What 'Exists' Means

Everyone already uses the word without difficulty. "There is a bus at four" and "there is no evidence for that" are ordinary English, and nobody stumbles over them. The trouble starts when the same word is asked to do the same job for numbers, for holes, for the colour red, and for a character in a novel. Some of those cases feel settled and some do not, and what makes the difference is not obvious from the outside.

The word is also used in a narrower sense by computer scientists, for a formal specification of the categories a database or a knowledge base recognises. That use grew out of this one and is not the topic here.

The answer nobody disputes

Asked in full generality, the question has an answer everyone accepts: everything. It is correct and it is useless, because both sides of any actual disagreement accept it. Someone who denies that numbers exist is not denying that everything exists. They are denying that numbers are among the everything.

So the general question is never where the work happens. The work happens over particular cases, and the cases that generate arguments are fairly predictable. Ordinary middle-sized objects are uncontroversial. Numbers, properties, possibilities, fictional characters, holes, boundaries, corporations, and laws are all contested, and each is contested for its own reasons.

Existing, being real, and being talked about

Three uses of the vocabulary run together in ordinary speech, and separating them removes a good deal of apparent disagreement.

To call something real is usually to contrast it with something else: real as against counterfeit, real as against imagined, real as against merely apparent. The contrast is what carries the meaning, and it shifts. A real diamond is not a zircon. A real problem is not one you invented. Neither use is quite the existence question, and an argument conducted in the vocabulary of realness will often turn out to be about which contrast is in play.

Being available to talk about is weaker still. Sherlock Holmes can be described, argued over, and distinguished from Poirot. All of that is possible without his being among the things there are, in the way the bus at four is. The move from "we can talk about it" to "it exists" is the single most common error in this area, and it has a name in the literature: it is sometimes called Plato's beard, after the tangle of trying to say what it is that does not exist.

Existence claims

Most of the action in ontology happens through a particular kind of sentence.

English hides these in several forms. "There are wolves in Poland" is explicit. "Some wolves are radio-collared" is an existence claim with different clothes. So is "wolves have returned to the Netherlands", where the bare plural does the work. Denials are existence claims too: "there is no largest prime" is a claim about what the world of numbers does not contain.

Learning to spot the pattern matters because ontological disputes are almost always disputes about which existence claims a body of talk actually commits its speakers to. Someone who says the average British family has 1.8 children has not thereby said that there is an average family with a fractional child. The sentence is a compressed report about a ratio, and rewriting it that way removes the appearance of a commitment without changing what is being said.

That manoeuvre is the second of the two moves available in any ontological dispute. The first is to accept the entities and put them on the list. The second is to rewrite the sentences that seemed to demand them, so that the demand goes away. Both moves are legitimate, and which one is better in a given case is exactly what the arguments are about. The article on ontological commitment later in this tier makes the second move precise.

Being

The interchangeable use is the safe default, and this site follows it unless an article says otherwise. Two qualifications are worth having in advance.

Alexius Meinong is the standard reference for the view that the two come apart. On his account there are objects that do not exist, and the golden mountain is one of them: it has properties, it is golden and a mountain, and it simply lacks existence. Whether that is a real distinction or a confusion is a live question, and the metaphysics ladder treats it directly.

Separately, several twentieth-century European philosophers use "being" for something that is not an inventory question at all: the topic is what it is for anything to show up as an entity in the first place. This site is written from an analytic vantage point and does not treat that tradition in the beginner tier. Noting that the same English word covers both projects will save you from reading one as a bad attempt at the other.

What ontology is not

It is not a science of a special extra realm. Nothing in the subject requires a hidden layer of the world that only philosophers can inspect. The questions are about the ordinary world and about what the correct description of it has to include.

It is not decided by looking. Whether numbers exist is not settled by any observation, and this is not because they are hiding. The question is about what the best account of mathematical practice is committed to, and that is an argument about arguments.

It is also not merely verbal, though the suspicion that it might be is a serious position with serious defenders. Rudolf Carnap argued that questions asked from outside a way of speaking have no factual answer, and that what looks like a deep dispute about numbers is really a choice about which language to adopt. The advanced tier takes that up. For now it is enough to know that the objection exists and that the discipline has not simply ignored it.

Where this goes

The next article takes the contested cases one at a time and asks what each would have to be. After that the tier fixes the vocabulary of objects and properties, of identity and counting, and of what a theory commits you to when you assert it. Nothing later assumes that any particular answer to the existence question is correct.