Questions Aren’t Free

I don’t fully buy the advice, “If you’re not sure, just ask.”

It’s usually presented as the polite, healthy, low-risk option—especially in workplaces, classrooms, and relationships where people want to discourage guessing and prevent mistakes. But there’s a missing variable in that rule: asking a question is not free. It creates cognitive load, and that load lands on someone else.

What bothers me isn’t questions themselves. It’s how casually we treat them, as if they’re weightless, as if they don’t interrupt, as if they don’t require work. And because that work is invisible, it rarely gets acknowledged, measured, or protected.

Questions are an attention tax

A question looks like a sentence. But receiving a question triggers a whole series of mental tasks that aren’t optional if you want to answer responsibly.

When someone asks you something, you’re often doing at least three things:

• Reconstructing context. What is the person actually asking? What have they tried? What do they know? What are they assuming? What are they leaving out?
• Deciding the appropriate level of response. Is this a one-line answer? A walkthrough? A cautionary tale? Is the person asking for facts, advice, permission, validation, or something else entirely?
• Accepting a slice of responsibility. If you answer and they act on it, and it goes sideways—what does that mean? Are you now part of the outcome?

Even if you give a short answer, you still paid the switching cost: you stopped what you were doing, loaded their problem into your head, and made room for it.

And because the cost is mostly mental, it blends into the background. The person asking feels like they did something “communicative.” The person answering just feels slightly more tired.

“Ask vs. assume” is a false binary

The moral framing around questions often goes like this:

• Asking = humble, collaborative, safe
• Assuming = arrogant, risky, antisocial

But that’s not the real decision most of the time.

The real decision is usually:

• Resolve uncertainty internally (read the doc, search your notes, infer from patterns, run a small test, make a reversible choice, fail quietly)
• Interrupt another mind (make your uncertainty someone else’s active problem)

Those are not equivalent moves. They’re different strategies with different costs, and the “just ask” advice tends to pretend the second one is costless.

In high-functioning environments—good teams, good friendships, good relationships—people usually develop an instinct for “self-resolution first.” Not because they’re afraid to communicate, and not because they’re trying to be lone wolves. Because attention is finite and interruptions compound.

That’s not arrogance. It’s respect.

Some questions are disguised dependency

There’s also a category of questions that aren’t really about missing information.

They’re about:

• wanting permission (“Is this okay?”)
• shifting liability (“You told me to do it this way”)
• seeking reassurance (“Am I doing this wrong?”)
• avoiding the discomfort of being publicly wrong (“Can you confirm before I send this?”)

These questions can look harmless. They can even look conscientious. But they create a dynamic where one person carries the thinking, and the other person “checks in.”

That dynamic is exhausting for the person who becomes the default checker. It’s also quietly corrosive, because it trains the asker to outsource judgment instead of building it.

If you’re in a role where people constantly seek your approval—manager, senior engineer, teacher, oldest sibling, “responsible friend”—you can feel your day being eaten alive by micro-requests that aren’t about knowledge. They’re about emotional risk transfer. And to be clear: sometimes reassurance is appropriate. Sometimes permission is necessary. Sometimes you truly do need a decision-maker. But if “asking” becomes a reflex for avoiding internal commitment, it turns into a dependency loop.

Good questions reduce total load; lazy questions relocate it

Not all questions are equal. Some questions are net-positive: they save everyone time, prevent expensive mistakes, and clarify shared reality. Others are just outsourcing.

A good question usually does a few things:

• shows prior effort (“I read X and tried Y”)
• narrows the uncertainty (“I’m deciding between A and B”)
• includes relevant context (“The constraint is Z”)
• makes it easy to answer (“Can you confirm which rule applies here?”)

A bad question is often maximally open-ended:

• “What should I do?”
• “Is this right?”
• “Can you check this?”
• “Any thoughts?”

Those aren’t questions so much as invitations for someone else to do the work of thinking, prioritizing, and taking responsibility—without being explicitly asked to take that role.

If you want a quick self-test: before sending a question, ask yourself whether you’ve made it easier for the other person to answer than for you to figure out yourself.

If the honest answer is no, it’s probably not a question; it’s a handoff.

The hidden politics of who gets asked

There’s also a social layer people avoid naming.

Questions don’t distribute evenly. They cluster around certain people:

• the most competent person in the room
• the most approachable person in the room
• the person who can’t afford to be “mean”
• the person who replies quickly
• the person who is perceived as safe

Over time, those people become bottlenecks—not because they asked to be, but because everyone else learned that they’re the easiest path to certainty.

This is how you get “helpful” colleagues who can’t do deep work, senior people who become human search engines, and emotionally steady friends who are quietly drained by everyone’s “quick question.”

And when the culture is “just ask,” the people bearing the cost are often pressured to treat their own attention as public property.

Why neurodivergent people feel this more sharply

This is where the “nobody talks about this enough” feeling hits hardest.

If context-switching is expensive for you—because you process deeply, because you need to re-orient carefully, because ambiguity lingers longer—then being interrupted with questions is not a small ask. It can derail your entire cognitive stack.

Even the phrasing “it’ll just take a second” can be maddening, because the visible time to answer is not the real cost. The real cost is:

• losing your thread
• rebuilding momentum
• carrying someone else’s uncertainty inside your brain
• managing the anxiety of giving the “right” answer

For someone with cheap switching, “just ask” works fine. For someone whose attention is fragile or precious, it’s like being told to keep opening tabs when your RAM is already maxed out.

Well-meaning advice often assumes a mind that can pivot without friction. Many people don’t have that mind. They just get blamed for being “rigid” or “unhelpful” when they protect their bandwidth.

A better rule than “ask if you’re unsure”

A sharper principle is:

Only ask when the expected cost of being wrong is higher than the cost of interrupting another mind.

That reframes “asking” as a tool, not a virtue. It also forces you to consider alternatives:

• Can I run a small reversible experiment?
• Can I look for a precedent?
• Can I make a best-effort guess and correct later?
• Can I document what I find so the next person doesn’t ask?

This doesn’t mean “never ask.” It means treating questions like what they are: requests for attention and responsibility. Sometimes that request is absolutely worth it. Sometimes it’s not.

Conclusion

Asking questions is part of being human, and in the right moment it’s generous, efficient, and connective. But the blanket rule “just ask” hides the real cost: questions consume attention, and attention is a limited resource. The healthier norm isn’t silence or guessing—it’s effort first, clarity when you do ask, and respect for the fact that every answer requires someone else’s mind.