Friday, January 17, 2014

James P. Johnson's "Carolina Shout" in Performance: a Recording and Paper.

Here's a paper and video of a performance I submitted for one of my classes at Columbia last semester. The class was titled "Post-1965 Jazz," and focussed on critical responses to jazz in America.

The paper outlines the argument behind, and gives an analysis of, my performance of James P. Johnson's Carolina Shout. I cite Australian musicians Scott Tinkler and Allan Browne, American musicians Anthony Braxton, Wynton Marsalis, and Ethan Iverson as well as writers/thinkers such as Jacques Derrida, Paul Lopes, Lydia Goehr to critique neo-classical modes of jazz performance.

I welcome any productive responses to this paper: please leave them in the comments below.

First: the video, then: the paper.

Carolina Shout in Performance


This term paper will discuss my solo piano performance of James P. Johnson’s most famous composition Carolina Shout.

My performance was conceived of to demonstrate what Paul Lopes describes as the “traversing of boundaries” around jazz practice. “Traversing” may evince those boundaries’ composition when conceived of along the lines of modernity and authenticity (Lopes 2005). Typically, the “modern” element of jazz is conceived of in a teleological vein that comes to the fore in be-bop and reaches a climax in the “new-thing” movement of Ornette Coleman et al. (Lopes 2002), and “authenticity” is used as a legitimising and canonising concept for modern performance as well as earlier jazz “genres.” I am particularly interested in how the Harlem “stride” tradition can influence modern-day piano improvisation without demanding that a player remains wholly “in the style” of that genre. I argue that Lopes’ “traversing” makes a pointed critique of the notion of authenticity as perpetuated by neo-classical figures such as Wynton Marsalis, as well as the binary dualism of modernity versus tradition; a closer scrutiny of the Harlem stride aesthetic suggests that one can reinterpret Carolina Shout in a personalised way while maintaining authenticity.

This paper will outline my motivation for this project, the reasons and historical precedents for it, explanations for the decisions I made in interpreting the piece, and a judgment of the project’s success. In doing so I hope to shed light on how one might engage in the current discourse that surrounds contemporary improvised music and jazz by the way of performance and writing.

My introduction will outline why I have chosen Carolina Shout in particular and the motivating forces behind my engagement with pre-bebop jazz piano. It employs an historical account of James P. Johnson during the “Harlem renaissance,” theorisations of improvisation by Lydia Goehr and of art by Jacques Derrida, Australian musicians Scott Tinkler and Allan Browne’s positioning of Australian jazz and improvisation in terms of notions of authenticity as perpetuated by Wynton Marsalis, and also looks towards Anthony Braxton as a synthesiser of disparate musical histories, to demonstrate the impetus behind this project.

The second chapter outlines concepts that helped shape my interpretation. It discusses multiple performances by Johnson, Thomas “Fats” Waller and Marcus Roberts in conjunction with essays by Henry Martin and Bruce Barnhart to outline musical and critical concepts that I used to inform my interpretation of Carolina Shout.

My third chapter will discuss my performance of Carolina Shout to show how the concepts discussed in Chapter Two were realised in performance, while my conclusion will offer final thoughts on the challenges I faced in realising this project, as well as possible future research projects.

Chapter One – Why Carolina Shout?

Carolina Shout is arguably James P. Johnson’s crowning piano piece. At one time it was the testing piece for many Harlem pianists, who “were hurling ‘Carolina Shout’s at each other like hand grenades” as part of cutting contests that formed a part of burgeoning Harlem rent-party scene (Jasen 2002 74). Bruce Barnhart summarises the historical importance of the piece thusly:

James P. Johnson’s performance practice has had an impact on virtually every pianist who has ever played jazz. The appeal that his playing had for musicians like Waller and Ellington lies at least partially in the way that he was able to synthesize an emerging blues style with the established compositional practices of ragtime, stride, and New York jazz, and in doing so to add a new level of fluidity, elasticity, and improvisational élan to an already compelling music (Barnhart 2010 842).

Even though cutting contests are less explicit in jazz today than they were (although jam sessions still exist as a proving ground in many respects), they still provide a link between the Harlem renaissance and today. One such instance of a modern cutting contest can be see at amateur nights such as the one at the Apollo Theatre. Although not conceived of as cutting contests as such, amateur night at the Apollo Theatre still carries a subtext that lies at the core of the cutting contest: that of one-upmanship. In “The Agon of Improvisation - On Broken Strings: Towards a Theory of Fit and Wit,” Goehr uses the Apollo Theatre open mic night to theorise improvisation as a combination of “impromptu” and “extempore” practices, the former where one overcomes obstacles within a confined time-frame, and the latter, a more “open-ended” practice where one “works out” from a given starting point (Goehr (forthcoming) 10). Hence improvisation gives performers a chance to “demonstrate our humanity and sociability, but also when we want to exhibit our winning, individual selves" (Goehr (forthcoming) 2). I conceive my performance of Carolina Shout as an effort to simultaneously and agonistically demonstrate both “fit” and “wit,” that is, demonstrate how I, through improvisation, negotiate obstacles that arise during a “working through” of Carolina Shout.

Two avenues have facilitated my own connection to pre-bebop piano: Wynton Marsalis’ canonisation of past jazz masters and pieces and his effect of jazz education, and my performances with the Australian musicians, drummer Allan Browne and trumpeter Scott Tinkler.

Wynton Marsalis’ use of the “Jazz at Lincoln Centre” program as method of canonising, normalising and legitimising past masters and works of jazz has had widespread effect on jazz education. This was evident in my own education at the Australian National University, Canberra, where great emphasis was placed on being familiar with as much of the jazz canon as possible. Hence I have always regarded the entire history of jazz as potential material as a conduit for self-expression. This, however, raises some problems between my own practice and Marsalis’ doctrine.

Marsalis has continually championed and held up as relevant past masters such as Duke Ellington, Thelonious Monk and “Jelly-Roll” Morton. While these musicians received some recognition for their contribution to American culture, they have certainly not received the recognition they deserve. This has a long precedent in the critical reception of jazz; the ignorance, wilful or otherwise, demonstrated by critics in the 30s of the artistry of “swing,” for example, has strong ties to racism in the United States (Gennari 1991). Hence part of Marsalis’ project is for greater recognition of black jazz artists as creators and masters of America’s black art form and to dispel the “noble savage” myth linked to black jazz musicians (Ake 2002 37).

In order to do so, Marsalis has adopted, consciously or otherwise, certain performance practices that resemble those from western art music concert tradition, as well as an essentialist stance towards jazz’s ontology. Western art music is the epitome of legitimised music. As Lydia Goehr has shown, many of the performance practices associated with western art music were developed as part of its practitioners vying for legitimacy in Europe (Goehr 1992). These include many of the practices we now regard as “normal behavior” in a concert hall: a hushed and darkened atmosphere that symbolically separates the audience from the performers, all performers (perhaps with the exception of the opera diva) dressed in black to subsume them under the invisible glare of the composer or conductor, and most importantly, the fixing of the musical works into reified, never-to-be-altered objects. Many of Marsalis’ concerts at Lincoln Centre exhibit these traits, which are being used to help legitimize jazz as an American art form.

DeVeaux points out a certain kind of irony in Marsalis’ practice that explicitly links it to what many see as the stifling “fixed-ness” of western art music practice:

Marsalis is careful to present Jazz as a cultural heritage and, in a sense, a political reality, entirely separate from the European tradition. But his celebrated feat of winning Grammy awards for both jazz and classical recordings underscores the extent to which jazz has become another kind of classical music-one indigenous to black culture and reflecting black values, but following the same pattern of institutionalization in conservatories and repertory groups, and demanding of its musicians an empathetic response to aesthetic sensibilities of the past. Historical narrative plays a crucial role in the formation of a canon, in the elevation of great musicians as objects of veneration, and in the development of a sense of tradition that casts a long shadow over the present. The goals of the neoclassicists will have been admirably fulfilled if and when busts of Armstrong and Parker stand alongside busts of Beethoven and Bach in practice rooms and music studios across America (DeVeaux 1991 552).

Although discussions by Bailey and Benson point to varying degrees of improvisation being present in many musical practices outside of jazz (Bailey 1992 , Benson 2003), the performance of most western classical music remains relatively fixed. As Deveaux points out, Marsalis runs the risk of “fixing” jazz in a very similar way.

The Lincoln Centre’s (with Marsalis as artistic director) mission statement expresses the essentialist stance they have taken towards the question of “what jazz is.” Blues and swing, according to them, are indispensible traits that help jazz act as a metaphor for democracy (Jazz at Lincoln Centre 2013). This essentialist position has led Marsalis to discount anything that deviates from blues and swing as “not jazz.” This friction is a major motivation behind this performance project. Tying authenticity top history is an effort to use the jazz canon as a legitimizing force for Marsalis’ own contemporary jazz practice, the implication being that every self-respecting jazz pianist should revisit what is often broadly referred to as “stride piano.” Pianists in various bands led by Marsalis - Kenny Kirkland, Marcus Roberts, Eric Reed and Eric Lewis - all reference the “stride” tradition in explicit ways. Marsalis’ position is summed up nicely by his verbal introduction to Eric Lewis’ solo piano version of Cherokee:

Here’s the inimitable, top professor, Eric Lewis in a solo performance of a song that was made popular by Charlie Barnett, and famous by Charlie Parker, composed by Ray Noble . . . it’s entitled Cherokee. This is what the modern piano is about. (Lewis)

Not only does Eric Lewis receive the decoration of “top professor,” he also is credited by Marsalis as definitively demonstrating what it means to play “modern piano.” The second claim, that this performance is exemplary, hinges on the first. Even Marsalis’s verbal introduction itself demonstrates the legitimizing force of history. By citing the composer, and two musicians who helped popularise the piece, Marsalis is demonstrating his knowledge of the history of this particular piece, the implication being that he is therefore qualified to bestow the title of “top professor.”

So, how might a performer seriously engage in the music of James P. Johnson, for example, without necessarily subscribing to Marsalis’ essentialist stance? This is the question I asked myself as I developed an that led me away from Marsalis’ essentialism, while not necessarily foregoing the musical material itself. Australian musicians Allan Browne and Scott Tinkler are exemplary in this regard; they both describe what it means to find your own voice as an Australian musician who is engaged in the history of jazz.

Browne’s first contact with jazz was with the music of Ellington, Armstrong and New Orleans clarinetist George Lewis. While his reverence for the music ran deep, his comment, “we were totally connected with that early Ellington period . . . as much as middle-class boys in Australia could,” outlines the inherent distancing many Australian jazz musicians feel towards American jazz. Scott Tinkler echoes this in a 2009 interview:

For me, the tyranny of distance has been much more a positive than a negative. I have found that living in Australia and not having the history and expectation of jazz hanging over my head is somewhat liberating. I've always loved jazz but do not by any means profess that I am a jazz musician. Jazz is a style that I've studied and played but I consider myself an improvising musician (Vantrikt 2009).

Similarly for Browne, American jazz is something that can be re-appropriated towards Australians’ own expressive ends. Browne uses the example of the press-roll, a technique used by New Orleans drummer Baby Dodds, and emphasises both its historical importance and its use as a contemporary technique: “I am very likely with Sam and Marc to play press rolls in the style of Baby Dodds, simply because it's something I do naturally and it just happens, I don't think about it . . . that sort of drumming . . . is the same as Tony Williams with Miles or Brian Blade now . . . really, it's interactive” (A.B.C. 2010). In the end, for both Browne and Tinkler, playing music together is not about whether the music is jazz under Marsalis’ definition, but how Australians can interact through music and develop a sense of self-identity:

I'm into interaction . . . whether it might be very contemporary, or it could be totally in the pocket, like in a New Orleans band (A.B.C. 2010).

For me that I had to continue to work on creating my own style through playing music with my peers and develop a language that was relevant to us here in our “motherland” . . . being from Australia and not really feeling a direct connection to jazz, I don't feel the need to be, or at least to call myself, a jazz musician (Vantrikt 2009).

This is a position I strongly identify with. Particularly relevant to my project of interpreting Carolina Shout is the notion of exploring jazz as a means of articulating a personal conception of art making and establishing group identity. As Kofsky shows, the complexity of be-bop was, in part, an effort by black musicians to assert group identity and agency against appropriation by white promoters (Kofsky 1971 423). In the sense that any jazz musicians outside of the United States (or even New York) is automatically “othered” from mainstream jazz practice, this performance project is partly a continuation of the Australian jazz tradition of establishing and enacting group identity that places Australian jazz musicians adjacent to, but also outside of, most America-centric jazz practice. While Marsalis might see this position as ignoring, and perhaps even working against, his efforts to have black American jazz musicians properly recognised, I feel it does more to honor the tradition in jazz of articulating a personal artistic outlook and forming group identity through music.

Eric Drott, in his discussion of French free jazz musicians’ look towards the music of Ornette Coleman, Eric Dolphy, Cecil Taylor, and other civil-rights-minded black American jazz musicians, shows how, ultimately, "the attempts of French free musicians to align themselves with American political movements only reinforced the perception that their appropriation of free jazz was illegitimate. Not only had the music been imported, but so too had been the injustice against which it protested" (Drott 2011 140). Claiming a legitimate connection to black American’s struggle for recognition through their art form while “importing” these injustices against a distinct social background runs the risk of becoming cultural appropriation; an antithesis to the aims of my project. Hence my performance will eschew notions of authenticity as defined by Marsalis and will combine my own aesthetic of improvisation with Carolina Shout by the way of broader aesthetic, sociological and philosophical discussions. While in doing so I will traverse boundaries that define jazz’s ontology for an essentialist like Marsalis, my dialectical approach pays more respect to the figure of James P. Johnson, who continued to expand the scope of his music-making into his last years (Jasen 2002 78/79), than any imitation could.

Paul Lopes and Jacques Derrida describe the usefulness of this “traversing of borders” in more detail. While the notion of jazz as an “art-music” has morphed according to the social dynamics of the time, Lopes suggests that by traversing and exploring the boundaries of what makes jazz “art” we highlight the what these very borders are and what role they play:

The evolution in the meanings and practices of jazz music over time traversed numerous boundaries of cultural distinction in America. This traversing of cultural boundaries forces us to understand why such boundaries existed in American culture and why musicians and others were compelled to transgress them (Lopes 2002 9).

This is very similar to Derrida’s notion of the parergon as something that reflects, even epitomises, the ideology of the framer. Rather than being a neutral, objective and universal accessory to the work of art, the parergon, if interpreted as integral to the artwork, elucidates the relationship between how the artwork is “framed” and its relationship to what is outside the frame:

What is incomprehensible about the edge about, the a-bord appears not only at the internal limit, the one that passes between the frame and the painting, the clothing and the body, the column and the building, but also the external limit. Parerga have a thickness, a surface which separates them not only (as Kant would have it) from the integral inside, from the body proper of the ergon, but also from the outside, from the wall on which the painting is hung, from the space in which statue or column is erected, then, step by step, from the whole field of historical, economic, political inscription in which the drive to signature is produced. No 'theory,' no 'practice,' no 'theoretical practice' can intervene effectively in this field if it does not weigh up and bear on the frame, which is the decisive structure of what is at stake, at the invisible limit to (between) the interiority of meaning (put under shelter by the whole semioticist, phenomenologicality, and formalist tradition) and (to) all the empiricisms of the extrinsic which, incapable of either seeing or reading, miss the question completely. (Derrida 1987 60/61)

If what Derrida is suggesting, that critiquing the “frame” of the artwork is the most effective way to show “what is at stake,” is correct, then my performance should critique the “frame” around the notions of modernity and authenticity by traversing them.

One final example of “musical traversing” comes in the form of Anthony Braxton, who, as a means to self-expression, drew equally on the history of Afrocentric spiritualism, jazz history and twentieth-century western art music. As Ronald M. Radano notes, “both [Arnold Schoenberg and Ornette Coleman] spoke in a radical, artistic tongue that conflicted with the enforced homogeneity of mass cultural existence” (Radano 1993 74). While Braxton’s aesthetic was developed alongside and in response to the civil rights movement in the United States, Radano suggests that Braxton’s protest comes in the form of a highly personalised and radical aesthetic; one that cuts across the boundaries of race in order to gesture towards class, rather than racial, strata. The fact that Braxton sees jazz as forming only part of his musical vision suggests his music might be simultaneously thought of as exemplar of a personalised musical fusion of styles and as a critique of narrow definitions of genre. Braxton’s music is not jazz, but it is a critique of others defining jazz narrowly. While my project doesn’t have the same sort of connection to the civil rights movement as Braxton’s, it certainly is a commentary on the ongoing discussion surrounding race and contemporary jazz. In a climate where lines are drawn, along aesthetic and racial grounds to demarcate authenticity, my project aims to utilise Braxton’s methodology in performance to form a modern jazz cultural critique.

Now that I have shown the reasoning behind this project, I will discuss writings and recordings that have informed my interpretation of Carolina Shout.

Chapter Two – Interpreting Carolina Shout

My first point of call for learning Carolina Shout was recordings, ones made by Johnson in particular. As George Lewis has noted, one primary difference between “eurological” and “afrological” approaches to music-making revolves around traditional western notation. A “eurological” approach to music generally stems from western forms of notation, whereas the “afrological” approach makes primary use of the oral tradition (Lewis 2002). Lewis is also points out that “ ‘afrological’ and ‘eurological’ systems of improvisative musicality refers to social and cultural location and is theorised here as historically emergent rather than ethnically essential” (Lewis 2002 217). Hence I chose to learn Carolina Shout via an “afrological” route: by transcribing it from Johnson’s 1944 Decca recording. I used this recording in particular because I had listened to it the most, and the recording fidelity seemed high enough to as to make transcription achievable. I purposively avoided using any sort of notation as a mnemonic device, and memorised each section as I learnt it. If I forgot the section the following day, I re-learnt it in the same way.

Ethan Iverson suggests that this kind of “afrological” (although he doesn’t use that word) method of learning helps avoid the pitfall involved in regarding one particular version as an urtext; it opens up a space for a personal approach to interpretation: “Since ‘authentic’ sheet music has never been available anyway, this approach makes a lot of sense.  That may be one of the most important yet misunderstood aspects of the real Harlem Stride tradition:  take the basic material and make it your own” (Iverson September, 2009). My avoidance of sheet music thus positions my initially learning of Carolina Shout away from the “eurological” ideal of pieces existing in idealised states, and more towards the variations and contingencies that fall in line with the jazz tradition.

Having explored one version of Carolina Shout, I began investigating four other recordings by Johnson, as well as two by other pianists to inform my interpretation. The four other recordings by Johnson are the Artempo piano roll 12975 from 1918, the QRS piano roll 100999 from 1921, a 1921 recording on Okeh, and a recording as part of a John Hammond 1938 concert, which was then released on 1999 on Vanguard records, titled From Spirituals to Swing. Thomas “Fats” Waller, who was Johnson’s protégé, made a recording in 1941 on Victor, and Marcus Roberts presents his version on his 1993 album If I Could Be With You. Each of these renditions offers suggestions as to the way I might develop and personalise my own version of Carolina Shout.

Also pertinent to this discussion are two papers specifically concerned with Carolina Shout: “Balancing Composition and Improvisation in James P. Johnson's ‘Carolina Shout’ ” by Henry Martin, and Bruce Barnhart’s “Carolina Shout: James P. Johnson and the Performance of Temporality.”

I will now discuss some of the elements that these recordings and papers present that have informed my rendition.

The Artempo piano roll 12975 exhibits an asymmetrical form during its third, or “B” strain, beginning at 0:51 and lasting to 1:36.[1] Generally speaking, the structure of each strain in most pre-swing jazz follows a similar pattern: the first eight-bar section introduces a new idea and concludes with a modulation that acts as a harmonic turn-around, the second eight-bar section begins identically to the previous eight bars, but replaces the modulation with a cadence in the local tonic. Thus a typical strain may have a structure similar to that presented in Table 1.

Table 1: Typical structure of a single strain
Introduce new thematic material
Modulation and harmonic turn-around
Repeat thematic material
Cadence in local tonic key
4 bars
4 bars
4 bars
4 bars

It should also be noted that the B section tends to begin in a similar way to the A section, and only presents its harmonic deviation in its second half, while the C section begins its deviation turn right away.

Instead of following this pattern, the second strain of the Artempo piano roll has the form: ABAB (16 bars) followed by A (4 bars) and C (3 bars), before completing a full ABAC 16-bar cycle. This 3-bar C-section, which has its final bar removed, creates an asymmetry that is never subsequently resolved. If we regard form as, in some sense, rhythmic, then we can view this truncation as a kind of form-related syncopation that resembles, although on a different scale, what Henry Martin refers to as a “change step” (Martin 2005). According to Martin, a “change step” disrupts the “march-like rhythm of the left-hand” (Martin 2005 279) and is one of the salient features of “stride” as opposed to “ragtime” piano. “Change steps” are evident in all of the recordings of Carolina Shout discussed here, demonstrating their centrality to Johnson’s style. Hence the asymmetrical C-section in this recording metaphorically reflects the left-hand syncopations of Johnson’s style, which in turn points to the larger project of increased rhythmic complexity in stride piano. Jasen and Jones make this explicit, stating “stride takes syncopation even further than ragtime does - takes it, in fact, to its human limit - by freeing the left hand to syncopate with the right. It is the most sophisticated of all popular piano playing.”

Martin shows how subsequent versions of Carolina Shout performed by Johnson refine and solidify the rhythmic complexities of earlier versions. He also discusses “Fats” Waller’s original embellishments during the “trio” section (Martin 2005 291/292). While it seems that Johnson in some sense reifies the variations he presents in the “trio,” it is Waller who injects new life into Carolina Shout by developing his own approach to its performance. But before I turn to Waller, I will briefly discuss to elements Johnson’s other recordings that I relate to my own practice.

While Johnson’s 1921 recordings differ dramatically from the Artempo piano roll in their form and melodic figuration, they differ in one important regard.[2] A comparison between the two confirms Martin’s point that “in general, the trio is where the stride pianist is most likely to elaborate thematic material” (Martin 2005 286). The trio, which appears as the “F” section on the QRS piano roll, 2:45 – 3:18, houses Johnson’s variations, which don’t appear at all on the Okeh recording. This was a clear indicator to me that the trio section was the point at which I should begin introducing my own variations.

Johnson’s 1938 live performance as part of John Hammond’s From Spirituals To Swing concert contains curiously little extemporisation in the final trio section. It does, however, exhibit the same form as the Decca recording, confirming Martin’s point that Johnson solidified his approach to Carolina Shout as the years progressed (Martin 2005 291). Johnson also makes increasing elaborate use of the “change step” technique, sometimes prolonging the “changed” beat for longer than ever before. Table 2 uses similar notation to Martin to show how the Johnson’s left-hand “change step” works; “1” denotes a bass note, while “2” denotes a chord.

Table 2: Increasingly complex “change step” techniques by Johnson
Measure 1
Measure 2
Measure 3
Measure 4
“Standard” left-hand pattern
1 2 1 2
1 2 1 2
1 2 1 2
1 2 1 2
Basic “change step” pattern (from Martin’s example 2)
1 2 1 2
1 2 1 2
1 2 1 1
1 2 1 2
Complex “change step pattern” (0:33 – 0:38 of From Spirituals to Swing)
1 1 2 1
1 1 2 1
1 1 2 1
2 1 1 2

Table two shows Johnson’s development of the “change step” idea, which reaches its peak in between measures three and four of this excerpt from his live performance. The four underlined numbers here show how the “standard” left-hand pattern of 1 2 1 2, where “1” traditionally coincides with the strong beats of the bar, has been displaced so that it instead begins on the weak second beat of the bar. This heightened syncopation is disorienting for the listener, and again demonstrates Johnson’s penchant for rhythmic complexity.

To summarise, the Artempo piano roll strongly suggests to me a notion of asymmetrical rhythm that I then used as an inspiration for my own rhythmic approach to extemporisation (to be discussed later). The 1921 recordings demonstrate where those extemporisations should, in the traditional sense, take place, and the live performance confirms the form exhibited on the Decca recording as the most developed version of the piece. Of course, all versions exhibit “change steps” and heightened syncopation between right and left hands, with the live recording from 1938 exhibiting Johnson’s most highly developed form of left hand, “change step,” syncopation.

As mentioned previously Thomas “Fats” Waller’s 1941 recording presents a highly personalised alternative to the recordings by Johnson, making it the archetypal example of Carolina Shout as realised by someone other than the composer. Although the characteristic rhythmic buoyancy of Waller on this recording immediately sets it apart from Johnson’s recordings, there are four other features contained in this performance that inform my interpretation.

Firstly, Waller’s alters the formal structure. The “trio” section first appears as the third strain (In Johnson’s performances it had always appeared as the fourth), is followed by two more strains (that also appear in Johnson’s recordings), and then reappears as the strain containing extemporisation. Secondly, and as Martin points out, whereas Johnson’s “trio” extemporisations remained similar one another, Waller’s (1:41 – 1:57) are all his own (Martin 2005 292). This opens the space for me to take a more personalised approach to improvisation in this section in my interpretation. Thirdly, Waller re-plays the opening “A” section as the closing strain of the piece: something Johnson never did. This “recapitulation” seems much more akin to the sorts of rounded AABA forms used by swing and bebop musicians in the 1930s and 1940s; perhaps a scene the younger Waller was more familiar with than Johnson. As Paulo Lopes points out, the swing movement of the 1930s were seen by many as a radical departure from “hot jazz” (Lopes 2002 157), arguably “progressing” to the high modernism of be-bop. I conjecture that this reappearance of the “A” strain at the end of Waller’s Carolina Shout is yet another signifier of modernity, thus becoming another argument for my project of adapting Carolina Shout to my own modernist ends, and against any attempt at mere recreation. Johnson even suggests this through one of his other titles: You’ve Got to be Modernistic. Barnhart suggests that the return of harmonic sections in Carolina Shout renounces “any transcendence that would move away from the subjective (time) and social (space) conditions of its existence” (Barnhart 2010 847). The return of sections or material in Carolina Shout can be seen refusal for the music (as well as Johnson, Waller, and later, Roberts and I) to let go of the past; something that African Americans themselves could and should not do given their recent and continuing mistreatment in America: Carolina Shout insists that the present is always insuperably attached to both the future and the past, and that repetition, if handled properly, is more likely to open onto a future possibility than onto a regressive past” (Barnhart 2010 845). Further to this, Barnhart remarks that Carolina Shout continues to “open onto future possibility” by remaining incomplete, not in the sense of performer or composer error, “but that it adheres to an aesthetic of incompleteness, and that it takes as its subject the incompleteness of time and of the ‘now’ that an objective, linear version of it presents.” How might my interpretation embody this “aesthetic of incompleteness”? My approach is to maintain a sense of open-ness in performance; I allow myself the freedom to explore whatever musical suggestions appear during the course of a performance. This does not mean I have to explore all of them, but that I’m free to explore should I wish to do so. While I have a rough idea as to the structure of my performance, particularly prior to “trio” section, I allow myself to be “waylaid” and led into previously uncharted territory, meaning (to use the terms suggested by Lydia Goehr) that my skills of “fit” and “wit” are displayed via both impromptu and extempore performance. By refusing to “complete” my arrangement by predetermining its structure, I preserve Barnhart’s notion of Carolina Shout opening onto “future possibility.” Finally, to return to Waller, his coda (2:10) continues the theme of rhythmic complexity by the way of polyrhythmic dotted crotchets. Thus the performance ends on a note of rhythmic irresolution that works against the concurrent V7-I progression.
To summarise, Waller’s rendition of Carolina Shout confirms the “trio” section as the point where extemporisation might begin, also suggests that the element of form might also be manipulated towards a personalised performance, and introduces polyrhythm into the rhythmic language. This final point will prove to be a major one in light of the sorts of rhythmic techniques I plan to include in my own interpretation.
Marcus Roberts’ 1993 rendition exhibits a much more open approach to form, improvisation and an increasingly developed rhythmic language. Roberts waists little time introducing improvisation, using the “A” strain as a platform for improvisation immediately after the opening theme rather than waiting for the “trio” (0:29 – 0:46). Even so, Roberts acknowledges the “trio” section as an important point of departure in his sudden contrast in textual clarity, using liberal sustain pedal prior to, and none after, the arrival of the “trio” at 2:02. This creates an extremely clear auditory division between the two sections, as if the highlight the “trio” as an important arrival point.

Roberts’ improvisations also have an interesting relationship to those of Johnson and Waller. Martin makes the point that improvisation, for Johnson, was not a linear affair in the way we might associate with be-bop pianists such as “Bud” Powell: “Johnson never abstracts a chord progression from a block for ‘blowing’ (improvising with extensive departure from the melody); instead, the strain remains compositionally focused on its original melodic material or on a pattern specific to each variation that relates to it” (Barnhart 2010 294). Barnhart discusses the importance of “call and response” figures, saying that they are “not a back and forth between discrete subjects, but a coeval call-response rhythm in which listeners, dancers, and performers collaborate in revising and refiguring their shared time and space” (Barnhart 2010 847). “Call and response” patterns, unlike linear improvisation, form an “unending chain of antecedent and consequent responses and calls, even if they occur at the beginning or end of a piece. Calls are responses to previous responses, and responses become calls to which subsequent calls respond. In such a pattern, no musical figure can claim priority as origin” (Barnhart 2010 846). By forming an unending chain of antecedent and consequent phrases where no single event has priority over the other, Roberts’ performance ensures the kind of incompleteness that allows it to remain open to “future and past.”
Roberts, in direct lineage from Johnson and Waller, also introduces a new level of rhythmic complexity by displacing the typical “1 2” left hand pattern by a quaver, placing it on each “off” beat. Thus the “change step” technique, instead of being realised at the level of crotchets, is now realised at the smaller level of quavers (0:56 – 1:04).

To summarise my survey of recordings and papers, there is a clear development in rhythmic complexity through the original and subsequent versions of Carolina Shout, realised primarily through left-hand figuration, but also in Waller’s introduction of polyrhythm and Roberts’ greater emphasis on right-hand extemporisation. Also, each rendition posits a flexibility of form in order to realise the performer’s personalised rendition. Martin’s writing supports this chronological development via his survey of Johnson’s, and then Waller’s, realisations, while Barnhart suggests that concepts of incompleteness and return are integral to any performance that is to remain true to the spirit of the original.

Continuing this train of thought, stride piano is partly a modernist movement (another of Johnson’s compositions is titled You’ve Got to Be Modernistic), and this is, in part, realised in the form of rhythmic complexity. I hence feel warranted, in order to “make it my own,” in introducing my own rhythmic complexities into my rendition of Carolina Shout.

Chapter Three – In Performance

I will now discuss my performance of Carolina Shout. Most of the reasoning for my decisions should obvious from the previous chapter, and will need only minimal explanation. My goal in this section is not to detail exactly what I played (hence I will no provide a transcription), nor is it to discuss my practice methodology in arriving at this performance (although this discussion would be fruitful). I will instead discuss some of the materials that occur in my improvisation, which begins at 1:44 in the accompanying recording.

Before I begin, I should say something about the musical language I bring to this scenario. Perhaps the biggest influence on my language for musical improvisation is the piano music of American composer Elliott Carter. A detailed presentation of this research can be founding in my paper “Elliott Carter's Rhythmic Language: a Framework for Improvisation” (Hannaford 2011). Since then I have spent time developing Carter’s harmonic materials, in particular the 0146 all-interval chord, the 012478 all-triad hexachord, and a small selection of all-interval, twelve-note chords, for use in my compositions and improvisations.[3] Hence these harmonic materials are bound to influence my improvisation; they may sometimes appear explicitly, but probably more often act as underlying structures for what I play.

Entering into this performance my basic template for performance was as follows:

  • 1.     Perform Carolina Shout as per the Decca recording, with some variations
  • 2.     Begin more overt improvisation at the “trio” and introduce more personalised musical language
  • 3.     Develop this material gradually and without necessarily adhering to the harmonic structure of Carolina Shout
  • 4.     Gradually return to the final “shout” strain of Carolina Shout and end the performance

The first 1:44 of my performance remains more or less faithful to the Decca recording by James P. Johnson. It is the material I transcribed and memorised and then shaped through practice. There are some minor differences in left-hand chord voicings and right-hand figuration, but, by and large, each of Johnson’s strains closely resembles those on the Decca recording. I would describe the variations that occur in these sections as products of the “afrological” approach I have taken; when one learns a piece by ear and without formal notation as a mnemonic device there are bound to be artefacts of difference.

The “trio” section begins at 1:44, and marks the beginning of my extemporisation. Beginning my extemporisation at this point creates a tangible connection the 1921 Artempo piano roll as well as Waller’s rendition, discussed above. I begin my extemporisation with a series of percussive chords whose character are derived from the second, third and fourth strains of the composition. They are organised into two-bar phrases to form a question and answer pattern, a motif that Martin suggests is intrinsic to the Harlem stride style (Martin 2005 286). I return frequently to similar   chordal passages as a way to “point back” to these opening question and answer patterns, thereby maintaining the sense of “return” discussed by Barnhart as well as strengthening structural unity. One such instance can be heard at 2:17, which follows on from a section of more linear extemporisation.

A primary goal, during 1:44 – 2:35, is to create continuity between my improvisation and the main thematic material of Carolina Shout. I therefore gradually emphasise my linear improvisation over and above the Harlem stride aesthetic while attempting to maintain a tangible connection between the two.  At 2:35 the sudden disruption of the (heretofore regular) left-hand pattern signifies the beginning of more drastic rhythmic developments. Although it is similar in conception to the sort of left-hand syncopation used by Roberts in his interpretation, it points towards something much different.

The next level of rhythmic modulation/disruption comes at 2:49, where my right hand plays five crotchets for every four in the left hand. This (for me) is rhythmic language derived from Elliott Carter, and has no obvious antecedent in Johnson’s music that I know of. The left-hand synchronisation with this right-hand polyrhythm suggests a faster tempo during 1:56 – 3:01, again pointing towards the continuing rhythmic development of my performance.

At 3:05 I move to Johnson’s opening strain as a harmonic basis for improvisation. Roberts is the only pianist who improvises on this section, although, as mentioned, Waller performs the opening material as a recapitulation. It is another of Barnhart’s “returns,” and also lends more harmonic momentum to the ongoing improvisation through its increased harmonic diversity. My right-hand improvisation continues to develop harmonic and rhythmic complexity, and my left-hand develops complex “change-steps” that take their cue from Johnson’s live recording.

At 3:37 I hint at one of Johnson’s other famous compositions, The Charleston, composed in 1923. Although I didn’t conceive of this rhythmic gesture as such, it helps create the sense of simultaneously “looking forward and back” discussed by Barnhart.

At 3:35 I return to rhythms based on a quintuplet subdivision. The right-hand improvisation at 3:56 sounds all of the quintuplet semi-quaver before moving into 5:2 (five semiquavers for every two crotchets) and 5:3 (five semiquavers for every three crotchets) polyrhythms.[4] Increasingly, but gradually, my performance moves away from explicitly referencing Johnson’s material and towards my own as a basis for improvisation. I propose that doing so is a musical realisation of the “traversing of boundaries” discussed (see Chapter One) by Lopes and Derrida; by making relatively gradual transitions between contrasting musical materials I hope to highlight these contrasting musical world’s similarities, thereby showing that the boundaries at the edge of jazz (or any music for that matter) are not rigid but fluid.

At 4:05, partway through the sixteen-bar section, I forego the harmonic form for a more improvisative approach. At first my left-hand maintains the “1-2” pattern, again creating continuity, but later moves to more arpeggiated figures (see 4:13). Harmonically, many of the materials I cited as coming from Elliott Carter are foregrounded in this section, and most of the rhythmic material comes from the quintuplet-based division introduced in my earlier improvisation.

From here until 5:05 I extemporise on the Carter-derived harmonic and rhythmic materials in an increasingly personalised way. Listening back to this section, I personally find it interesting as to how much it reminds me of Carolina Shout despite not explicitly referencing it. Perhaps, in drawing attention to the inter-connectedness of different “genres” of music, my project is at least partially successful, although it’s difficult to tell whether this is a product of the interpretation itself or inextricably tied to my being the performer.

At 4:45 I briefly reference the harmony of Johnson’s opening strain, while at 5:05 I strongly reference the left-hand “1-2” figure and the harmony. Although interrupted by more Carter-derived material, this marks the beginning of my return to Carolina Shout; my goal is the final “shout” section, which will serve as a gateway out of my improvisation and towards the end of the piece.

Before the end arrives though, there is one more exploration: at 5:35 I metrically modulate using the 5:2 polyrhythm, transplanting the harmonic form of Johnson’s opening strain into a faster tempo. At 5:52 my right hand plays a series of chordal dotted crotchets that simultaneously references the question and answer phrases at the opening of my improvisation and “Fats” Waller’s polyrhythmic coda. This section embodies the synthesis of Carter (metric modulation via 5:2 polyrhythm), Johnson (the harmonic form, the Harlem stride tradition) and Waller (dotted crotchets).

At 6:06 I re-enter Carolina Shout with the final “shout” section, albeit at the faster tempo. The increased tempo embodies the ecstasy this final section is intended to stimulate; rousing dancers to a frenzy of collective frenzy or, to use Barnhart’s terms, “shared time and space” (Barnhart 2010 847). Following this I play the coda from the Decca recording and finish the performance.

Chapter Four – Conclusion

This performance project and accompanying discussion attempted to draw our attention to what makes up the “frame” around jazz practice. My goal was to call into question the notions of authenticity posited by Marsalis, and to substitute a more philosophically informed, flexible and dialectical approach to performance that takes its cue from Anthony Braxton as well as the Australian jazz tradition. I think of the mode of performance outlined here as part of an established tradition in Australian jazz that intersects with musicians such as Braxton who are able to carve out a place in the American jazz tradition for their own personalised approach. The juxtaposition of music from Johnson and Carter, realised through my improvisation, posits that the boundaries that (for some) define what makes authentic jazz are much less reified than they might appear, and that this investigation of the history of jazz practice reveals a tradition of experimentalism and the cultivation of a personalised approach to music-making. This is not to say that one can simply ignore the history of jazz and replace it with “individualism”; the history of jazz is a rich and varied bank of music. Rather, jazz history provides countless examples of musicians developing their own voice. In fact, it is the selective skimming of jazz history that often results in a normalised view of jazz and its practitioners; musicians who fall outside of categories such as “be-bop,” “hard-bop” and “free jazz” are often marginalised. There are countless examples of this (Herbie Nichols immediately springs to mind). David Ake has a pertinent discussion regarding Louis Jordan, who becomes a problematic figure if one adheres to fixed notions of genre (Ake 2002 42).

One of the major challenges for this project is to be convincing through performance. It is up to the reader and listener as to whether my performance and discussion does ample justice to the jazz tradition. I am the first to admit that my performance of the main section of Carolina Shout is far from flawless; I wonder if a more polished rendition would lend my argument more weight. According to the argument presented in this paper, the “authenticity” of my performance should no depend on my flawless reproduction of one of Johnson’s recordings, but on the sorts of concepts identified by Barnhart and Martin. My rendition as it stands might also be heard as a critique of the overly polished stylings that dominate jazz practice today. As Nicholson and Ake both point out, many jazz students today seem to forego emotional content for technical prowess (Ake 2002 , Nicholson 2005). I’d also like to think of my pianistic technical deficiencies as evidence of my performance as in a state of continual “becoming,” rather than conclusion.

I have already mentioned some of the discussions that I feel would be worth investigating further. An “auto-ethnographic” discussion, in particular, that outlines how I practiced and developed this performance, would perhaps be one of the most fascinating and instructive; it would provide a template for other musicians looking to develop a similarly personalised approach to jazz. Also, a more nuanced discussion of the interaction between jazz practice, education, and concert curating would shine more light on how jazz practice is regulated and defined, for better or worse. Another avenue worth exploring would be how, if at all, the performance of music “shows” things to listeners. Throughout this paper I have spoken of my performance as “demonstrating” and “showing” things that, in a way, presume the listener would be able to hear these things without me pointing them out. A discussion that engages more fully with music cognition, phenomenology and aesthetics might illuminate how this, in fact, takes place. Finally, the issue of “otherness” I used to contextualise the Australian jazz tradition brings with it discussions around race, gender and sexuality. Another project I of mine, “Lost in the Stars,” juxtaposes portions of the Zodiac suites of Mary Lou Williams with Karlheinz Stockhausen, not as a way of showing that Williams’ music is as “legitimate” as Stockhausen’s – Williams’ music can be judged on its own terms – but as a way of bringing each composers music into a common realm by the way of improvisation, thus preserving difference and identity while creating connections between seemingly disparate music (Allan Browne 2013). More performance projects and discussions that engage with the figures which fall outside of mainstream jazz would be a way of starting to reverse the normalised conception of the jazz canon that remains prevalent in pedagogy, academic discussion and practice.


[1] When using terms like “A” or “B” I am referring to the 1926 Publication of Carolina Shout by Barry Glover, Sr. and the James P. Johnson Foundation (
[2] The QRS piano roll is the one both Duke Ellington and “Fats” Waller learnt Carolina Shout from.
[3] More information about these harmonic materials can be found in Hopkins, N., and John F. Link, Ed. (2002). Elliott Carter Harmony Book. New York, Carl Fischer.
[4] Incidentally, this section reminds me of some of the piano rolls of Conlon Nancarrow, of whom I’m also a fan.