A woman sits across from a friend at dinner and says, with total sincerity, "I just don't know what I want anymore." Not about a big thing. About anything. The job, the city, the relationship, what to order. The friend nods, and says the thing friends say: "Think about it. What does your gut tell you?"

She has been thinking about it for months. That is the problem.

The uncomfortable truth, supported by a half-century of psychology, is that the tool we all reach for first, looking inward and asking why do I feel this?, is far less reliable than it feels. The sense that you can see your own motives clearly is vivid, confident, and often wrong. This is good news in disguise, because there is a better method, and it looks a lot like curiosity.

The window that isn't there

In 1977, Richard Nisbett and Timothy Wilson published a paper with a blunt title: "Telling More Than We Can Know." Reviewing many experiments, they argued that people often have little direct access to the causes of their own judgments and behavior. When asked to explain themselves, participants did not report what actually influenced them. They produced plausible stories, the kind an outside observer might offer, and delivered them with full confidence.

In one well-known example, shoppers evaluated identical items laid out in a row, and they tended to prefer the one on the right. Asked why, they cited fabric quality or texture. Almost nobody mentioned position, and when it was suggested, they dismissed it.

This is sometimes called the introspection illusion: the feeling that looking inward gives you a direct read on your own mind, when much of what you're doing is interpreting. You aren't reading a gauge. You're writing a caption.

When thinking harder makes it worse

It gets stranger. Wilson and his collaborators ran studies on what happens when people analyze their reasons for a preference.

In one, students tasted jams that had been ranked by trained experts. Those who simply tasted and rated them produced rankings fairly close to the experts'. Those asked to write out the reasons for their ratings drifted further from that benchmark. In another set of studies, people chose a poster to take home. Those who had listed reasons for their preferences tended to choose differently from those who hadn't, and later reported being less satisfied with their choice.

The proposed mechanism is intuitive. A feeling is a whole, fuzzy, whole-body judgment. A reason is a sentence, and sentences can only capture what's easy to put into words. When you ask yourself to justify a feeling, you grab the most verbalizable factors, such as price, practicality and what sounds sensible, and quietly overweight them. The harder-to-name signal, the one that was actually doing the work, gets outvoted.

This doesn't mean reflection is useless. (Other findings suggest that careful reflection helps in some settings.) It means the specific habit of interrogating your own inner state for causes, especially when you're already unsure, is a shaky way to find out what you want.

Know yourself the way you know a stranger

Here is the flip side. In 1972, the psychologist Daryl Bem proposed self-perception theory: that people often infer their own attitudes the same way they infer other people's, by observing behavior and the circumstances around it. When you see a friend choose the same café three weeks running, you conclude they like it. Bem's suggestion was that, especially when our internal signals are weak or ambiguous, we do exactly that with ourselves.

Notice what this reframes. If "what do I want?" can't be answered by staring inward, it can be answered by watching, with the patient, slightly amused interest of a biographer.

  • Not "why do I feel flat about this job?" but "on which days did I lose track of time?"
  • Not "do I love him?" but "when I see his name on my phone, what does my body do?"
  • Not "what do I want to do with my life?" but "what did I read, watch, or ask about this week without being made to?"

The first kind of question demands a verdict. The second kind collects evidence. Curiosity turned on yourself is just that: treating your own behavior as something interesting that hasn't been explained yet, rather than something you're supposed to already understand.

Why this is harder than it sounds

There are two obstacles.

First, we observe ourselves badly because we're busy narrating. The story we already believe, such as "I'm not a creative person" or "I'm someone who doesn't need much," filters what we notice. Behaviors that contradict it slip by unrecorded.

Second, memory edits. We tend to remember the peaks and the endings of experiences rather than their texture, so a retrospective "I think I enjoyed that" is a weaker data source than a note made at the time. Observation works best when it's close to the moment.

That's why the practical version of this idea is less about insight and more about logging.

Your next moves

None of these require a journal habit or an hour of quiet. They require a phone and about two minutes.

  • Run a three-line daily log for two weeks. Each evening write: one thing that gave you energy, one that drained you, one you did without being asked. No explanations. After fourteen days, read the whole thing in one sitting and underline repeats.
  • Swap "why" for "when." The next time you catch yourself asking "why don't I want this?", rewrite it as "when, specifically, did I last feel the opposite?" Answer with a date and a place, not a theory.
  • Check your free-time defaults. Open your screen-time report, your search history, or your library holds, and ask what you return to on your own. This is behavior you didn't have to justify, so it's the closest thing to an unedited signal.
  • Decide something small before you reason about it. Pick a menu item, a movie, or a route home, then notice your gut reaction to your own choice. Relief? A tug in another direction? That reaction is information. Take it down before you build the argument.
  • Ask one person who watches you, not who knows you. Choose someone who has seen you in action, such as a colleague or old friend, and ask: "What do I light up talking about?" Outside observers often see your pattern more cleanly than you do, which is exactly the point of Bem's idea.

What you're really practicing

Underneath all of this is a small shift in posture. "What do I want?" sounds like an exam you're failing. "What do I keep doing?" sounds like a mystery you get to investigate. The second is lighter, and, if the research on introspection is right, it is also more accurate.

The woman at dinner didn't lack desires. She lacked data she had bothered to collect, and a gentler way of asking.

This is the kind of attention we built curiokit around: small, daily prompts that point your curiosity at ordinary things, including your own habits and reactions, so that what you notice gets caught while it's still fresh and you're not left reconstructing it from memory. If you'd like a low-pressure way to start that kind of logbook, you can look around at curiokit. And if you'd rather just start tonight with three lines on the back of an envelope, that works nearly as well.