On 1 January 1917 the periodical Nyugat* printed Menyhért Lengyel’s story, titled Pantomimegrotesque, which became later the libretto for Béla Bartók’s music for the pantomime The Miraculous Mandarin. The story interested the composer and he agreed to write the music for its stage presentation.
Béla Bartók completed the sketches for the composition by about May 1919, but he laid them aside and only produced the orchestra score in 1924. At the same time he prepared a piano version, for one piano four hands. This version followed the second sketch, which also was written on four staves, and both piano and full score versions incorporated the same modifications wherein they departed somewhat from the first sketch.
A handwritten copy of the complete orchestra score was made by the composer and his first wife and his second wife and delivered with the four-hand piano version to the publisher, Universal Edition, who printed the latter in 1925. First performance of the stage work with orchestra had to wait until 27 November 1926, in Cologne, conducted by Jenő Szenkár, but portions of the four-hand piano version were played by the composer and pianist György Kósa in a Radio Budapest broadcast on 8 April 1926.
The performance in Cologne was a disaster, as the audience could not tolerate the realistic elements in the story. Although a more successful performance took place in Prague on 19 February 1927, (not heard by Bartók), with the banning of the complete work by the Hungarian authorities in the same month prospects for further stage performances appeared dim, leading the composer to suggest to his publisher that promotion be concentrated on the concert version that was already contemplated as early as 12 September 1926. A final formula for such a version was sent to the publisher on 3 February 1927 and the concert version became published by the end of that year.
For possible stage performances where the length of the work may be a problem, the composer directed an optional cut, the section from  – , that would eliminate the second half of the chase of the girl by the mandarin (letter to Universal Edition, 3 December 1926). This is the only cut in the score that the composer authorized.
Performance of the work by the Budapest Opera House was contemplated as soon as the orchestration was completed. In 1925 the full score manuscript was copied by a professional copyist at the Opera House. A two-hand piano reduction was prepared for rehearsal purposes, hand copied also by the composer and his first wife Márta and his second wife Ditta, and presumed to have been available for performances first planned in 1927, but not permitted by the authorities.
A second attempt to perform the work in Budapest was planned for 25 March 1931, the composer’s 50th birthday. In order to reduce the work’s features that were offensive to the censors’ delicate sense, the story was somewhat modified. As László Márkus, the producer, explained at a closed dress rehearsal a day before the planned performance: ‘The story was reworked by Melchior Lengyel. The action does not take place in a room, as in the original libretto, but out in the open. The surroundings are reminiscent of an empty lot in the suburb of a big city. The emphasis is on the suffering of the girl who is reluctant to participate in the robberies. Regarding the mandarin, he is an absurd spider-like vision, the nightmare of the greedy bandits. The Miraculous Mandarin became more a Grand-Guignol-like play in Lengyel's new modified version. This is further enhanced by the unreal environment that gives the work a dream-like character.’**
Despite the new setting, however, the censors deemed it their duty to protect the public from viewing the performance the following day. However, as the project progressed towards the dress rehearsal, it appears that not only the stage action but also the music had to be modified, apparently with Bartók’s consent. The copied score of the Budapest Opera House shows pencil deletions, mostly in the second half of the work, that seem to follow some of the cuts which the composer marked in his printed copy of the four-hand piano version.
The rationale for the cuts could be explained by the revision of the story: As the scene is outdoors, there is neither bed nor pillows with which the mandarin was to be suffocated, so the corresponding music was eliminated.
Likewise, the mandarin could not be hung from a chandelier outdoors and the chandelier could not crash to the floor. On the other hand, the scenes with the girl dancing and being chased by the mandarin also work outdoors and no cuts appear in those parts of the music. Without a complete documentation it can be only speculated that the remaining cuts in the music were likewise dictated by the altered libretto. Altogether 42% bars of music are marked to be cut in the Budapest Opera House’s copy and it seems probable that the cuts were only made for the 1931 planned performance.
Prior to this planned performance the composer prepared a new ending to the work to replace the earlier version. This is the only modification in the score which he apparently wished to be permanent, as he had requested the publisher to incorporate it into their copies of the score and parts and then return the manuscript (letter to Universal Edition, 29 January 1936). The original ending is reproduced in the present edition, as an appendix to the study score.
It is not known which score was to have been used in Budapest when the work was next planned to be performed (in 1941, also aborted), or when the performance finally took place on 9 December 1945. It is only known that the copy of the score, prepared by the composer and his family, was also found decades later in the library of the Budapest Opera House, with no cuts marked.
The publisher finally printed the full score of the complete Miraculous Mandarin in 1955. This made use of the plates for the previously published concert version, with the remaining pages newly copied and engraved. It appears that the source used in preparing this edition was not the full score manuscript. From those parts of the score which were not part of the concert version, 30 bars of music are missing in this printed edition. Most of these missing bars correspond to certain bars marked for omission in the Budapest Opera House’s copy, presumably for the 1931 aborted Budapest performance of the work. A small size study score was also published in 1955; a newly engraved score but with the same contents as the large one. It can be surmised that the publisher sought to obtain a copy of the score from Budapest for the 1955 printing. It seems that it received a new handwritten copy from Budapest of what purported to be a complete score, rather than the one prepared by the composer’s family or the copy made by the copyist of the Budapest Opera House. In good faith the editor decided to print the last third of the work (the part subsequent to the ending of the concert version) as it appeared in this copy. Apparently the copyist in Budapest followed the full score that had been marked with cuts for the 1931 performance and did not include the marked bars in his copy.
The four-hand piano version was first reprinted in 1952. This reprint contained the complete work with the new ending which the composer had prepared prior to the planned 1931 performance. When the full score was printed in 1955, the four-hand piano version was reprinted again, changed by deleting sections so as to match the new, inadvertently incomplete full score. By mistake, however, only 28 of the 30 bars were deleted from the piano version.
Aside from the fact that some of the cut parts correspond to specific stage action, it is noted that Béla Bartók had carefully worked on some of those segments of the music that were later cut in the printed edition of 1955. In a precise list of alterations, ‘Änderungen im Mandarin’, sent to the publisher (letter to Universal Edition, 3 December 1926), he made modifications, fine-tuning the dynamics or other details. In a communication he lamented the impossibility of having the complete work performed (letter to Universal Edition, 3 February 1927): ‘Because, in my opinion, this is the best orchestra work I have written so far and it would be really a pity that it be left buried for years.’ It would be unlikely that he would have, soon afterwards, decided that some 30 bars of that well-written music should be eliminated forever. It is likewise improbable that the librettist would have wanted to permanently omit the corresponding details from his creation. If the cuts were approved by the librettist and composer, they doubtless did so only in the interest of enabling a performance to take place at that time. It is concluded that the deleted 30 bars of music need to be restored in this newly engraved corrected edition of the full score. A documentation can be found in the appendix.
The original story of the Mandarin is reproduced here in its entirety, translated from the Hungarian as it appeared in the 1 January 1917 issue of Nyugat. What the composer wrote in the score is of necessity a skeleton of the story, suitable to indicate when events on the stage are to occur, but all details could not be included. For instance, at  in the score the stage direction text reads: ‘The three tramps finally leap out from their hiding place, seize the old rake and throw him out.’ In the complete story this reads (second last paragraph of chapter two): ‘... the three tramps jump forward, – attack the old cavalier, – they form a line from the table to the door – they pass the old man from one to another until he flies out the door, rolls down the steps, they throw his top hat after him, – it can be heard as the hard hat falls down knocking on each step one by one.’ In the score at  the tramps grab the old man, pass him from one to another in the first three bars; at the end of the third bar (grace notes) the old man flies out the door, begins to roll down the steps at +4 and his hat bounces down after him about 3 bars later (descending scale, staccato). The one major abridgement in the libretto for the stage version is that there are only three attempts to kill the mandarin; in the original story there are four (including an old pistol after the rusty knife).
This edition contains both the complete score and the concert presentation. The formula for deriving the concert version from the complete score was provided by the composer on two occasions. Soon after completing the score, he advised the publisher (postcards, 6 and 12 September 1926) that for concert presentation a short version of the score should be performed, that would have consisted of two segments: from  to , then from 6 bars before  to 2 bars after , followed by a special closing section written only for this version. The composer advised the publisher that the concert version should not be called a Suite, but either Music from ... or Two Scenes from ‘The Miraculous Mandarin’. The definition is probably in view of the nature of the stage work, a pantomime, rather than a ballet or a series of dances. After the adverse reception of the work at its first performance, he suggested to Universal Edition concentrating on the concert version, since the likelihood of stage performances in Germany seemed remote; and he enlarged the concert version (letter of 3 February 1927). Accordingly, the concert version is to consist of the following: ‘From the beginning up to the 2nd bar (inclusive) after , with the following two cuts: from the 3rd bar before  to the 5th bar (inclusive) after , and from  to the 2nd bar (inclusive) before . Following the 2nd bar after  come the 11 closing bars composed earlier.’ The previously suggested title Two Scenes from ... became obsolete, leaving Music from ... as the only possible title.
In a communication subsequent to the printing of the concert version score in 1927, Béla Bartók asked the publisher to make an addition (letter dated 16 October 1928; also found hand-written by the composer at the intended position in his printed copy of the score):
‘After the table of contents of the full score of the Mandarin music the following should appear (properly corrected by you):
The music adopted for concert presentation contains approximately the first half of the music of the pantomime, virtually without abridgement, and it comprises the following sections attacca:
I. Introduction (street noises); the command of the hoodlums directed at the girl.
II. The girl’s first inviting gestures (clarinet solo), in response to which the old gentleman appears, who gets thrown out in the end by the hoodlums.
III. The girl’s second inviting gestures, upon which appears the young lad who is also thrown out.
IV. The girl’s third inviting gestures; the mandarin appears (tutti ff).
V. The girl’s seductive dance before the mandarin (at first slow, then increasingly faster Waltz).
VI. The mandarin catches up with the girl after an ever wilder chase.
It is requested that this description be printed in the program.’
In preparing this edition all available sources have been considered. These include, in addition to those already mentioned, a copy of the concert version, printed in 1927, with some correction marks in the composer’s hand. The entire score has been newly engraved on the basis of the manuscripts of the full score, the new ending, the four-hand piano reduction and the composer’s list of alterations, sent to Universal Edition in 1926.
I am grateful for the work of Nelson Dellamaggiore whose examination of all manuscript sources and their comparison with previous printings made this revised edition possible. All problematic details have been analyzed together with editor Heinz Stolba of Universal Edition. We also thank László Somfai, director, Bartók Archívum, Budapest, for his suggestions and providing copies of source materials.
Homosassa, Florida, August 1999
* ‘West’, a prestigious Hungarian literary periodical
** Az Est (‘The Evening’, a Budapest daily newspaper), 17 March 1931; quoted in: Ferenc Bónis, A csodálatos mandarin (‘The Miraculous Mandarin’), published by: Magyar Biblofil Társaságì Tevan Kiadó, Budapest 1990