Sunday afternoon, November 8, 2-3:00 CST

Add this to your calendar

New Perspectives on Referents in Analyses of Improvisation

Andrew Goldman (Indiana University Bloomington), Chair
Matthew W. Butterfield (Franklin & Marshall College), Respondent

What do we mean when we say an improvisation is based on something? The late music psychologist Jeff Pressing proposed the term “referent”—“an underlying formal scheme or guiding image specific to a given piece, used by the improviser to facilitate the generation and editing of improvised behavior on an intermediate time scale” (1984, 346)—to encompass the myriad musical and conceptual structures underlying an improvisation. While Pressing initially characterized the referent as one input among many in a model of improvisers’ cognitive processes, later scholars sought to retool the concept to address particularities of specific improvisatory styles as well as analytical perspectives. Notably, Michaelsen (2019) has discussed how referents act as interactional influences during improvisations, and Love (2017) recontextualized the term from the ecological perspective of “affordances.” These attempts at retooling suggest that a critical reflection on the concept of the referent is in order. Are there aspects of the original definition that have disappeared in today’s discourse? Does the instability of improvisation’s work-concept, as discussed by music ontologists (Alperson 2016; Brown 2011; Gracyk 1996; Kania 2011), weaken (or strengthen) the legitimacy of the referent? How might the concept of the referent be updated to reflect new improvisatory techniques and music-theoretical models that came after its inception?

Recomposing the Referent: An Ecological Perspective

Marc Edward Hannaford (University of Michigan)

Marc Hannaford is a music theorist whose interests lie at the intersection of jazz and improvisation, identity (especially race, gender, and disability), and performance. He is an Assistant Professor of Music Theory at the University of Michigan’s School of Music, Theatre, and Dance, and completed his PhD at Columbia University with a dissertation on Muhal Richard Abrams, pianist, composer, and cofounder of the Association for the Advancement for Creative Musicians (AACM). His publications appear in Music Theory Online, Women & Music, and Sound American. He is also an improvising pianist, composer, and electronic musician who has performed and/or recorded with Tim Berne, Ingrid Laubrock, Tom Rainey, Tony Malaby, and William Parker.


Jeff Pressing’s original formulation of the referent encompasses many possible structures that undergird musicians’ improvisations, including a theme, motive, mood, image, emotion, or social situation (1984, 346). Importantly, he states that free improvisers create their referent as the performance unfolds (346). Pressing additionally asserts that improvisers vary their “attitudes” to the referent, encompassing imitative, antagonistic, and independent stances (348). Free improvisers therefore continuously modulate their relationships to an unfolding, multifarious, and coauthored referent.

In this paper, I argue that the concept of affordances, borrowed from ecological psychology, furnishes a compelling analytical embodiment of Pressing’s description. The concept of affordances helps characterize relational qualities between improvisers’ sonic gestures and the unfolding referent. It embraces relations of similarity and contrast, and encompasses connections facilitated both by musical segments’ theoretically defined parts and their “extra-musical” references.

Additionally, I offer a novel analytical methodology grounded in embodiment and recomposition. By playing along with a recording at the piano, stopping the recording immediately following a decisive juncture, and improvising various continuations, I explore various recomposed responses, ranging from the congruous to the incongruous and referring to multiple sonic and referential characteristics.

I employ both my ecological framework and embodied-recomposition methodology to discuss excerpts from two performances: “Focus, ThruTime…Time→,” by Muhal Richard Abrams and Fred Anderson, and “Kasumi” by Ingrid Laubrock and Aki Takase. In both cases, the analyses demonstrate my analytical methodology, as well as compelling aspects of these performances.

Being Self-Referential

Garrett Michaelsen (University of Massachusetts Lowell)


While jazz improvisations are often based on melodies and repeating harmonic frameworks, many other materials make up the pre-improvisational structures on which an improvisation may be based. The late music psychologist Jeff Pressing coined the term “referent” to describe these materials: “an underlying formal scheme or guiding image specific to a given piece, used by the improviser to facilitate the generation and editing of improvised behavior on an intermediate time scale” (1984, 346). In previous work, I have construed the referent as a “domain” of musical interaction that “operates like an additional band member, influencing the utterances of the musicians throughout and occasionally being modified by the musicians’ alterations or additions” (Michaelsen 2019, [26]). While this additional, invisible band member often takes the form of a set of musical structures, it sometimes takes the form of a person. In this paper, I explore the ways in which Miles Davis often refers to himself—in the form of his own prior utterances—while performing. In doing so, Davis creates complex chains of influence from one performance to the next. I will focus on one particular and notable chain in Davis’s recorded output: his existing recordings of the Tin-Pan-Alley standard “My Funny Valentine.” In reflection of his idiosyncratic improvisational aesthetic, Davis expressly refers to his own customary way of playing a tune over any other version. His playing establishes a self-referent that is inextricable from his musical identity; Davis doesn’t play “My Funny Valentine,” he plays himself playing it.

The Problem of Invisible Transcribers: Towards a Materialist View of the Referent

Joon Park (University of Arkansas)


Recently, transcription YouTube channels have gained in popularity. These YouTubers monetize their transcriptions either through a membership platform Patreon or through direct sale of the notation. In doing so, the transcribers claim a role as co-creator by requesting a fair share for their labor, contrasting the previous ways musicians bought transcriptions. Current copyright law allows publishers to omit transcribers while highlighting the improviser’s name on the cover, rendering the labor of transcription invisible. These YouTube channels, therefore, lay bare the publishers’ interest in diminishing the labor of the transcribers by presenting the transcription as a mere change of the medium (from recording to notation) as well as the inadequacy of the copyright law.

In this paper, I argue that treating improvisation as an ideal object contributed to the problem of invisible transcribers. This idealist view of improvisation is inherent in Jeff Pressing’s term “referents,” which portrays an improvisation as a result of the improviser’s interaction with ideal objects (e.g., a musical theme, a motive). An alternative is the “materialist view,” where performance’s documentation is itself considered as a creative act. I consider the recent theories of embodiment in music as a first step to situating improvisation within a materialist context. This investigation into an improviser’s body and an instrument’s physicality reveals the material dimensions of improvisation and proffers transcribers’ interpretations in creating notations. This view erases the conceptual border around improvisation as the only manifestation of creativity and considers improvisation’s dissemination as a continuation of creative instances.

Referents in the Palimpsests of Jazz: Disentangling Tune from Improvisation in Recordings of Ellington and Strayhorn’s “Satin Doll”

Sean R. Smither (The Juilliard School)

Sean R. Smither received his Ph.D. in music theory from Rutgers University and also holds an M.A. in music theory from Rutgers and a B.F.A. in jazz drumming from the New School for Jazz and Contemporary Music. His research focuses on the analysis of jazz standards, improvisation, and group interaction, and utilizes tools from linguistics, philosophy, and cognitive psychology to shed light on musical concerns. His doctoral dissertation, “Conceptualizing Tunes: Avant-Textes, Referents, and the Analysis of Musical Structure in Jazz,” explores the ways in which jazz improvisers conceptualize musical structure and develops a model for the analysis of jazz tunes. His research has been published in Music Theory Online and Theory and Practice (forthcoming), and he has presented his work at national and regional conferences throughout North America.


Benjamin Givan has compared jazz improvisations to palimpsests, writing that improvised utterances “provide us with clues to the improviser’s underlying conception” of the tune being played (2002, 41). The sounding music represents a negotiation between extemporization and fixed text wherein it is not always clear which elements are “part of the tune” and which are improvised. While recent research has sought to better understand both what a jazz tune is (Born 2005; Kane 2018) and how such tunes may be analyzed (Stover 2013), the palimpsests that are jazz improvisations continue to pose problems for analysts due to the dense entanglement of improvised utterances with the referents (Pressing 1984) that underlie them. In this paper, I take Givan’s above observation as the starting point for a methodology that disentangles improvisations from the tunes with which they intersect. 

I begin by outlining a theory of referent defaults, prototypical features grounded in a tune’s performance network that represent an improviser’s understanding of the musical structure of the referent. Using nine recordings of Duke Ellington’s “Satin Doll” as a case study, I show how various segments of the musical surface may be understood to convey referent defaults in varying degrees. By examining the interplay between referent and improvisation, we can better account for the tune as both an inspiration for and constraint upon the sounding music. This methodology reconfigures the objects of jazz analysis, suggesting new ways of understanding the relationship between improvisation and referent in jazz practice and beyond. 

The Referent’s Double

Michelle Yom (The Graduate Center, CUNY)


Jazz scholars generally agree that musical improvisation makes reference to preexisting schemas—these are collectively called “the referent.” There are two prevalent ways of understanding the referent. In the computational view, the referent is a fixed object, unaffected by the improvisation (Pressing, 1984). In what I will call the interactive view, the referent is a living, evolving process that changes in response to the improvisation (Love, 2017). Both views, however, fail to take sufficient account of the role of the observer: the listener/theorist. In a close analysis of “period 2” in Tree of Life by Cecil Taylor, in which I will play the role of the observer, I will show that taking the observer into account complicates a stable notion of the referent due to the self-referentiality of the listener. In what I call the referent's double, the listener's innumerable references in the process of listening include the relationship of notation to sound, embodiment of gesture, emotion and affect as it is encoded in music, and the relationship of listening to analysis. I argue that factoring in the register of the listener, or the observer who makes sense of the music to nominate the factors from which to draw and theorize the referent, destabilizes the notion of the referent. Listening to improvisation, like improvisation, involves inter-referentiality inherent to any listening subject, including the music theorist.