Truth, mention, and use

Truth is correspondence to reality
Truth is correspondence to reality

by Jeremy W Bowman

1. Mention and Use

Philosophers and logicians are familiar with a distinction between mention and use. (Or at least they are supposed to be.) The distinction is crucial for logical and grammatical clarity. It is important — and useful for anyone who deals with words, or cares about truth.

If I hit you over the head with a chair, you would be seriously hurt. But if I hit you over the head with the word 'chair', you wouldn't be hurt at all. A chair is an object — four legs, wooden, etc. — but 'chair' is a word — five letters, ink on paper, etc.

Simple, isn't it? A word usually refers to a thing in reality, and although any word is itself an abstract sort of thing, it's usually entirely different from what it refers to.

Simple and all though that is, we easily get confused. When we talk about words — that is, when words stand for themselves — it sounds the very same as when words stand for things by referring to them. If I were speaking these words instead of writing them, and you were hearing them instead of reading them, you wouldn't be able to tell when I was talking about chairs, and when I was talking about occurrences of the word 'chair'. That is why academics do the “rabbit's ears” gesture with their fingers to show when they are using quotation marks.

We do not have to look silly to be clear about the difference in writing. We simply use inverted commas to mark a bit of language when we are talking about it, and omit inverted commas when we are referring to things.

Words refer to things by default. That is, unless we specifically mark them as having a different purpose, they refer to things. So naturally enough, the way we usually use words is called “use”. But when words stand for themselves rather than referring to things (i.e. when they are flanked by quotation marks) it is called “mention”. We just mention words when we don't mean them to refer to things beyond themselves.

For example, in the second paragraph above, the word 'chair' is used twice and mentioned twice. In the current paragraph, the word 'chair' is mentioned three times but not used at all, because nowhere in this paragraph does the word 'chair' refer to the four-legged wooden things we sit on.

In all of the examples of mention in this essay, I have used single quotation marks to indicate when words stand for words rather than referring to things. However, I have also used double quotation marks to indicate quotations, i.e., bits of language that I want to attribute to someone else. For example, when I introduced the words 'use' and 'mention' above, I was using them rather than mentioning them, but I was also attributing them to other people, namely, the language experts who decide how these terms should be used. So I used them, but marked them with double inverted commas as a gesture towards other people, and the way they used the words.

By using double quotation marks for quotation, I am departing from standard British punctuation, which normally uses single quotation marks for first-level quotation, and double quotation marks for second-level quotation (i.e. quotations nested inside quotations). In this as in many other aspects of English usage, I think the Americans have got it right, and the British could get it better.

We have to distinguish between mention and use, but we also need to distinguish between quotation and mention. If we adopt the British system of using single quotation marks for first-level quotation, that becomes difficult. If we adopt the American system of using double quotation marks for first-level quotation, it's easy, but of course the problem arises again when we have to revert to single quotation marks for second-level quotation. However, second-level quotation is much rarer than first-level quotation, so the problem arises less often in the American system.

2. "The Bewitchment of our Intelligence by means of Language"

So much for punctuation. Many people will be rather surprised to hear that what I've just discussed has a very real bearing on some pretty deep philosophical issues. “Philosophy is a battle against the bewitchment of our intelligence by means of language”, wrote Ludwig Wittgenstein. Let us consider one example of bewitchment by means of language. How often have you heard “postmodernist” types talking about “my reality” as opposed to “your reality”? Or claiming that people from different cultures live in “different worlds”? Obviously, there is at most one world, and we all live in it. The reason why “postmodernists” suppose that people can live in different worlds begins with the perfectly correct thought that different people from different cultures can believe different theories. Various theories talk about various entities (such as ghosts, electrons, etc.). We are never able to talk about anything unless we have a theory of the thing we are talking about, because we cannot talk about things without having reasonably clear ideas about the things we are talking about. The things we talk about are posited (i.e. assumed to exist) by theories. So in a strictly limited sense, all of the entities we can ever talk about “belong to a theory”. But theories do not consist of entities. The assumption that they do is a classic case of the confusion of mention and use.

London is a big city in England, but the word 'London' is the six-letter name of a city. Similarly, an electron is a very small negatively-charged particle, but the word 'electron' is an eight-letter word that refers to very small negatively charged particles. Over the course of the last century or so, a succession of different theories of matter spoke of electrons, and tried to describe accurately how they behave. Before that, no one spoke of electrons because they didn't have theories that described them. But the stark, cold reality out there stayed the very same. Theories consist of laws, terms, etc. (i.e. bits of language), and when people adopt new or different theories, they use different words, and talk slightly different languages. But the reality the languages describe remains the same, regardless of the words people use. To assume that the reality itself changes when a new theory is adopted is a confusion of mention and use, analagous to thinking that the city of London pops into existence as soon as we start using the word 'London', and pops out again as soon as we stop using the word.

3. Quotation and Truth

"What is Truth?" asked Pontius Pilate. The best traditional answer is that truth is correspondence between language and reality. A sentence is true if and only if it corresponds to a fact (i.e. a state of affairs that exists in reality), so that the words that occur in the sentence refer to things, which actually are arranged the way the sentence says they are arranged. (Beliefs are also true or false, but for every belief there is a linguistic sentence that expresses its content, so the truth or falsity of beliefs is closely analogous to the truth or falsity of sentences.) For example, let's take a childishly simple example of something true or false: the sentence 'The cat is on the mat'. As I wrote that sentence down, I had my brother's cat in mind, asleep as usual on his own mat in my brother's house. I'll guess that the sentence is true. It's a long shot, because I don't have much evidence, but that's irrelevant for its truth. What makes it true (if it is true) is a four-legged furry object sitting on top of a flat object, at a remote location, at this moment. If the four-legged furry object is not in fact on top of the flat object at that remote location right now, then the sentence is false. But if it is, then the sentence is true, and my lack of certainty is completely irrelevant.

When a sentence is true, it's a bit like a person being an uncle. When you become an uncle, a special sort of event has to happen to someone else: your sibling has to become a parent. It might happen on the other side of the world, without your knowledge — but those details don't matter. Analogously, when a sentence is true a special sort of fact has to exist somewhere else, usually at an entirely separate location. And your certainty or lack of it has absolutely nothing to do with it.

This admirable "correspondence" idea has never quite lived up to its promise, because different types of sentences tend to make very different claims about how things are arranged. The fact that makes the sentence 'the cat is on the mat' true is easy to visualize, but what about the fact that makes the sentence 'what goes up must come down' true? Like a scientific law, this sentence describes what would follow if something happened — even if it didn't actually happen.

Over the course of the last 50 years or so, philosophers have increasingly come to think that the concept of truth is “a device for semantic ascent”: that is, it enables us to talk about linguistic sentences from a higher or “meta”-level, in which we can express agreement or disagreement with what the sentence says without actually uttering the sentence itself. For example, suppose I want to express the thought that everything Al Gore says is true. I can just utter the words 'everything Al Gore says is true', which only takes a moment, or I can repeat everything Al Gore actually said during the course of his life, which would take years, and it might seem like decades.

If I were to take the second route, it would be very inconvenient, but it would enable me to avoid using the word 'true'. The so-called “redundancy” theory of truth begins with the following idea: the word 'true' is redundant, that is, it can be avoided albeit at the cost of great inconvenience. This is very similar to the traditional “correspondence” theory of truth, only it acknowledges that the correspondence we are looking for differs in kind from one sentence to the next, and acknowledges that any account of it is bound to be expressed in language. (After all, what else would it be expressed in?) The sentence that is claimed to be true expresses the fact that makes it true more clearly than any other sentence. Hence the following famous “disquotational schema”:

'p' is true = p

In effect, this says that whenever a claim is made that a sentence is true, that claim is made at a higher linguistic level that the sentence itself. One can descend from the “meta”-level by simply uttering the sentence itself, omitting the quotation marks and the word 'true'. In other words, the word 'true' is strictly “redundant”.

The classic example goes like this: the sentence 'snow is white' is true if and only if snow is white. If that sounds empty or circular, just look at those quotation marks. The left-hand-side of the equation is saying something about language, a mere representation of reality, but the right-hand-side is saying something about reality itself.

By understanding truth as a means of talking about language from a higher level, as with “mention” described above, this theory sees truth itself as neither more nor less mysterious than all of the myriad ways language can describe the world accurately. Although this may seem empty at first, it is a genuine step forward, because although we do not know everything there is to know about language, we have a pretty good idea of how to go about studying it, and how to find out more about how it works. Before that, our task was to answer a question — 'What is truth?' — that can seem utterly mystifying. Poor old Pilate was not the only one to have been nonplussed by it. If only he had known about mention and use!