Introduction: On the Usefulness of Archival Sound Recordings in Performance Practice Study

That the study of performance practice per se is important to the study of music and its creators and performers cannot be doubted. Further, an organized investigation into the historical aspects of performance practice--i.e., performance standards and options within a particular milieu during a set time period--can offer understanding to the nature of the music and its aesthetics in a manner ranging from merely noteworthy through useful to indispensable.

Certainly many factors have influenced changes in performance practice from the middle of the nineteenth century to the present date. In terms of orchestral performance, the most influential factors must include the following: the existence of fully-professional orchestras, standing bodies with set personnel and planned schedules of rehearsal, performance, broadcast and recording, thus affording the development of technique and ensemble; the advent of virtuoso perfectionist conductors, beginning (for the purposes of this study) with the much-beleaguered Berlioz and stretching through such éminences grises as Hans von Bülow, Artur Nikisch, Gustav Mahler, Felix Weingartner and Arturo Toscanini whose demands in orchestral performance practice and whose personalities shaped not merely the orchestras which they conducted but composers', performers', critics' and dilettantes' ideas of the possibilities of orchestral style and performance; the social and economic conditions which have allowed for the creation and maintenance of these professional orchestras and circumscribed the responsibilities and privileges of their members, while paradoxically limiting the power of the autocratic conductors from Mahlerian or Toscaninian rank, and further influenced the public and critical tastes toward performance practice and style; and the technological advances which have permitted the increased travel of conductors and orchestras, as well as the dissemination of their performances via radio and television broadcast, audio and video recording, and the associated communications procedures which allow as well for publication and dissemination of musical scores, criticism, reproductions of manuscripts, etc.

This last circumstance--that of the technical advancement as heavy influence upon performance practice--cannot be sufficiently stressed. It is tritely remarked that the advent of broadcast and of recordings has brought music closer to a larger audience than has ever existed previously in human history; the depth of this influence becomes almost unfathomable whenever it is casually remarked, for example, that a single televised performance of a particular opera has been seen (and may continue to be seen, by recording and repeated playing or broadcast) by more people than had ever seen that opera in live performance up to that time.

To what extent archival sound recordings may be taken as accurate reflections of prevailing performance practice during the period available for actual study (effectively, about the 1890s to the present day) can be a major point of contention. Admittedly commercial sound recordings from past eras of the recording technique have their technical limitations, and these archives' use as documents must be viewed as of limited use, since they can merely suggest the sound of the live performances, not accurately represent them.

It may therefore be seen that an analysis of performance practice through recordings has its distinct liabilities, for the aesthetic values of what goes on in a recording studio may differ widely from what might occur in a concert situation. It is suggested by the present author that the researcher select with great care the nature of the repertoire in which archival recording performance practice might best be applied. Should one apply it to a small work or to a large one (whatever merits this terminology might have on an aesthetic basis); to a Classical or pre-Classical work or to one recent enough for a performing "tradition" (perhaps even based on indications and/or performances by the composer); to an opus performed very frequently or hardly at all?

Let us examine the case of a frequently-performed work, such as a Chopin Waltz. Obviously a great number of recorded performances would be available, beginning with many from the acoustical recording era, through the many electrics indexed in The World's Encyclopaedia of Recorded Music and continuing through the eras of the long-playing record, stereo, quadraphonics, direct mastering, digital-analog hybrids, and the Compact Digital Disc even now being introduced commercially.1 One might also include private label issues, radio and television airchecks, and such live performances as might be heard in a major musical center in the course of a year. The scope of such a survey would be a broad one indeed, perhaps too broad for the drawing of any but the most general of conclusions. Further, the performance practice of Chopin piano works is an intensely personal one, and noted pianists will often change their views on the performance work depending on such unmeasurable parameters as "inspiration of the moment" and the attention-span of the audience. Furthermore, the early and middle years of the 20th Century saw a great many performers with markedly individual styles, perhaps because of the growing personality cult of the solo pianist as a competitive performing (and recording!) artist: Ignacy Jan Paderewski, Sergei Rachmaninoff, Josef Hofmann, Ossip Gabrilowitsch, Josef Lhevinne, Leopold Godowsky, Mischa Levitsky, Harold Bauer, Egon Petri, William Kapell, Dinu Lipatti, Vladimir Horowitz and Arthur Rubinstein--all artists of markedly differing resources and temperaments.

Furthermore, the production of a recording of a piano work is (and has always been) a relatively simple one, from the earliest days of the recording technique, dating back at least to 1889, when Brahms made his one cylinder, of a portion of his Hungarian Dance #1 in G Minor. It is the work of but a moment of decision for a pianist, perhaps in the studio for some other purpose, or perhaps for none at all, to decide to record a Chopin Waltz which happens to be in his repertoire; and if he is aware that a rival has been playing a particular work, he might decide on a moment's thought to record that one instead of a planned item. Too, a Chopin Nocturne which in the pianist's typical concert performance might require six and a half minutes for performance, on account of the time restrictions afforded by the early electric process might be "sped up" to take no longer than the four and a half minute side length of that medium. Additionally, a pianist may think ill enough of the very process of recording that his interest may lie less in the notes of the recording than the numerals on his payment check; or a performance which has not been approved by the performer for artistic or other considerations may still be released by the recording company. These particular recordings might not then be representative of that artist's style, thus rendering questionable the basis of this sort of research on determining what actually may have happened in concert performances.

Similar problems might befall a study of vocal performance practice, since singing is if anything an even more personal form of performance than pianism; Enrico Caruso's records might display his voice and general stylistic approach, but surely cannot pretend to show him how he was "in the hall," and Francesco Tamagno's and Claudia Muzio's acoustics, made as they were in these singers' later years, similarly cannot be held fully representative of their art.

Symphonic performance practice in the 20th Century has undergone a great variety of changes in tradition and style, chiefly and most importantly at the hands of three of the most influential European conductors of the late 19th and early 20th Centuries: the Bohemian-Austrian Gustav Mahler, the Dalmatian-Swiss Felix Weingartner, and the Italian Arturo Toscanini. The Hungarian Artur Nikisch, as chief conductor of orchestras in Berlin, Leipzig, and London, also had a powerful impact upon orchestral performance quality and style in Europe, as indeed can be witnessed in his recording of the C Minor Symphony of Beethoven, made with the Berlin Philharmonic Orchestra in 1913, and therefore the first recording of a complete symphony, but the purifying tendencies of Mahler, Weingartner and Toscanini were to exemplify the 20th Century drive toward perfection of realization of the composer's intentions, la musica comè scritta, freed from the stultifying (though occasionally imaginative and creative) performing "traditions" which grew up around many works in the 19th Century. Mahler's drive as a conductor of opera and symphonic works is well-documented in texts and personal reminiscences--his famous remark of "Tradition ist Schlamperei!" is frequently cited as an indication of his attitudes toward the pervasive traditions of late 19th Century performance-- though he unfortunately left no actual, fully-documented sound recordings as a conductor (some piano rolls exist, and there has been speculation that Mahler may have participated in a few early acoustic recordings as piano-accompanist for some vocal recordings, but conclusive evidence is lacking).

Weingartner, whose career took him from Vienna to Boston, Russia, and Japan, made a surprising number of recordings between 1910 and 1940; he was the first conductor to record all the Beethoven symphonies, some more than once, and the first to record all the Brahms symphonies; his writings on these and later Romantic symphonies are still cited for their insight into performance practice (many of which suggestions Weingartner did not follow on some of his own recordings!), and are still standard works to consult for suggested wind doublings in this literature.2,3

Toscanini's own career stretched from 1886, the year of his surprise debut conducting Aïda in Buenos Aires, to 1954, the year of his retirement from the helm of the NBC Symphony Orchestra. During his sixty-eight very active years as a conductor his was a strong voice in favor of the integrity of the composer's directions (though he was not averse to careful rewriting, judicious cuts, and some slight alterations in orchestration; Debussy, for one, consented to Toscanini's emendations to La mer, which clarified the musical texture).4 Toscanini's art has now fallen largely into disfavor as critics who have heard precious few of the recordings, spanning the years 1920-1954, disdain him for tempos which are "always too fast" (though a more studious and comprehensive approach will show that this was certainly not true), and blame the conductor for the primitiveness of the recording science in those years before stereo, quadraphonics, digital tape and laser-read Compact Digital Discs.

Despite some difficulties, chiefly based in the varying quality of the recordings and in the selections of repertoire and personnel, it would appear that the examination of performance practice of a standard orchestral work might offer useful opportunities for observations on 20th Century performing practice. Yet here these very difficulties must be noted. It is all too easy for a recording session to be set up on short notice and the resultant recording may therefore lack the approach, in true performance values, of the conductor's conception and the orchestra's actual playing style as might be encountered in the concert hall. It has in fact been known to happen that a Brahms symphony, the C Minor, has been recorded in the session time left over after a successfully-completed recording of a Mahler symphony, as was actually done by James Levine and the Chicago Symphony Orchestra in 1975. Indeed, in these days when professional orchestras maintain their entire basic repertoire up to playing standard at all times, it is not unknown for an ensemble to play the Mahler First Symphony on twenty minutes' notice, as the present author witnessed at a concert by the Los Angeles Philharmonic Orchestra under Gerhard Samuel in 1972. Whether such "spur of the moment" performances reflect accurately a specific method of approach and broad conception of an orchestral work is moot.

Similar problems might undermine a study of operatic performances since the repertory nature of most of the world's major opera houses and the "equalizing" approach of the most eminent of the operatic stars cannot ensure that a given recording of an opera is especially representative of a specific stylistic approach.

One is left, then, with the following suggested parameters for the selection of a composition upon which some basic performance practice analysis may begin to be performed:

  1. It should be a work of the standard literature, and yet not typically, commonly, frequently nor easily mounted;
  2. It must require special performance approaches, including but not limited to considerations of size of ensembles, placement of performers, allotment of rehearsal and performance time and space, and so forth;
  3. It should be a work of such monumental size, quality, gesture and effect that its performance may be noted in journals;
  4. It must be a composition of such significance and interest that it will be regarded as "worth the trouble," artistic and commercial, of recording, and a composition requiring serious artistic and stylistic decisions.

One general category of musical endeavor comes to mind immediately: that of extremely large-scale choral-and-orchestral works. Comparison of performances and their tempi in Händel oratorios might be fodder for the mill of such analysis, as has been shown by Riedel.5 Yet the performance choices inherent in mounting such a work as, say, Messiah are so numerous that a comparison of performance practices would need to begin with many pages of tables indicating which editions have been consulted, selection of alternate arias and alternate versions of arias and choruses, size of chorus, etc.--surely a monumental task for any researcher, Therefore we may consider adding one further parameter to our above list:

  1. The work should have a standard, basic performing edition, or at least be relatively free from exceedingly wide performance choices, in order that variant performances might be judged on essentially equal bases.

Several large-scale opera present themselves for consideration. Gustav Mahler's Symphony in E Flat might seem ideal for such a survey were it not for the fact that its size and complexity act to defeat it; it is so enormous a work that it can typically be given only under festival circumstances. Arnold Schönberg's Gurre-Lieder would seem a possibility, but its inconsistency of compositional and orchestrational style, its dissimilarity to the chief body of Schönberg's life work and the essential lack of any form of Schönberg "tradition" mitigate against it; also to be considered is the extremely high rental cost of scores and parts, which reduces the number of possible performances. Havergal Brian's Gothic Symphony, a work of colossal scope and tremendous invention and imagination, would be a perfect vehicle for a survey were it not so completely far out of the performance mainstream; this 1928 work has only enjoyed four performances, the earliest of those being in 1964.

If the emphasis is altered to take into account works of a liturgical or religious nature, there are several candidates: Beethoven's Missa Solemnis, Brahms' Ein Deutsches Requiem, Dvorak's Stabat Mater or Requiem, Verdi's Manzoni Requiem, and Elgar's sacred oratorio The Dream of Gerontius. Yet in terms of the combination of popularity with overweening size, and sheer difficulty from the standpoint of material and physical plant arrangement, the Grande Messe des morts of Hector Berlioz stands alone as a monument both to 19th Century largesse and the grandiose vision of its composer. Further, as much with Berlioz as with any of the other mentioned composers, a palpable performance tradition exists by which one might gather that performances of the Grande Messe des morts are prepared with an eye to the integration of this work within that stated style. The principal pioneering Berlioz specialists in this century, Felix Weingartner, Sir Hamilton Harty, and Sir Thomas Beecham, were all known far their performances of a wide variety of Berlioz works including the Grande Messe des morts; the possible influence of Berlioz specialists upon Berlioz performance practice in general will be examined in a later chapter.

With the selection, therefore, of Berlioz' Grande Messe des morts as the subject of this thesis, several points of agreement with the stated parameters may now be observed:

  1. The Grande Messe des morts --hereinafter referred to as GMdm--is a standard work, particularly insofar as it is the largest non-operatic/dramatic work of a commonly-performed composer;
  2. Its performance demands are such that special efforts must be made to ensure the availability of brass bands, numerous timpani, suitably large choral and orchestral ensembles, rehearsal and performance space, and scores and parts, which last are fortunately easily available;
  3. It is a work of such size and interest that its performances do not typically go unnoticed;
  4. It has been found to be both commercially viable and artistically successful, both in concert presentation and in recording, with the result that several notable renditions have been produced, many of which are still available as recordings;
  5. There are few exceptional difficulties with editions (although some variants exist, as will be noted later), and thus performances may be easily compared on a common background.

Copyright © 1983, 1995 by Matthew B. Tepper

Table of Contents


Chapter 1


Back to The Hector Berlioz Page

Back to Matthew B. Tepper's Home Page