What Kind of Thing Is a Work of Art?

Imagine a painting of a bowl of fruit. Not just any painting—one you’ve seen in a museum, say a Cézanne. Now imagine a perfect copy of that painting, so good that even an expert couldn’t tell them apart by looking. The two canvases are visually identical. Are they the same work of art? Most people would say no. The original is by Cézanne, made in his studio in the 1890s. The copy is by someone else, made last year. They look the same, but they seem to be different works.

But then what is the work of art? It can’t just be the painted canvas, because two different canvases could look identical and still be different works. So what kind of thing is it?

This question—the ontology of art—isn’t asked by most people who enjoy paintings or listen to music or read novels. But once you start thinking about it, it gets weird fast. Philosophers have been arguing about it for a long time, and they still don’t agree.

Do Works of Art Even Exist?

That sounds like a crazy question. Of course works of art exist. There’s a copy of Hamlet on my shelf. There’s a sculpture in the park. But some philosophers have pointed out that it’s not so simple.

Here’s one problem: You can hold a book in your hands. But is the book Hamlet, or is Hamlet something else that the book contains or expresses? If all copies of Hamlet were destroyed, the play wouldn’t exist anymore—except that people know it, and could write it down again. So the play isn’t exactly the same as any particular copy. It’s something that can have many copies.

Or take Beethoven’s Fifth Symphony. You’ve probably heard it. But where is it? It’s not in the score—the score is just marks on paper. It’s not in any particular performance—each performance is slightly different. And it’s not in your head—other people can hear it too. So what is it?

A philosopher named C. E. M. Joad suggested in the 1930s that works of art are “subsistent” objects—they have a kind of being that isn’t physical or mental, but something else entirely. They’re like numbers in that sense: the number 7 exists, but you can’t point to it anywhere in space.

Other philosophers have gone in the opposite direction. Jean-Paul Sartre, the French existentialist, said that a symphony doesn’t really exist in the normal sense at all. We imagine it. When we listen to a performance, we’re not really hearing the symphony—we’re hearing sounds and imagining the symphony in them. The work is a kind of fiction.

Nicolai Hartmann, writing around the same time, developed this idea more carefully. He said works of art have two layers. The first layer is the physical stuff—paint on canvas, sounds in the air, marks on paper. You perceive that with your senses. The second layer is what you imagine through that physical stuff—the story, the emotion, the beauty. The work of art is both layers together. And the second layer only exists when someone is paying attention in the right way.

This leads to a strange possibility. If nobody is looking at the painting, does it stop being a work of art? Some philosophers have said yes. The painting becomes just paint on canvas until someone looks at it with the right attitude. Other philosophers find this absurd. Surely the Mona Lisa was a work of art even when it was locked in a closet.

One Category or Many?

Most people assume that paintings, symphonies, poems, and dances are all “works of art,” so they must all belong to the same basic category of thing. But many philosophers think this is wrong.

Think about the difference between a painting and a symphony. You can perform a symphony many times, with different orchestras, in different halls. Each performance is a separate event, but they’re all performances of the same work. A painting, on the other hand, seems to be a single physical object. There’s only one original Mona Lisa. Copies exist, but they’re considered copies, not the work itself.

This suggests that different kinds of art might be different kinds of things. A symphony might be something like a “type”—an abstract pattern that can have many “tokens” (the individual performances). A painting, on the other hand, seems to be a particular physical object.

But even this isn’t settled. Some philosophers argue that all works of art are multiple, at least in principle. If we could make perfect copies of paintings, they say, those copies would be just as much the work as the original. Others insist that paintings are essentially unique—that’s part of what makes them paintings.

And then there’s the strange case of photography. A photograph can be reproduced endlessly. Is each print the work? Or is the negative the work? Or is the work something more abstract, like the image itself? Henri Cartier-Bresson’s famous photograph “Behind the Gare Saint-Lazare” exists in many prints. They’re all considered genuine. So photography seems to be a “multiple” art form, like music.

The Pierre Menard Puzzle

In 1939, the Argentine writer Jorge Luis Borges published a short story that has fascinated philosophers ever since. It’s about a French symbolist poet named Pierre Menard who sets out to write Don Quixote—not by copying Cervantes’s novel, but by writing it himself, from scratch, so that his text turns out to be verbally identical to Cervantes’s.

The story’s narrator claims that Menard’s Quixote would be a different work from Cervantes’s, even though the words are the same. Why? Because they were written in different centuries, by different authors, with different backgrounds and intentions. The same words mean something different when they come from a 17th-century Spanish writer than when they come from a 20th-century French symbolist.

This suggests that the identity of a work of art depends not just on its text or structure, but on its history—who made it, when, and where. A string of words is just a string of words. A work of literature is something more, tied to its origin.

The philosopher Jerrold Levinson took this idea and built a theory around it. He said that a musical work (at least a classical composition) isn’t just a pattern of sounds. It’s a pattern of sounds as indicated by a particular composer at a particular time. This makes the work a kind of “initiated type”—an abstract pattern that comes into existence through someone’s act of indicating it.

This would explain why two composers who independently come up with the same melody haven’t written the same piece. The melody might be the same pattern, but the works are different because they were created by different people in different contexts.

Are Works of Art Abstract or Concrete?

This might sound like a technicality, but it leads to a real puzzle. Consider these three statements:

Works of art are created.
Works of art are abstract objects.
Abstract objects cannot be created.

Each of these seems true. But they can’t all be true together. So something has to give.

Some philosophers give up on statement 1. They say composers don’t really create musical works—they discover them, like mathematicians discover numbers. The work was always there, waiting to be found. This saves statements 2 and 3, but it seems to deny something important about what artists do. When Beethoven wrote the Fifth Symphony, he wasn’t just discovering something that already existed. He was making something new.

Other philosophers give up on statement 2. They say works of art aren’t really abstract. They’re concrete things—physical objects, events, or performances. But this seems to have trouble explaining how one work can have many performances.

Still others revise statement 3. Some philosophers argue that abstract objects can be created. The American philosopher Amie Thomasson has argued that works of art are a special kind of “abstract artifact”—they’re not physical, but they’re made by humans at particular times and places. They depend on human activity to exist, but once they exist, they’re not located in any particular spot.

The Performance Theory

Gregory Currie took a different approach. He said that works of art aren’t objects at all—they’re actions. More specifically, they’re types of actions.

According to Currie, a work of art is something like: an artist discovering a structure through a particular process. The painting on the wall isn’t the work. The work is the activity that produced it—all those decisions and movements and choices that the artist made.

This theory has a strange consequence: you can’t actually read a novel or see a painting. What you can do is experience a token of the action type that is the work. But the work itself is something you can never directly encounter—it’s an event in the past, or a type of event.

David Davies developed this idea further, arguing that works of art are not action types but action tokens—particular, individual performances that happened at a specific time and place. The work of art is the artist’s “generative performance,” the actual doing that brought the artistic vehicle into being.

This would mean that the Mona Lisa isn’t the painting in the Louvre. The Mona Lisa is Leonardo da Vinci’s act of painting it—something that happened in the early 1500s and is now over. What we see in the museum is just the trace of that act.

Most people find this highly counterintuitive. “You cannot hang an event on a wall, only a picture,” as one critic put it. But Davies has a response: philosophy sometimes has to revise our ordinary ways of thinking if those ways are confused.

Why Should You Care?

This all might seem like philosopher’s nitpicking. But the questions matter for how we treat art in real life.

If a perfect copy of a painting is not the same work, then forgeries matter even when they’re visually identical. The value of an artwork isn’t just in how it looks—it’s in its history. This affects everything from museum curation to art prices.

If musical works are types that can have many tokens, then what counts as a genuine performance? How different can a performance be before it stops being the same piece? This matters for musicians, conductors, and critics.

And if works of art are created things that depend on human activity, then they can also be destroyed. A lost film or a burned manuscript isn’t just a lost copy—it’s the loss of the work itself. This gives weight to preservation efforts.

These are not just academic questions. They’re about what we value when we value art, and what we’re trying to protect when we try to protect it.

Appendices

Key Terms

Term	What it does in this debate
Ontology	The study of what kinds of things exist—here, applied to works of art
Type	A kind of abstract entity that can have many concrete copies or “tokens” (like a word type that can be written many times)
Token	A particular concrete instance of a type (like this specific written word you’re reading right now)
Abstract object	Something that doesn’t exist in space or time, like a number or a universal property
Initiated type	A type that is brought into existence by someone’s action, rather than existing eternally
Subsistence	A kind of being that is not physical existence—some philosophers say works of art “subsist” rather than “exist”
Vehicle	The physical stuff (paint, sound, words on paper) that carries or conveys the work of art
Indicated structure	A sound pattern or word sequence that has been “pointed to” by an artist in a particular context

Key People

C. E. M. Joad – A British philosopher who argued that works of art are “subsistent” objects, neither mental nor physical
Jean-Paul Sartre – A French existentialist who said works of art are imaginary, not real—we don’t really hear the symphony, we imagine it
Nicolai Hartmann – A German philosopher who argued that works of art have two layers: a physical foreground and an imaginative background
Jorge Luis Borges – An Argentine writer whose story about Pierre Menard showed that two identical texts could be different works of art
Jerrold Levinson – An American philosopher who said musical works are “indicated structures”—patterns that come into existence when someone points them out
Gregory Currie – A philosopher who argued that works of art are types of actions, not types of objects
David Davies – A philosopher who argued that works of art are particular actions (action tokens), not types at all
Amie Thomasson – An American philosopher who argues that works of art are “abstract artifacts”—not physical, but still created by humans

Things to Think About

If a perfect copy of a painting is not the same work as the original, what exactly is lost when you look at a copy rather than the original? Is it something you can detect, or something you just have to know?
If two composers living in different countries independently write down the exact same sequence of notes, have they written the same piece of music or two different pieces? What would make you say one or the other?
Does a museum owe it to visitors to display only original works, or could a perfect replica serve just as well? What would be lost if we replaced every painting in a museum with an indistinguishable copy?
Some philosophers say works of art don’t exist at all—we just talk as if they do. Could you live your life as an art lover without believing that works of art really exist? What would change?

Where This Shows Up

Art forgery and authentication: When experts determine whether a painting is “really” by the artist it’s attributed to, they’re making judgments about the identity and history of the work, not just its appearance
Music copyright law: Lawsuits about whether one song is a copy of another depend on what counts as “the same” musical work
Museum conservation: Deciding whether to restore an old painting or leave it as it is involves questions about what the work of art is—the original object, or something that has changed over time
Digital art and NFTs: Questions about whether a digital image can be an original work, and what ownership even means for something that can be copied perfectly, are exactly the kind of puzzles philosophers have been discussing