be the norm) up to about 150 Megabytes of storage space per symphony will be needed. On the plus side, pricing structures are generally favourable to long tracks. If you subscribe on the basis of x tracks per month, a Bruckner symphony is much more cost-effective than, say, a Schubert song. Indeed, for 35 tracks you can have a complete set. Outside such subscriptions, whilst most prices reflect track length this is not invariably the case. Thus there are quite a few Bruckner recordings that can be downloaded for 69 or 79 pence per track – say £3 per symphony. These include for example, some of Wand’s Berlin recordings on Amazon UK and, in particular, the Eighth, which is one of our highest recommendations.

Another point in favour of downloading is that it is potentially possible to extract a single work from a complete set, although this is often rather expensive. For example, some of Haitink’s first recordings have been issued on Philips Duos but not the Second. This is a fine performance that on disc can only be obtained as part of a set. If downloaded it costs £9.16 from Universal Classics and Jazz whereas the Duos, which contain two symphonies each, are priced at £9.99.

We should add the caveat that prices and availability may change but, generally, our expectation would be that such changes will be for the better. It is also worth mentioning that there are significant “freebies” around – the abruckner.com site which is mainly a [superb] discography offers a free download each month, usually of a historical recording. It also links to other sites offering free Bruckner downloads. At least at the time of writing, it was possible to follow the links from there to download a very decent Eighth conducted by Kent Nagano for free. There is also a four movement version of the Ninth freely available which is conducted by Hubert Soudant.

If there are recordings that one is only likely to want to hear once or storage space is an issue, streaming i.e. downloading without saving, is a useful option. Much the best site on the web for streaming of classical music is the Naxos Music Library and this has multiple versions of all Bruckner’s symphonies including all the Tintner and Paternostro recordings. 

Overall recommendations

Below we give one principal recommendation for each of the early symphonies (as far as No. 3) and two choices for the later works. It is worth reiterating here that we make no pretence to these choices being based on a complete review of available recordings and availability may change over time, so we have disregarded it as a factor in our choices. These choices merely represent what we believe to be the finest recorded performances we have yet heard; sound quality was not taken into account.

Symphony No. 00 Tintner (1998)

Symphony No. 0 Tintner (1996)

Symphony No. 1 Haitink (1972)

Symphony No. 2 Giulini (1974)

Symphony No. 3 Haitink (1988)

Symphony No. 4 Böhm (1973), Wand (1998)

Symphony No. 5 Horenstein (1971), Sinopoli (1999)

Symphony No. 6 Klemperer (1964), Haitink (2003)

Symphony No. 7 Wand (1999), Haitink (2007)

Symphony No. 8 Karajan (1988), Wand (2001)

Symphony No. 9 Walter (1959), Wand (1998)

Finally, with a nod to Nick Hornby’s book High Fidelity, our personal top five Bruckner symphony recordings, in order of preference, are:

Patrick Waller

Gunter Wand 8th (2001)
Carlo Maria Giulini 2nd (1974)
Giuseppe Sinopoli 5th (1999)
Bruno Walter 9th (1959)
Karl Böhm 4th (1973)

John Quinn

Günter Wand 8th (2001)
Karl Böhm 4th (1973)
Jascha Horenstein 5th (1971)
Günter Wand 4th (1998)
Herbert von Karajan 8th (1988)