Our stimulus does feature a claim that could not possibly be true even in principle. It’s this:

Impossible Claim: Most people sleep less than most other people.

And the (anticipated) main point of the argument is that some people are wrong to believe a certain claim. Here’s that untrue claim, though:

Claim Some People Falsely Believe: I sleep less than most other people.

These are two different claims.

Here’s a parallel: my friend and I are wondering which of us is smarter than the other. We know we can’t both be smarter. That’s not possible. No one believes that. But we do each believe ourselves to be the smarter person. One of us is wrong to believe this.

(It’s me 😭)

Among other things, this mixes up relative and absolute claims. To make the determination that you sleep less than others (which is what our stimulus is all about), you don’t need to know exactly how much you sleep.

Let’s say you’re driving down the highway with a broken speedometer. You’re playing it safe because, you know, your speedometer is broken. As you look around, you notice all the other cars on the highway are zooming past you. You correctly believe you’re going slower than most other cars, even though you don’t know exactly how slow you’re going.

This is right if not for the “most.”

Wait actually let’s just doublecheck the word “underestimate” real quick. It’s easily confused with “overestimate.”

Underestimating something means your estimate is lower than the reality. I think you’re 5 feet tall, but you’re 6 feet tall. I have underestimated your height. Underestimating how much sleep you get is thinking you sleep less than you actually do. Okay cool that’s the right direction we can keep going.

Recall the sleep deprivation contest, with rankings from #1 (least sleep) to #100 (most sleep). 51 people think “I’m in the top 50 sleep-deprived people.”

A correct version of (C) would point out that if 51 people think they’re ranked top 50, at least one of them must really be ranked lower than that. This person is underestimating their sleep.

But (C) says at least 51 people are ranked lower than they think they are. That’s no good.

Of these 51 people who think they’re top 50, imagine that every single one of them thinks they are ranked exactly #48 on the list. They all think “I sleep less than most other people, but not that much less. I bet I’m borderline. Call it #48.”

In this world, people really don’t think their rankings are very high. In fact, a lot of them (for example, the people who are actually ranked #4, and #27, and #45) are probably closer to #1 than they think. (These people are overestimating their sleep.) So we can’t guarantee 51 people are ranked lower than they think.

(D) invites us to leap across the dreaded is-ought gap – to jump from descriptive premises about how things are to normative conclusions about how things ought to be.

The stimulus’ premises are about how much sleep people think they do in fact get. (D)’s conclusion is about how much sleep people think they ought to get.

You’re not allowed to leap from “is” to “ought” – David Hume’s ghost will haunt you.

Recall the sleep deprivation contest. In our world of 100 people, 51 of them think they are in the top 50 when it comes to sleep deprivation. But it’s impossible for 51 people to be in the top 50 because of how numbers work. So some of those people are mistaken.