Current Research in Jazz
Home > V16 > Duke Ellington's Compositions: A Revised Survey of His Output as a Composer

Duke Ellington’s Compositions: A Revised Survey of His Output as a Composer

Jørgen Mathiasen

After World War I, black Americans began to assert themselves as central culture-creating personalities in the 20th century. Among these were the creators of jazz, including Duke Ellington (1899–1974), who was also an exponent of the cultural current called the Harlem Renaissance. In addition to being the leader of his own orchestra for around fifty years, he wrote a large number of pieces and works. Determining the magnitude of this number is the subject of this article.

History of Estimates

Broad-based research into Ellington’s output as a composer began with Erik Wiedemann’s 1986 articles in the Danish Musik og Forskning. In these he mentioned that there were estimates of the scope from Jan Bruér (2,000) and André Hodeir (6,000), respectively. [1] This was a significant difference between two connoisseurs of Ellington’s music. Based on the research Wiedemann had done since 1984, he gave a new estimate, and a few others followed before the publication of his second and final paper on the topic in 1991. These were the following estimates:

Author(s)ScopeYear
Erik Wiedemann [2] 1,200–1,3001986
André Hodeir and Gunther Schuller [3] 2,0001988
Ken Rattenbury [4] 1,0121990
Erik Wiedemann [5] 1,5001991

In his articles, Wiedemann cited the most important sources in the Ellington literature for a list of works, but subsequent to the 1991 publication of his article in the Annual Review of Jazz Studies, the picture changed. Klaus Stratemann’s filmography appeared in 1992, and new studies of parts of Ellington’s and Strayhorn’s music by John Franceschina and Walter van de Leur were published in connection with the hundredth anniversary of Ellington’s birth in 1999. At the same time, a team of European discographers with Luciano Massagli and Giovanni M.Volonté as editors published a new edition of the discography: The New DESOR, while the Smithsonian Institution published an inventory of its holdings of Ellington sheet music. There was thus a new basis for making a work list, and at a 2004 conference for jazz historians, I estimated, on the basis of the literature on Duke Ellington and Billy Strayhorn, the now known scope of the output to be around 1,700 compositions. In the meantime, there has been no reason to change this estimate, but not least the fluctuating assessments of Ellington’s music output is a reason to look more closely at how a list of works for his compositions can be provided or how a data model for a database that produces a work list and a chronology of works can look like. Erik Wiedemann developed some thoughts on this, and the Italian discography contains some others. In both cases, they are dictated by some circumstances of Ellington’s working habits, which undoubtedly also played a role in the fluctuating assessments.

Firstly, there is the question of crediting, that is, which pieces Ellington composed; secondly, it was Ellington’s habit to give his compositions several titles and to work the compositions into different versions — a problem area which has created much unevenness in the Ellington literature; and thirdly, it is the perception of concepts such as derivatives and contrafacts or adaptations of original pieces. These are conditions to which studies of Ellington’s total output as a composer must adjust, and which ultimately determine the scope of the output. There is enough elasticity in these concepts for one to conclude that music historians can arrive at only an approximate, yet qualified assessment of the scope and structure of the overall work.

As a model example for several problem areas in Ellington’s way of working, “C Blues” is suitable:

CompositionTitlesCredits
The first eight intervals
from the tonic:
5 5 5 5 5 5 5 1
C Blues (1941)
C-Jam Blues (1942)composer: Duke Ellington
Jam Session
Jump Blues
Duke’s Place (1957)composer: Duke Ellington
lyricists: Ruth Roberts, Bill Katz, Robert Thiele

A list of works must be composed of information from (many) different sources, and each column in the table has, in addition to a transversal problem (horizontal) also its own problem (vertical), and I will now look at both in more detail.

Credits

The “C Blues” theme consists of two notes (fifth and first scale degrees), and it has been given five different titles. In Music Is My Mistress, Ellington is credited as the composer of “C Jam Blues”, but when the composition had a lyric added, its name changed to “Duke’s Place”, and Ruth Roberts, Bill Katz, and Robert Thiele were added as lyricists. The Italian discography uses this second credit for all the titles of the composition, and it is a principle of the discography. Instead of the credit for each individual title, the complete credit for the composition is given, and thereby the credit also applies to the alternative titles, which otherwise have no credit information in the source material. For the above composition, it applies to “Jam Session” and “Jump Blues”.

With credit information for more than 1,000 titles, Music Is My Mistress is one of the most important sources for the origin of compositions. Other contributions have come from Wiedemann’s studies in the Copyright Office and from documents from Ellington’s publisher Tempo Music, from Stratemann’s studies of film, Franceschina’s of Ellington’s music for the theatre, and the Italian discography. There are smaller contributions from other sources, and thus approximately 1,700 compositions are provided with credit information. The inventory for the sheet music archive at the Smithsonian Institution contains only limited credit information, and it must be assumed that this material contains a number of uncredited compositions by Ellington.

Of the reported compositions, more than 900 have Duke Ellington credited exclusively as composer, while in a smaller group of around one hundred compositions he is also credited as lyricist. The remainder are compositions where Ellington is credited with others. He is included in around 200 credit combinations in the source material, of which the credit of Billy Strayhorn as co-composer is the most important set. Other of the credits sets, notably those with Irving Mills, indicate that Ellington did not have artistic but rather financial reasons for his copyright registration. The dividing line between Ellington as composer and as orchestra leader was not as sharp as he himself claimed.

Especially in the part that concerns Billy Strayhorn’s work for Ellington, there are examples of the complementary credit being in fact a conflicting credit, and therefore most of the objections to the credits also come from the Strayhorn literature. In some cases Ellington is completely written out of the credits, while in a few other cases it is Strayhorn who is crossed out. Added to this is the discussion of musical contributions from other orchestra members to Ellington’s compositions. Some, such as Ben Webster, did not accept his way of handling credits and refrained from contributing to the orchestra’s music once they became familiar with it. Others, such as Johnny Hodges, despite their periodic reluctance, returned to the orchestra and Ellington.

The crediting of Ellington as an author is, of course, a first decisive criterion for a list of works, but the assessments of his total output are affected by how one will relate to the crediting information in cases where there is conflicting information, and it must be possible for the crediting entity of a work list database to accommodate together with the complementary credits, which, in terms of technical data, do not differ much from the conflicting ones.

Titles of Pieces

Which corpus of titles for compositions one wants to include in an Ellington work list depends, among other things, on how one wants to relate to peculiarities of American orthography, the facultative use of grammatical articles in English, spelling errors, unauthorized titles, and mistakes by record companies. In the corpus that is the basis here, titles with and without articles are reduced to one title, the rest have been left out, and thus the title corpus is limited to around 2,900 titles. Without this reduction it would involve approximately 700 extra titles.

There are a number of cases such as Paris Blues, My People, or The River, where a composition title has also been used as a title of a work, and such sets must be kept separate by a numbering and a classification. Each case also counts as two units in the calculation of the scope. In approximately ninety cases, a composition falls into the work category, either because it is divided into single compositions or movements, or because it is of longer duration. Despite the fact that fourteen pieces are titled “Blues”, and that Ellington himself numbered five compositions titled “Dance” and seven others titled “Freedom”, in a compositional practice that spanned more than fifty years, there is strikingly little reuse of titles in the overall corpus. It is difficult to imagine that this would be a coincidence. It seems rather as if some (Ellington) had both an overview and a memory for titles, and this is only reinforced by the fact that some titles such as “Concerto for Cootie” versus “Cootie’s Concerto” have clearly been separated by an active linguistic effort. The collected material shows that Ellington produced titles on assembly tape, and that the tape got out of control on a few occasions (“Midnight Indigo”).

The authors of the Ellington literature and the archivists at the Smithsonian Institution, where Ellington’s music manuscripts are kept, have made great efforts to attribute the titles to the compositions, but have not been able to avoid creating contradictions — in some cases almost labyrinthine contradictions — which is why no Ellington work list can avoid deviating from one source or the other and in the complex cases must be drawing on an auditory comparison. The incompatible representations of the compositions’ title sets are supplemented by disagreement about how to deal with cases where thematic material has been partially reused for a new (derivative) composition. Alongside the challenges they pose for a list of works, these also affect the assessments of the scope of work.

In the schematic overview of “C Blues”, I have omitted giving the composition a title. This is, among other things, to draw attention to the fact that it is necessary to choose a title. All authors of literature about Ellington have faced that task, and in some cases the choice has fallen on the best-known title, which in the chart is “C Jam Blues”. But the rule is misleading, and it also becomes meaningless in cases where the popularity principle cannot be applied. In the model, I have used Erik Wiedemann’s interval string, which suits a computer perfectly, though on the other hand, a human being cannot distinguish 1,700 interval strings. One cannot avoid choosing a title that must be used as a reference point and sorting element for a list of works. A close choice would be the earliest, or in other words to make dating the criterion.

Dating

Up to this point, only fragments of a work chronology for Ellington have been made. Such can be found in texts by himself, Mercer Ellington, and Stanley Dance, in the discographies, and in the aforementioned studies of film recordings and stage music. But the dating of Ellington’s work as a composer in the source material is of such a fragmentary nature that one cannot use the date of composition as the sole dating category, and already early in his endeavour Erik Wiedemann concluded that one must use a different dating than the one normally aimed for in a list of works. He repeated this in 1991: “In my project, the place of the individual composition is determined by its earliest occurrence, be it a recording, a known public performance, registration with the CO [Copyright Office] or ASCAP, or publication as sheet music.” [6] To the categories Wiedemann mentions, one must add premiere dates for stage works with Ellington music, but in any case the publication of a composition is a point in Ellington’s chronology which has a consequence for the overall dating: the composition must necessarily have taken place before this point.

Since only a small portion of the compositions are dated in the composition category, a work chronology must use supplementary dating categories. Here, two in particular take a leading role: Music Is My Mistress has a long list of compositions dated with the copyright year, and the discographies have dates for recording in recording studios and in concert venues. Ellington’s stage music is only included to a limited extent in these works, but John Franceschina was careful with this aspect in his study, and thus the theatre music was dated with the time of publication. In some cases they did not take place until several years after the music was written, but they remain points in Ellington’s overall chronology.

The dating starts with the first composition in 1913 and ends in 2000: A smaller group of compositions has been made public only after Ellington’s death. There are just over 3,500 dates in the database, which concern 1,700 compositions and 2,000 titles. (In about 1,500 cases, a title has a date in more than one category.)

As mentioned, the source material for a dating of the composition is fragmentary, and this category makes up only about 10% of all datings. In almost all cases the dating contains a year, which should be read as the interval between January 1 and December 31. The same approximate dating applies to the copyright category, which accounts for about 33% of all dating. For the other dates, the majority specifies a specific date, while a minority specifies a time interval.

The first preserved live recording with Ellington was made in 1937 (aptly titled Swing Session), and live recordings were made at intervals thereafter, not least of which the live recordings of the first performances of Black, Brown, and Beige in January 1943 must be mentioned. In the period after 1945, the frequency increased, so that from the Newport Jazz Festival in 1956 until Ellington’s death in 1974, live recordings appear regularly. In a number of cases — not least (A Tone Parallel to) Harlem from 1949, recordings of piano versions of Ellington compositions constitute the earliest publications.

Klaus Stratemann’s collection of information on footage (mainly films) with Duke Ellington is a significant addition to a list of works, but Stratemann’s text is not just a filmography. With it, he also published comprehensive information on Ellington’s professional activities and a travel calendar, and this information can, in some cases, document the time of a composition’s creation or its publication. The dates must be sorted numerically, and since an entire year is smaller than a date that also contains month and day, the entire year comes first. In most cases, this is probably not the historical truth. The workings of Ellington’s organization make it more likely that a copyright notice was only made after a recording.

A large number of titles have no dating at all. It must be assumed that they have already been created in connection with the composition of the music, i.e., that they go back to the earliest sketches for the composition.

Entity of Composition

The earliest dated title for a composition appears as a point of reference in the work list, insofar as the sources allow a choice. Where this is not possible in individual cases, the choice of index title is arbitrary.

While a title is functional for the human reader, it is not acceptable as a key for a database because it opens up confusion and because it does not work in the face of the numerous orthographic variations anyway. Therefore, the database must contain an artificial key for compositions. It stands for the depth structure in the list of works, while the titles are surface representations of Ellington’s compositions, and not surprisingly, the composition key is also the most supporting structure in the database below the list of works.

The above mentioned entites are essential to a data model of Ellington’s output as a composer. As of this writing, the data of the sources representing his compositions have been entered into a database solution, which consists of a handful of databases with many thousand records, including records keeping track of the sources to each composition. The solution also produces various reports that, when combined, form a work list and a chronology of Ellington’s output and support the ongoing proofreading of the records.

Bibliography

References

[1] Erik Wiedemann, “Duke Ellington som komponist — en indkredsning,” Musik og Forskning 11 (1986): 81.

[2] Ibid.

[3] André Hodeir and Gunther Schuller, “Ellington, Duke [Edward Kennedy].” In The New Grove Dictionary of Jazz, Edited by Barry Kernfeld. (New York: St. Martin’s Press, 1994): 332.

[4] Ken Rattenbury, Duke Ellington, Jazz Composer. (New Haven: Yale University Press, 1990): 2.

[5] Erik Wiedemann, “Duke Ellington: The Composer.” Annual Review of Jazz Studies 5 (1991): 46.

[6] Erik Wiedemann, “Duke Ellington: The Composer.” Annual Review of Jazz Studies 5 (1991): 41.

Author Information: 
Jørgen Mathiasen was born in Denmark in 1956 and has lived in Berlin, Germany since 2003. He received MA degree from the University of Copenhagen with studies in musicology, information science for the humanities, and Germanic philology. He has been involved in music as a teacher and lecturer and as an editor for jazz radio. His work on Duke Ellington includes an index to Music Is My Mistress (1995); “Duke Ellington’s Oeuvre. To the Discussion of the Size and Its Structure” in DEMS Bulletin (2002); and a contribution to the August 2004 conference of Nordic jazz researchers in Odense entitled “Ellington’s Production as a Composer. A Survey of a Selection of Sources to His Entire Production and a Methodological Discussion.”

Abstract: 
Duke Ellington was an extremely prolific composer with a long career, but estimates vary widely as to just how much music he wrote. In this article, the challenges to determining an accurate number are identified, and a logical and systematic plan is proposed to approach the solving of this puzzle.

Keywords:
Duke Ellington, compositions, jazz

How to cite this article:

For further information, please contact:

Valid XHTML 1.0 Transitional Valid CSS! Creative Commons Attribution 3.0 License Open Access Dublin Core Metadata Initiative

Except where otherwise noted, content on this site is licensed under a Creative Commons Attribution 3.0 License

This page last updated December 31, 2024, 14:39