The romantic ideal of originality: a solitary figure elevated above the world, yet standing on ground formed by what came before. Image: Caspar David Friedrich — Wanderer above the Sea of Fog (1818)

 

April 28, 2026: This article has been substantially revised.

Originality is one of those words that feels self-evident until you try to define it. Few concepts are invoked more confidently in art, and few collapse more quickly under closer inspection. Artists are expected to be original. Works are valued for their originality. Legal protection hinges on it. And yet, anyone who has seriously studied or practiced art knows how fragile this idea is.

This tension is structural, and it is one we engage with directly. CORPUS uses the term originality on its contribution platform, fully aware of the frictions it carries. At CORPUS Live, the concept sparked some of the most intense discussions of the day. That was not an accident. We consider the discomfort around originality productive. But to understand why, it helps to look at where this concept comes from, what it was designed to do, and why it keeps breaking down, especially in music.

The romantic invention of originality

The idea that artworks emerge as radically new expressions from an individual creator feels deeply ingrained today. It is the dominant cultural image of artistic production: the solitary author, the inner voice, the unique gesture. But historically, this image is surprisingly young.

What crystallizes in European Romanticism is not creativity as such, but a specific moral framing of authorship. Artistic creation becomes tied to inner authenticity, personality, and genius. The artwork is no longer primarily a contribution to a shared tradition, but an emanation of an individual subject. Originality, in this sense, does not simply describe difference: it becomes a virtue (see Martha Woodmansee, The Genius and the Copyright).

This matters because it coincides with a period in which reproduction becomes economically and politically relevant at scale. Printing presses, publishers, music engravers, and later record industries require a way to assign exclusivity. The romantic author provides exactly that: a culturally compelling figure that makes private ownership of cultural expression appear natural, even necessary (see Mark Rose, Authors and Owners).

Originality, in other words, is not just an aesthetic category. It is a legitimizing narrative.

Copy, not creativity, was the original problem

Copyright did not emerge because society suddenly wanted to reward creativity. It emerged because copying became industrially feasible and economically disruptive.

Early copyright regimes are best understood as systems for regulating reproduction, distribution, and market order (see the Statute of Anne, 1710; historical commentary). The core concern was not whether something was metaphysically new, but who had the right to print, sell, and control copies. The language of “original works” enters later, as a way of stabilizing these regimes by tying rights to authors rather than printers.

This distinction is important, because it reveals a persistent confusion. We tend to read copyright as if it were about artistic value. It is about allocatable control in a system of reproduction (cf. Peter Jaszi, Toward a Theory of Copyright).

The romantic idea of originality helped fuse these domains. It allowed the law to lean on an intuitive moral claim: if a work is the authentic expression of an individual, then copying it without permission feels like a violation of something personal. The legal structure gains cultural plausibility.

But this plausibility comes at a cost.

What artists learn early, law struggles to accept

In serious artistic training, one lesson arrives quickly: absolute originality does not exist. Art is relational, referential, iterative, built from styles that are learned, forms that are inherited, and techniques that circulate.

This understanding is not a postmodern provocation but basic professional literacy, articulated explicitly across movements such as collage, appropriation, Fluxus, and postmodernism (see e.g. writings around Fluxus).

In music, it becomes unavoidable. Techno, hip-hop, sampling, remix culture: these genres are not exceptions to an originality rule but demonstrations that the rule itself is fiction (see Lawrence Lessig, Free Culture).

And yet, copyright continues to operate as if works were cleanly attributable to singular acts of creation.

The resulting gap is what many artists intuitively feel as a kind of violence: the law seems to demand something that artistic practice knows to be false.

Music as a stress test for originality

Nowhere does this contradiction surface more clearly than in music.

Western music operates within a finite system: a limited number of pitches, scales, harmonic functions, rhythmic structures. Musical meaning emerges from patterns, expectations, timbre, performance, and context; groove, voice, sound, and aura matter at least as much as melody or harmony, often more.

Copyright, however, has historically anchored musical originality in what can be abstracted, fixed, and compared: notated melody, harmonic progression, formal structure. The “work” is treated as something that can be reduced to a lead sheet.

This reduction systematically ignores precisely those dimensions of music that listeners often perceive as most distinctive: sound, texture, timing, articulation, performance.

The absurdity of recent plagiarism lawsuits follows directly from this mismatch. Claims are made on the basis of generic chord progressions, short ostinati, or stylistic resemblance, which function as shared musical vocabulary (e.g. Williams v. Gaye; Gray v. Hudson; Skidmore v. Led Zeppelin).

The result is strategic litigation. Copyright becomes a weapon.

Free culture and the attempt to escape the trap

The free culture and copyleft movements of the early 2000s were, in part, a response to this deadlock. They kept authorship; they rejected the assumption that exclusivity must be the default. Instead, they experimented with license systems that lower transaction costs, encourage reuse, and accept derivation as a cultural norm.

Creative Commons emerged from this context as a pragmatic compromise: a way to reprogram copyright from within (see Lessig).

If this movement appears less visible today, that does not mean it failed. Its tools are widely embedded. What has changed is the battleground. Platform economies, streaming, and now machine learning have shifted the locus of extraction and control. The question is no longer only who may copy, but who may learn.

The future of music in machine learning systems will be decided not by whether outputs resemble inputs, but by how we organize participation, attribution, compensation, and access in a world where learning itself has become industrial.

Why CORPUS still uses the term

CORPUS evaluates musical contributions along several axes, including originality (see our White Paper). We do so knowingly, and not without friction.

What we do not mean is artistic originality in the romantic sense: no judgment of genius, authorship, or expressive depth, and no ranking of art. What we mean is novelty relative to the already existing music library.

This is a technical, not an aesthetic, concept. It asks a simple question: does this contribution occupy territory that is currently underrepresented in the dataset? Does it add something that was not already statistically present?

In other words: where are the white spots on the map?

Originality, in this context, is a measure of informational difference. It is a tool for optimizing diversity. And diversity is not a moral stance but a functional requirement. A dataset that collapses stylistic variety will produce brittle, biased models. Rewarding novelty is how we counteract that.

We use the word originality because it is familiar, contested, and uncomfortable. It forces the discussion into the open. But its meaning within CORPUS is deliberately narrow and instrumental.

A concept under pressure: why that is useful

Originality is a story we keep telling, collectively, because it has been useful. It justified ownership. It stabilized markets. It made legal abstractions feel human.

But as soon as we look closely (as artists, as listeners, as technologists), the story frays. Art is relational. Music is collective memory. Human learning is accumulation, not theft.

Recent analyses have begun to describe copyright itself as a coming “stress test” in which originality shifts from a background assumption to a visible point of strain. Music lawyer Raffaella De Santis’s discussion of AI, music, and market power is a notable example.

As machine learning enters cultural production, originality is increasingly used as a stand-in for unresolved questions of access, attribution, and value creation. The resulting pressure is useful: it brings the concept back onto the table as a mechanism to be examined, rather than an ideal to be defended.

The task is to reform how originality functions within legal and technical systems, so that authors’ rights remain meaningful without demanding an impossible standard of absolute novelty. The real risk is clinging to an unchanged concept while asking it to regulate a world it was never designed to govern.