Why Can’t You Just Switch ‘Bachelor’ for ‘Unmarried Man’?
A puzzle inside little marks
You’re reading a sentence: “‘Bachelor’ has eight letters.” It’s true — count them. Next you try: “‘Unmarried man’ has eight letters.” Wait — that’s false? But “bachelor” and “unmarried man” mean exactly the same thing. The only difference is that the words are inside quotation marks. Something about those little hooked lines makes ordinary swapping impossible.
Philosophers of language love this puzzle. Quotation is the name for what happens when we put words inside those marks, or use air quotes, or italicize. It’s a device we use all the time without thinking. Yet it breaks some of the most basic rules of meaning. For example, usually if two words refer to the same thing, you can exchange one for the other without changing the truth of a sentence. But quotation slams the door on that — philosophers call this opacity. You also can’t grab a variable from the outside and put it inside a quotation, the way you can say “something has eight letters” and then ask “what has eight letters?” — because “ ‘x’ ” will always name the 24th letter of the alphabet, not whatever you wanted to point at. And here’s another strange thing: you can quote completely new squiggles. The sentence “ ‘❦’ is not a letter in any language” is true, and you understand it instantly, even though you’ve never seen that symbol before. Those are three of the basic features any theory of quotation must explain.
So what is going on? Over the past century, philosophers have built several families of answers. Each one tries to answer two questions: What part does the referring inside a quotation? and How does it refer? They range from the ultra‑simple to the creatively wild.
The name‑tag idea: quoting as a proper name
One of the earliest suggestions is the Proper Name Theory. Imagine that every time you enclose an expression in quotation marks, you create a brand‑new name for that expression — just like “Paris” is a name for a city. Willard Van Orman Quine (1908–2000) described the idea vividly: from the point of view of logic, a whole quotation is a single solid word, and the parts inside “count for no more than serifs or syllables.” Alfred Tarski (1901–1983) put it even more plainly: a quotation‑mark name is a constant individual name of a definite expression, just like the proper name of a man.
The name theory makes quick sense of opacity. If “‘bachelor’” is just a name — a long, weird‑looking label — then it doesn’t contain the word “bachelor” at all. You can’t swap “Tully” into a name like “Cicero” in the middle of a word, any more than you can swap “cat” for “dog” inside “cattle.” And because the whole quotation is a single lump with no internal structure, you can’t quantify into it — there’s just nothing to reach inside.
The problem is that the name theory can’t explain why quotation feels productive. We don’t learn quotations one by one. The first time you see “‘❦’ is not a letter,” you know exactly what symbol is being talked about. If quotation marks were truly meaningless serifs, coming across a brand‑new quotation would be like hearing a new name with no clue what it names. Some defenders have pushed back. Mario Gómez‑Torrente suggests that some names can be generated and understood automatically — street names like “East 6th Street” tell you which street they refer to by their pattern. Michael Johnson points to onomatopoeia: the word “buzz” sounds like the noise it names. It’s possible that quotations could work the same way, but many philosophers think these clever patches don’t fix the deeper flaw: the name theory severs the intimate tie between a quotation and the expression inside it. “‘Lobster’” feels as though it contains the word “lobster”; the name theory says that’s just an accident of spelling.
The pointing theory: quotation as a finger aimed at words
The most influential paper on quotation in the twentieth century is Donald Davidson’s (1917–2003) “Quotation.” Davidson proposed the Demonstrative Theory, which says that quotation marks are like a finger pointing at a shape. The words sitting between the marks aren’t part of the sentence at all — they are simply displayed, like a sample held up and shown. The real referring device is the quotation marks themselves, and they mean something like “the expression of which this is a token.” So, instead of “‘Bachelor’ has eight letters,” the sentence has a logical form like this: “Bachelor. The expression of which that is a token has eight letters.”
This bold move explains a lot. There’s no upper limit on what you can point at, so you can quote the brand‑new squiggle ‘❦’ without ever having learned its name — you just demonstrate it (or a token of it). Opacity is immediate: pointing at one object doesn’t tell you anything about a different object you point at with another gesture. Mixed quotation, where you both use and mention words at the same time (“Quine said that quotation ‘has a certain anomalous feature’”), becomes natural: you use the words to report what someone said and simultaneously point at the very shapes they used. And quantifying in is impossible, because the quoted material isn’t even inside the sentence logically — there’s nothing to bind.
But critics spotted trouble. If quotation marks introduce a demonstrative, then in theory they ought to be able to demonstrate anything — a nearby penguin, a cloud — yet not a single utterance of “‘bachelor’ has eight letters” can refer to a penguin. A deeper worry is the problem of relevant features. Any squiggle on a page instantiates countless shapes at once: letters, fonts, sizes, even abstract patterns. How does the demonstrative narrow down which one is meant? And what about spoken language, where no visible quotation marks exist? Corey Washington and others noted that people say “My name is Donald” without any audible marks at all, yet the word is still clearly quoted — the name theory and demonstrative theory struggle to explain why.
The mirror idea: quotation as an identity machine
Maybe the simplest idea is the right one. The Disquotational Theory says that the function of quotation marks is nothing fancier than returning the expression they enclose. Mark Richard states it with a crisp schema: for any expression e, the left‑quote mark followed by e followed by the right‑quote mark refers to e itself. The marks work like a mirror: you feed in “bachelor,” and the mirror gives you back the very word “bachelor.”
This theory is wonderfully straightforward. It explains opacity at once: “‘bachelor’” and “‘unmarried man’” refer to two different strings of letters, no matter what they mean. There’s no internal structure to wriggle a quantifier into, so that puzzle evaporates. And the special closeness between a quotation and its content is explained perfectly — they are, in a deep sense, the very same thing, so of course they’re as close as it gets.
Yet the mirror theory has its own cracks. It must specify a list of basic expressions that can be quoted, but then how can we quote a symbol that isn’t in any existing language, like ‘❦’? If the rule only covers letters and punctuation we already have, the theory seems to shrink quotation’s reach. More pressing, the mirror doesn’t handle mixed quotation at all. If the quote‑marks simply serve up the expression, then “Quine said that quotation ‘has a certain anomalous feature’” would be grammatically identical to “Quine said that quotation Ted” — with a noun where a whole clause belongs. That doesn’t match how we actually read the sentence. And just like the demonstrative theory, the disquotational approach has no ready explanation for how we can quote by italicizing, by capital letters, or by just using words in a special way with no marks at all.
Why it matters: quoting, meaning, and the words you play with
Right now you might be thinking: so what? Why should anyone care what theory is correct? The answer is that quotation is one of the deepest ways language can turn around and talk about itself. Without it, you couldn’t say “‘literally’ literally means ‘in a literal sense’,” you couldn’t report what someone said while also signalling that you’re sticking to their exact words, and you couldn’t play with new symbols at will.
Mixed quotation, where you use someone’s words and mention them at the same time, is everywhere — pick up any newspaper and you’ll find sentences like “The court said this language ‘makes no sense.’” That tiny trick packs a lot of information: what the person said and that they said it in just those words. If you’ve ever made air quotes with your fingers, you’ve performed the same move. Quotation lets us hold words up to the light, examine them, and talk about the very tools we’re using in the middle of using them. It’s a piece of language machinery so ordinary that we barely notice it, yet so mysterious that after more than a century of work, philosophers still disagree on the most fundamental questions: Who does the referring? How does it latch onto a word? And what exactly does it pick out? The puzzle of those little marks hasn’t been solved — and that is precisely why it still matters.
Think about it
- If a friend says “My name is Chris” without any finger quotes, is Chris a person or a word? Could it be both at the same time? How would you decide?
- You hear a car horn and later say “‘honk’ has four letters.” Is the honk itself inside the quotation, or only the word “honk”? Can you quote a sound that isn’t a word at all — and if you do, what are you quoting?
- Why can’t you switch “love” for “a feeling of deep affection” inside quotation marks? If you think about it, we make those switches in normal sentences easily. What is it about those marks that stops the swap?
