Facets of Musical Existence: A Critical Overview of Five Issues in Music Ontology ()
1. Introduction
The ontology of music, a field situated at the intersection of philosophy, musicology, and aesthetics, confronts foundational questions about nature and existence of musical works. Contemporary scholarship in this area increasingly reveals a significant paradigm shift in progress: a move away from traditional metaphysical inquiries into what a musical work is toward a focus on methodological and phenomenological questions concerning how musical works are engaged with through human practice. This paper aims to illuminate this intellectual trajectory by surveying five critical issues in contemporary musical ontology, positioning them not as discrete debates but as key moments in this overarching development.
The analysis begins at the traditional starting point of ontological debate: the opposition between Musical Platonism and its critics. Centered on the seminal arguments of Peter Kivy, who posits works as eternal sound-structure types awaiting discovery (Kivy, 1983), and Jerrold Levinson, who counters with the historically situated concept of “indicated structures” (Levinson, 1980), this section outlines the foundational tension between music as a timeless abstraction and a created artifact. Following this, the second issue explores attempts to reconcile Platonism and Aristotelianism. By decoupling a work’s “existence conditions” from its “identity conditions,” theorists such as Nemesio Gracia-Carril Puy have opened the door for a “historically attuned Platonism” (Puy, 2020), challenging the rigid dichotomy that long defined the debate.
The paper then argues that the third issue, the methodological schism between descriptive and revisionary approaches, functions as the fundamental axis for all other ontological disputes (Kania, 2008; Stecker, 2009). Whereas revisionists like Goodman (1968) prioritize theoretical coherence, even if it contradicts common intuition, descriptivists such as Thomasson (2004) insist that any viable ontology must remain faithful to the actual practices of musicians and listeners. The subsequent issues demonstrate how external pressures force this methodological question to the forefront. The fourth issue examines the impact of recording technology. The arguments of Gracyk (1996) and Davies (2001), particularly concerning rock music, reveal how technological mediation destabilizes classical ontological models centered on scores and live performance, thereby demanding genre-sensitive and practice-focused frameworks. Finally, the fifth issue addresses the turn toward phenomenological and imaginative experience. Scholars like Walton (1994) and Bourbon (2018) shift the focus away from the work’s static properties entirely, relocating music’s ontology within the listener’s active and imaginative engagement.
By tracing this path through these five issues, this paper contends that the trajectory of musical ontology is not a random series of debates but a coherent movement from abstract entities toward human practices. This intellectual journey ultimately leads to a crucial contemporary question: in an era of AI-generated music that dissolves traditional notions of authorship and creation, does an ontology focused merely on “what a work is” retain its relevance? Or, as Aaron Ridley’s critique suggests, does such a project risk becoming detached from the very aesthetic experiences that give music its value? (Ridley, 2003) This survey aims to provide a clear map of the landscape that gives rise to these urgent questions.
2. Issue 1: Musical Platonism and the Challenge of Creation
The foundational debate in contemporary music ontology centers on the metaphysical status of the musical work as an abstract object. This discourse has been largely shaped by the tension between Musical Platonism, which posits that musical works are eternal, independent entities discovered by composers, and various creationist theories insist that works are brought into existence through historical acts of composition. This initial issue serves as the quintessential example of the traditional metaphysical project, the limitations of which motivate the methodological and phenomenological turns examined later in this paper.
The most uncompromising defense of Musical Platonism, articulated by Kivy (1983) and rigorously developed by Dodd (2007), identifies the musical work with a “pure sound structure”, an eternally existing type. According to this “simple view,” a work’s identity is determined solely by its acoustic properties. Individual performances are merely tokens of this pre-existing, abstract type (Dodd, 2007). A necessary and controversial consequence of this position is that composers do not create works ex nihilo; rather, they engage in an act of “creative discovery,” identifying and selecting a specific sound structure from an infinite realm of possibilities (Dodd, 2007). For the strict Platonist, Beethoven did not bring the Archduke Trio into existence; he discovered it.
While this model elegantly accounts for a work’s repeatability across multiple performances, its “discovery” thesis sharply conflicts with the deeply entrenched intuition that art is a product of human creation (Levinson, 1980; Predelli, 2001). To fully appreciate the debate, however, one must understand the deep philosophical motivation behind Platonism, which extends beyond mere metaphysical preference. For thinkers like Dodd, Platonism is the most robust way to secure the objectivity of a work’s aesthetic properties. If a work is an eternal, unchanging type, then its properties (elegance, coherence, sublimity) are also fixed and objective, independent of any particular performance or historical interpretation. This provides a strong defense against relativism, grounding aesthetic judgment in the stable features of an abstract object. Thus, the “discovery” model is not merely a counter-intuitive byproduct of the theory; it is a necessary feature to guarantee the kind of objectivity that Platonists believe is essential for meaningful aesthetic discourse (Dodd, 2007). The price of creation, in their view, is the loss of this stable aesthetic foundation.
In response, Jerrold Levinson proposed a highly influential alternative: a modified Platonism that reincorporates the act of creation. Levinson argues that a musical work is not a pure sound structure, but an “indicated structure”, specifically, a sound-and-performing-means structure (S/PM structure) historically tethered to a particular composer at a particular time (Levinson, 1980). On this view, the abstract sound structure may pre-exist, but the work itself, the structure-as-indicated-by-Beethoven-in-1811, is brought into existence by the composer’s intentional act of “indication” (Levinson, 1980). Levinson’s theory thus laudably attempts to historicize the artwork and restore creative agency to the composer, offering a clear corrective to the ahistorical nature of pure Platonism.
However, while Levinson’s theory of “indicated structures” elegantly solves the creation problem, its own ontological underpinnings are less secure than they appear. As Simon Evnine’s critique suggests, viewing the work as a “structure-as-indicated” risks creating an ad-hoc entity metaphysically analogous to what Kit Fine calls a “qua object” (Evnine, 2009). This approach can lead to a problematic proliferation of entities, creating ambiguity as to whether the work is the structure-as-indicated-by-Beethoven or the structure-as-indicated-by-Beethoven-in-1811, without offering a robust theory of constitution that explains how the abstract sound structure actually makes up the work (Evnine, 2009). The theory seems to stipulate a new kind of object into existence to fit our intuitions, rather than explaining the relationship between pre-existing materials and the created artifact. This impasse within the metaphysical debate highlights its limitations and motivates not merely a search for a better theory, but a critical re-examination of the foundational assumptions, specifically the rigid binary of Platonism versus Aristotelianism, upon which the entire debate rests.
3. Issue 2: Deconstructing the Platonic-Aristotelian Binary
Whereas the previous section detailed the specific theoretical impasse between Kivy and Levinson, this section takes a step back to analyze the broader philosophical architecture of the Platonist-Aristotelian debate itself. The traditional framing of this debate presents a stark choice: are musical works eternal templates awaiting discovery, as a strict Platonist like Dodd (2007) might contend, or are they historically contingent entities forged by human practice, as argued by Aristotelians like Wolterstorff (1975)? However, this established dichotomy crumbles under closer scrutiny. Following the critical insights of Puy (2020), this section argues that the received view, which rigidly aligns Platonism with noncontextualism and Aristotelianism with contextualism, is based on a false premise. By decoupling a work’s existence conditions from its identity conditions, we can demonstrate that Platonism can indeed be contextualized, and Aristotelianism decontextualized, thereby dissolving the binary.
Platonism’s most challenging claim is that musical works pre-exist their composition, existing eternally as abstract types (Dodd, 2007). This radical independence from culture and history seems to commit Platonism to a noncontextualist account of a work’s identity, where its essence is defined purely by its sonic structure. However, as Puy (2020) persuasively argues, one can accept that a type exists eternally without committing to the idea that its identity conditions are ahistorical. A work’s identity can be fixed not just by its intrinsic sonic properties but also by relational properties, such as its relationship to a composer or historical context. A Platonic type might pre-exist its discovery, yet remain individuated via these context-sensitive relations. Thus, a “historically attuned Platonism” becomes possible: works persist as structural invariants, but their ontological identity as specific artworks is fixed by the historical act of indication.
Conversely, Aristotelianism’s insistence on historical embeddedness has been traditionally linked to contextualism. Wolterstorff’s influential theory of “norm-kinds,” for example, grounds a work’s identity in the socially established conventions for its correct performance, making the work contingent on its historical praxis (Wolterstorff, 1975). However, this position risks collapsing the work into its historical contingencies, making it difficult to account for the work’s enduring identity across changing performance practices. A more sophisticated Aristotelianism can avoid this by reframing types as emergent entities. Drawing on concepts like Simon Evnine’s “constituted objects” (Evnine, 2009), we can propose that a work attains its full ontological status only through cumulative human engagement. Here, context is not a fixed constraint but a constitutive act; each performance actualizes latent potentials within the sonic structure. The work’s identity is thus forged dialectically between the composer’s “indicative act” (Levinson, 1980) and the socio-historical forces that shape its reception and continued practice (Evnine, 2009).
The critical breakthrough, therefore, lies in recognizing that the Platonist-Aristotelian debate is fundamentally about existence conditions (Are works eternal or created?), not necessarily about identity conditions (What defines a work?). A Platonic (eternal) type can have contextual identity conditions, just as an Aristotelian (created) type might, in principle, have purely sonic, noncontextual identity conditions (Puy, 2020). This decoupling dissolves the rigid opposition. The divide between abstraction and embodiment becomes a false choice, replaced by a more nuanced understanding of how structural invariants and historical practices interact to forge the complex entities we call musical works.
4. Issue 3: The Methodological Core: Descriptivism vs. Revisionism
The ontological impasses discussed in the preceding sections, such as the deadlock between Platonism and its creationist critics, do not merely reflect differing metaphysical commitments. They expose a more fundamental, methodological schism that runs through the entire philosophy of art: the conflict between descriptive and revisionary approaches to ontology (Stecker, 2009; Kania, 2008). This debate is not simply another issue among the five; it is the foundational axis around which the others revolve. It forces a choice about the very purpose of our inquiry: should the ontology of music aim to accurately describe our concepts and practices as they exist, or should it seek to correct and replace them with a more metaphysically rigorous framework?
The descriptive approach, championed by philosophers like Thomasson (2004), argues that since artworks are cultural artifacts, their nature is determined by the beliefs, intentions, and practices of the communities that create and engage with them. On this view, the task of the ontologist is not to discover a mind-independent reality, but to unearth and systematize the assumptions about existence, identity, and persistence that are already embedded in our artistic world (Thomasson, 2004). As Stecker (2009) frames it, this methodology adheres to a “pragmatic constraint” where ontological theories are assessed by their fidelity to the ways people interact with and value music. To a descriptivist, a theory is successful if it explains our practices, not if it proves them wrong.
In stark contrast, the revisionary approach inherits the ambitions of traditional metaphysics. Proponents like Goodman (1968) and more recent Platonists like Dodd (2007) argue that the philosopher’s task is to construct a theory that is logically consistent and metaphysically “clean,” even if it requires us to abandon our most cherished pre-theoretical beliefs. Goodman’s infamous conclusion that a performance with a single wrong note is not a genuine instance of the work at all is the classic example of this impulse (Goodman, 1968). For the revisionist, if common sense conflicts with a more parsimonious or powerful metaphysical system, it is common sense that must give way.
However, the descriptive project is not without its own profound challenges. As critics like Kania (2008) have pointed out, a purely descriptive approach faces a critical problem of normativity. If our practices are inconsistent, confused, or based on factually incorrect beliefs, should an ontology simply reproduce these errors? Whose practices count as authoritative: those of elite composers, professional performers, critics, or the casual listener? Descriptivism risks collapsing into a form of sociological relativism, where the “ontology” of a work changes with every community’s beliefs. Furthermore, it struggles to account for how our artistic practices can be subject to rational critique or reform. If practice is the ultimate arbiter, on what grounds can we ever say that a particular way of treating or understanding a work is wrong? This challenge, how to derive normative standards from descriptive facts, is a central, unresolved tension within the descriptivist program, suggesting that a successful ontology may need to find a middle ground that respects practice without being enslaved by its every inconsistency (Kania, 2008).
Despite these challenges, I contend that a descriptivist methodology offers the more productive path forward for the ontology of music. The counter-intuitive consequences of revisionism, such as the Platonist denial of creation or Goodman’s “wrong note” paradox, arise precisely from the attempt to force music into pre-existing metaphysical categories (e.g., eternal types) that are ill-suited to capture the contingencies of a human artistic practice. By prioritizing fidelity to the concepts of musicians and listeners, a descriptive approach remains anchored in the lived experience that gives music its value. This commitment does not eliminate philosophical problems, but it ensures that the problems we seek to solve are those that matter to our understanding of music as a cultural and aesthetic phenomenon, rather than artifacts of an unrelated metaphysical agenda. While Stecker (2009) is correct that the choice between these approaches is ultimately a pragmatic one, the pragmatics of aesthetics should surely favor a philosophy that illuminates our artistic world rather than one that declares it to be in error.
This methodological preference for practice-based accounts provides a powerful lens through which to analyze the remaining issues. As we will now see, the destabilizing impact of recording technology is best understood not as a problem for an abstract theory, but as a concrete evolution of musical practice that our ontological concepts must strive to catch up with.
5. Issue 4: The Technological Challenge: Performance,
Recording, and the Fracturing of the Work
While the debates between Platonists and Aristotelians revolved around the abstract nature of the musical work, the technological shifts of the 20th century posed a more concrete challenge, forcing a crisis in the traditional ontology of music. The advent of recording technology and the studio-centric practices of genres like rock and electronic music called the very identity of the musical work into question. This section examines how this technological disruption destabilizes classical ontological assumptions, revealing the deep dependence of ontology on practice and pushing the discourse toward either a pluralistic or a radically skeptical conclusion.
Traditional musical ontology, developed primarily with the Western classical tradition in mind, is fundamentally a work-for-performance model. On this view, the work is an abstract entity, a type or kind, specified by a score and made audible through its various instantiations in performance (Kivy, 1983). Recordings, within this framework, are ontologically secondary: they are merely reproductions of performances, which are themselves instantiations of the primary work. The work, therefore, precedes both its performance and its recording.
This established hierarchy was decisively inverted by the practices of rock and other studio-based genres. As Theodor Gracyk argues, in rock music, the recording is not a reproduction of a prior work; it is the primary work. The final master recording, itself a product of composition, performance, and technological manipulation such as overdubbing and mixing, becomes the primary text from which all other instances (LPs, CDs, digital files) are copied (Gracyk, 1996). For Gracyk, this reorients the ontology of rock around a new kind of object: the track, which is a repeatable sound structure embodied in a master recording. Live performances of rock songs, in this view, are often attempts to reproduce the canonical recorded version, reversing the classical relationship.
The shift from analog to digital recording and distribution further complicates this picture. An analog recording, such as a vinyl LP, is a physical trace, a continuous inscription of sound waves. In contrast, a digital file (like an MP3 or a stream from a service) is not a physical trace but a set of abstract instructions, a sequence of code for reconstructing a sound event. This dematerialization introduces another layer of abstraction between the listener and the supposed work. Does the work now reside in the algorithmic code of the file, the specific data packet transmitted, or the resulting sound waves generated by a digital-to-analog converter? This “digital object” is ontologically distinct from both the score and the analog recording, possessing a fluidity and manipulability that challenge traditional notions of a fixed, stable work. The rise of streaming services, where the musical object is never possessed by the listener but is accessed fleetingly, arguably completes this process of dematerialization, pushing the ontology of music even further away from stable objects and toward ephemeral processes of access and generation.
In an attempt to preserve a performance-centric model, Davies (2001; 2003) offers a nuanced counterargument. While acknowledging the centrality of the studio, Davies contends that rock music is still fundamentally a performance art. He re-characterizes the creation of a recording as a “studio-performative” act, a distinct form of performance intended for the medium of recording rather than a live audience. For Davies, the work is a thin entity specified by the song, and both studio recordings and live concerts are distinct, equally valid performances of it. This perspective seeks to adapt the traditional work/performance distinction to the realities of studio production, rather than abandoning it entirely.
I argue that the Gracyk-Davies debate does more than just challenge classical music-centric models; it demonstrates the contingency of ontology on technology and practice. The rise of recording does not simply replace one ontological object (the score) with another (the recording). Instead, it fractures the very concept of “the work,” suggesting that a single, unified ontology of music is no longer tenable. This forces a choice between two radical conclusions. The first is to embrace ontological pluralism: to accept that “musical work” is a genre-relative concept, meaning one thing in a classical context (e.g., a norm-type specified by a score) and something entirely different in a rock context (e.g., a track-type specified by a master recording). The second, more skeptical conclusion is that this technological fragmentation lends powerful support to an anti-ontological stance, as championed by Ridley (2003). For Ridley, the fact that our ontological models are so easily broken by shifts in practice suggests that the models themselves are philosophical artifacts, detached from the actual aesthetic criteria we use to appreciate music in its various forms. In this light, the endless debate over the “true” location of the work becomes an idle distraction.
Ultimately, the challenge of recording technology reinforces this paper’s central thesis. It forces a crisis in the traditional search for a single, stable metaphysical object, pushing the conversation away from what the work is and toward a descriptivist analysis of what artists, producers, and listeners do. This turn to practice and experience provides the perfect entry point for our final issue: the phenomenological account of music, which locates ontology not in an object at all, but in the act of listening itself.
6. Issue 5: The Phenomenological Turn: Music as Imaginative
Experience
If the metaphysical debates over the identity of the musical work have reached an impasse, and if technological and cultural practices continually challenge our ability to locate a stable ontological object, then perhaps the entire object-focused inquiry is misguided. The fundamental weakness of any object-based ontology, whether Platonic or creationist, is its necessary commitment to a set of fixed identity conditions, a core set of essential properties that define what the work is. However, the history of musical interpretation and adaptation demonstrates that the musical work is characterized by a profound “modal flexibility” (Rohrbaugh, 2003). Properties once considered essential, from specific instrumentation and harmonic language to rhythmic structure, are constantly being challenged and altered in performances, arrangements, and new versions that are still legitimately connected to the original work. This suggests that the musical work is not a static entity with a stable essence, but rather an entity whose boundaries are porous and perpetually renegotiated through practice. The object itself appears to be ontologically unstable. This inherent instability of the musical “object” provides the strongest impetus for a phenomenological turn. If no fixed set of properties can definitively ground the work’s identity, then the only stable ground for a philosophy of music may lie not in the object itself, but in the structured activity of engaging with it.
The final issue we will consider, therefore, represents a radical turn away from metaphysics and toward experience, relocating the foundation of music’s ontology from the work itself to the listener’s active engagement with it. This phenomenological approach, by reframing the central question of the field, offers a compelling path forward. A crucial step in this direction can be found in Walton’s work (Walton, 1994). Walton argues that even “absolute” music is representational, functioning as a “prop” in a game of make-believe that prompts listeners to engage their imaginations. In his view, we hear melodies as “rising” and “falling,” we sense “tension” and “release,” and we perceive the presence of anonymous, fictive agents who seem to struggle, strive, or express emotion. These imaginings are not optional additions but are integral to a proper musical experience. While Walton’s theory still treats the musical work as a “work world,” an object that prescribes these imaginings, it marks a significant shift in focus from the work’s intrinsic properties to the dynamic experience it generates in the listener.
Bourbon (2018) takes this phenomenological turn to its logical conclusion. In her essay “Beyond Musical Metaphysics,” Bourbon argues that music is not distinguished from mere sound by its special metaphysical status or psychological effects, but by the specific way in which we listen to it. She characterizes musical listening as an active and intentional practice of recognizing and attending to patterns of sound as an end in itself. For Bourbon, music is not an object that we experience, but an experience that we constitute through a particular kind of active attention. The distinction between sound and music, therefore, is not ontological but phenomenological; it resides not in the object but in the activity of “listening with imagination”.
I posit that this phenomenological reframing offers the most compelling resolution to the problems that have plagued the ontology of music. By shifting the central question from “What is a musical work?” to “What distinguishes the act of listening to music from merely hearing sound?”, this approach effectively sidesteps the irresolvable dilemmas of abstract objects. It is no longer necessary to choose between eternal discovery (Kivy, Dodd) and metaphysically suspect creation (Levinson, Evnine), because the “work” is not a pre-existing object at all but is continuously brought into being through the act of engaged listening. This listener-centric model effortlessly accommodates the diversity of musical practices, from classical and rock to improvisation, as different contexts simply foster different modes of listening.
This perspective does not render the musical object irrelevant, but it redefines its role. The score, recordings, or performances are not the work itself, but rather the catalysts or prompts for the imaginative activity that constitutes the musical experience. The ontology of music, in this view, is ultimately an ontology of a specific kind of human practice. This conclusion leads directly to the final challenge for the field, one that makes this phenomenological turn all the more urgent: the rise of posthuman creativity.
7. Epilogue: Posthuman Creativity and the Future of Musical
Ontology
For over half a century, musical ontology has oscillated between metaphysical abstraction and contextual embeddedness. Yet, the recent emergence of AI-generated music poses a challenge that is not merely additive, but transformative. It serves as a stress test for the entire field, threatening to render its foundational debates obsolete and forcing the paradigm shift from object to practice to its logical conclusion. As Ridley (2003) questioned the ultimate value of determining a work’s identity conditions, the rise of algorithmic creativity compels us to ask what, if anything, is left for musical ontology to do.
The challenge of AI music lies in a profound paradox. On one hand, the process of algorithmic composition offers a powerful, contemporary vindication of Platonism. A neural network trained on vast datasets of existing music navigates a latent space of sonic possibilities, “discovering” novel structures in a manner strikingly analogous to the creative discovery model championed by Kivy (1983) and Dodd (2007). This automated exploration of an abstract sound-space directly challenges contextualist theories like Levinson’s (Levinson, 1980), as the “author” is an ahistorical algorithm and the “act of indication” is a process of statistical inference, devoid of the human intentionality and cultural specificity that grounded the creationist model.
On the other hand, this new form of Platonic discovery simultaneously deconstructs the very humanistic concepts—authorship, artistry, genius, and expressive authenticity—that made the traditional debates meaningful. An AI-generated “Bach chorale” belongs neither to Bach nor to its programmers; it challenges the very notion of creative lineage that is central to our musical practices (Kania, 2008). This crisis validates the anti-ontological skepticism of critics like Ridley (2003) and complicates the work of descriptivists like Thomasson (2004), as our practices for engaging with AI-generated music are still nascent and unstable.
This paradox that AI is Platonic in mechanism but deconstructive in practice signals the exhaustion of ontologies focused on static, pre-defined objects. It pushes us beyond a simple “phenomenological turn” and towards what might be called a generative ontology. In this view, the musical object is not something that is constituted prior to our engagement (whether as an eternal type or an indicated structure), nor is it merely a projection of our listening (as a simple phenomenology might suggest). Rather, the musical object is co-constituted in a dynamic loop of interaction between sounding structures and engaged listeners. It is not a thing, but a process of becoming. AI music becomes the paradigmatic case for this model: the algorithmic system generates sonic potential, and the human listener, through acts of selection, interpretation, and imaginative engagement, actualizes this potential into a determinate musical experience. The “work” exists only within this generative feedback loop.
Therefore, the essential question for a posthuman ontology is no longer “What is the musical work?” but “Through which interactive processes does music come into being?” This shifts the focus from identity conditions to the conditions of emergence. It aligns the philosophy of music with contemporary thought in fields like enactivism and affordance theory, which understand meaning as arising from the coupling of an organism and its environment. In this framework, the algorithmic output offers sonic “affordances” that the listener takes up to create musical meaning. The future of the discipline, then, lies not in classifying the outputs of human-algorithmic symbiosis, but in investigating the rich and complex practices of co-creation through which music’s being is continually, and generatively, forged.
Funding
This research was funded by the 2023 Scientific Research Project “Analytical Aesthetics and Postmodern Musical Ontology Research” of Shanghai Conservatory of Music (2023XJ1D05).