Can You Think Directly About People, or Only Through Descriptions?
The Puzzle of Being Surprised
Maya just finished reading Huckleberry Finn and Tom Sawyer. In American literature class, her teacher says, “Mark Twain wrote both.” Maya nods. Then the teacher adds, “Of course, Samuel Clemens also wrote them.” Maya blinks. She knows that Samuel Clemens is Mark Twain’s real name — but when the teacher said it, the sentence felt different. If someone had asked her five minutes earlier, “Do you believe Samuel Clemens wrote Huckleberry Finn?” she might have said no. Does that mean Maya’s mind held two pictures of the same person without knowing they fit together? And what does that tell us about how thoughts hook onto the world?
This kind of confusion caught the attention of a German mathematician and philosopher, Gottlob Frege (1848–1925). He noticed that when we use names, we seem to see the same object through different lenses. How can one person — Mark Twain, a single flesh‑and‑blood man — show up in our heads as two separate “characters”? The answer, Frege argued, is that a name carries not only a reference (the actual thing it picks out) but also a sense — a way of thinking about that thing, like a mental direction sign. And for Frege, thoughts are built entirely out of senses, never out of the individuals themselves. A century later, philosophers are still arguing about whether that picture is right.
Frege’s Answer: Two Layers of Meaning
Frege’s core distinction was between reference (the person, place, or thing a word stands for) and sense (the particular mode of presentation, or mental route, that gets you there). For the name “Mark Twain,” the reference is the man himself; the sense is something like “the author of those funny books about a boy on a raft.” For “Samuel Clemens,” the reference is the same man, but the sense might be “the real name of the writer from Hannibal, Missouri.”
Because the senses are different, Frege said, the two sentences “Mark Twain was an American author” and “Samuel Clemens was an American author” express different propositions — different complete thoughts. A reasonable person can believe one and doubt the other. Maya did exactly that. Frege’s slogan: sense captures cognitive value — the difference that shows up when you learn something new. If you already know one way of thinking about Twain, learning that the other way points to the same man is a genuine discovery.
Frege was strict. He insisted that a proposition contains only senses, never the actual individual. Why? Because if the proposition contained the man himself, the two sentences would be the same thought, and you couldn’t rationally accept one while rejecting the other. There would be no puzzle, no “aha!” moment when you learn that Twain is Clemens. For Frege, all thinking about concrete individuals is indirect — it goes through a descriptive layer that could exist even if the individual did not. A proposition like “Twain is funny” is made of senses, not of Twain.
Russell’s Direct Thoughts (But Only About Your Own Mind)
Bertrand Russell (1872–1970) agreed with Frege that identity confusions need an explanation at the level of thought contents. But he gave a very different account of how our minds reach objects. Russell held an acquaintance principle: you can have a thought that directly includes an object only if you are acquainted with that object — and acquaintance, for Russell, meant a kind of mental contact so immediate that you couldn’t possibly mistake the object for something else. If you can reasonably wonder “Is this the same thing I saw before?” you are not acquainted with it; you are thinking of it only descriptively.
Russell thought we are acquainted with very little: our own fleeting sense‑data (the patches of color, sounds, and feelings that appear in our experience right now) and maybe abstract universals like “redness.” Everything else — tables, trees, other people, even our own past selves — is reached only through descriptions. When you think about your friend, your thought has the form the person who caused this perception I’m having right now is thus‑and‑so. The only object that appears directly in that thought is your own sense‑datum, not your friend.
So Russell agreed with Frege that thoughts about extra‑mental particulars are always indirect. But he didn’t invent senses as a special layer between words and things. Instead, he used definite descriptions (“the first chancellor of the German Empire”) and logical analysis to build complex thoughts from simple ones that do contain direct constituents. The simple, atomic thoughts are singular propositions — propositions that have individuals (sense‑data, for Russell) right inside them, like ingredients in a cake. Frege’s atomic thoughts were made of senses; Russell’s were made of individuals, but only very private ones.
Can You Point at a Person and Think Directly?
Many later philosophers found Russell’s list of direct‑thought objects too stingy. Neo‑Russellians — thinkers like David Kaplan and John Perry, writing in the 1970s — argued that we can think directly about ordinary things: the friend in front of us, the day we are living right now, even ourselves. If they are right, then singular propositions are all over the place in our minds, not locked inside private sense‑data.
They offered simple, vivid arguments. Imagine David points to the person on his right and says, “He lives in New Jersey.” That utterance expresses a proposition about Paul. Now imagine a counterfactual scenario: Paul and Charles switch seats, but everything else — including their home states — stays the same. In that scenario, the sentence “He lives in New Jersey” would have been about Charles and false. But the proposition David actually expressed is still true in that counterfactual world, because Paul still lives in New Jersey. If the proposition had been about Paul only in virtue of a description like “the person on my right,” it would turn out false in that world — since in that world Charles is the one on his right. The fact that we can track Paul across possibilities suggests the proposition is directly about Paul, not about a role Paul happens to play. This is the modal argument for singular propositions.
Still, many philosophers think the strongest push comes from words like “I,” “here,” and “today” — indexicals. Their meaning depends on who speaks, when, and where. If the only contents of thoughts were descriptive, what would “today” contribute? John Perry tested several Fregean answers and found them wanting. Suppose a man wakes up with total amnesia and says, “I am alive.” He’s clearly thinking about himself, even though he possesses no uniquely identifying description. No collection of qualities — “the person with brown hair,” “the one lying in a hospital bed” — can replace “I” without changing what he believes. The thought he has is a singular proposition with himself as a direct constituent.
Why “I” Is Not Just a Very Good Description
Perry sharpened the point with a story about Hume and Heimson. Heimson believes he is the famous philosopher David Hume. He knows every fact about Hume’s life, yet when Heimson says, “I wrote the Treatise,” he says something false — Hume himself says something true. The description “the philosopher who argued against the rationality of induction” fits both men from the inside, but the proposition expressed is tied to the actual person speaking. That tie cannot be captured by any purely qualitative description that both could share. So “I” must contribute the speaker herself, directly, into the proposition. Frege himself, late in his career, hinted that the first‑person might be special and not shared, but the neo‑Russellians turned that exception into a whole picture of thought.
This isn’t just a technical squabble. Perry designed a famous example to show why it matters for action. Imagine you see a bear charging while you’re staring into a mirror. You recognize the reflection as a person — but don’t realize it’s you — so you calmly watch. The moment you realize “I am being attacked,” you run. The belief state shifts from third‑person to first‑person, even though the object (you) is the same. Explaining behavior, Perry argued, requires distinguishing the belief state (how you grasp a proposition — as “I” or as “that person”) from the belief content (the proposition itself). The content might be a singular proposition about you; the state is the first‑person way you hold it. This split solves the puzzle that Frege’s all‑in‑one senses couldn’t handle.
Why It Still Matters: Your Thoughts, Your World
When you say “I’m hungry,” you aren’t just announcing that a certain description is satisfied; you’re making a claim that only you can make in that direct way. Your friend can think about you, but she can’t think your “I”‑thought. That unique grip on yourself is part of what makes you a self. The debate between Fregeans and Russellians is really about how closely our minds can wrap around the world — whether all our thinking filters through a net of descriptions, or whether some thoughts reach out and touch a thing directly.
This matters for everyday understanding. When a news report says “The winner spoke,” you might later learn it was Emma; you add a fact. But if someone said “Emma spoke,” you already grasped the proposition with Emma inside it — no extra step of matching descriptions needed. The existence of singular propositions would explain why some pieces of knowledge feel immediate, while others feel like solving a puzzle. It also affects how we hold people responsible: if a promise made with “I” really involves a direct reference to the speaker, then breaking it is more personally tied to them than a description like “the tallest kid in class” would be.
Philosophers are still working out the deep consequences. The modal and temporal problems that singular propositions bring — how can a proposition about George Bush exist in a world where Bush does not? — push us to rethink what existence and time even mean. Those are big questions for another day. For now, the next time you point and say “that one,” you might wonder: is the thing itself inside your thought, or only a good description you happen to hold? The answer could reshape how you see every sentence you speak.
Think about it
- If your best friend had total amnesia and forgot who they are, would their thought “I am alive” be exactly the same as the thought you have when you say “They are alive”? Why or why not?
- Could a scientist ever build a complete description of you that captures everything you mean when you say “I” — or does something always escape?
- If you learn a surprising fact about a person you know only by a nickname, did your thought about them change because you added a new description, or because you touched the same person in a new way?
