Existence and Non-Existence

"There are things that do not exist" reads like a contradiction, and one way of taking it is exactly that. Another way makes it the plainest thing in the world: Pegasus is one of them, Sherlock Holmes is another, and the round square is a third. Which reading is correct is not a matter of being careful with words. It is the question of whether the world's contents divide into the existing and the rest.

What the position has to answer to

Three sorts of fact push in the same direction, and any account has to handle all of them.

True denials come first. "Pegasus does not exist" is true. Whatever else is said, that has to come out true, and it has to be true in a way that explains why "Bucephalus does not exist" is false. The two sentences differ in truth value, so something is distinguishing them.

Directedness comes second. When someone searches for the fountain of youth, the searching is directed at something rather than at nothing. Two people searching for different nonexistent things are doing different things, and the difference has to be located somewhere.

Fiction comes third. Holmes lived in Baker Street and Poirot did not. It is possible to be wrong about Holmes, which suggests there is something to be wrong about, and it is possible to compare him with Poirot, which suggests there are two of something to compare.

Taking the appearances at face value

The proposal is that the three facts above are exactly what they look like. There is something the sentence "Pegasus does not exist" is about, and what it says about that thing is that it does not exist. There is something the searcher is looking for, and it happens not to be there to be found.

The engine of the view is a principle of independence: what a thing is like is independent of whether it is. The golden mountain is golden and is a mountain, and it fails to exist, and there is no tension between these because having properties was never the same as existing.

Meinong pushed the principle further than most of his readers were willing to follow. The round square, on his account, literally is round and is square. He held that the law of contradiction governs what exists rather than objects as such, so an object can be impossible without that impossibility being a reason to deny that it is an object.

Russell's attack

Bertrand Russell pressed two objections that shaped everything written since.

The first is that the theory admits contradictions into the world. The round square is round, and being square it is not round, so there is an object that both has and lacks a property. Meinong's restriction of the law of contradiction to existent objects looks less like an answer than a restatement of the problem.

The second is subtler and turns the theory's own principle against it. If existence is a property, and objects have whatever properties their descriptions specify, consider the existent golden mountain. By the theory's own rules that object is golden, is a mountain, and is existent. So it exists, and it does not, and no empirical fact was consulted either way.

The alternative

The mainstream response is to deny that any of the three facts requires a nonexistent object.

For the true denials, the analysis from the ontology tier does the work. "Pegasus does not exist" says that nothing satisfies a certain condition, and this is a claim about the condition rather than about a thing. Russell's theory of descriptions supplies the details, taking the apparent name and treating it as a description, so that no object has to be referred to for the sentence to be true.

For directedness, the response is that the state of the searcher can be fully characterised without there being anything at the far end of it. Searching for the fountain of youth is a matter of how the searcher's mind is configured, and two searchers looking for different nothings differ in their configurations rather than in their targets.

For fiction, the leading response is to accept entities of a kind, but not Meinong's kind. Fictional characters, on this treatment, exist as artefacts created by their authors, in the way symphonies and legal systems do. Holmes exists in that sense, and the sentence "Holmes does not exist" is then read as denying that he is a person rather than denying that he is anything at all.

Whether that treatment is a solution or a relabelling is a fair question, and the intermediate tier presses it.

What is actually at stake

Both sides agree on every observable fact. What divides them is where the complexity goes.

Take the appearances at face value and the ontology becomes very large, containing objects for every describable condition, including inconsistent ones, with a corresponding difficulty about which of them are identical to which. Refuse them and the ontology stays small, at the cost of an elaborate account of how ordinary sentences that look like they are about things are really about something else.

Neither cost is obviously worse, which is why the dispute is still open more than a century after Russell thought he had closed it. What the dispute demonstrates is the standard shape of a metaphysical argument: not a disagreement about the data, but a disagreement about which theory pays the more acceptable price for accommodating it.

The next article steps back from existence to a division that runs through everything after it, between what is in space and time and what is not.