I recently saw a video in which someone is posed a high-stakes, but vague, yes-or-no question. I’m not here to discuss that incident. Rather, I wish to point out something that I hope is obvious: yes-or-questions are bad. More severely, they are dangerous: give someone the power to impose yes-or-no questions, and you give them the world.
You can insert anyone into almost any picture of reality you want with yes-or-no questions. And, unfortunately, many people will believe the picture. Or at least pretend to believe it. I’m hopeful that if we endeavor to publicize the dangers of yes-or-no questions, it will make it harder to (pretend to) believe in the pictures they manufacture.
As far as I can tell, entailment is the core phenomenon exploited by the worst sorts of yes-or-no questions.
For example, suppose we ask an agnostic, under a compelled yes-or-no paradigm, “Do you believe God exists?” The answer must be “no.” Of course, the answer to “Do you believe God does not exist?” is also “no.” Being agnostic entails both of these things. But a yes-or-no question doesn’t capture that.
Many people think agnosticism is just as bad as atheism, so perhaps I should produce a better example.
On a recent episode of Ezra Klein’s podcast, he said (@ 1:11:20): “we’re barely above monkeys, we’re barely above monkeys, and the idea that we’re going to figure it all out is just an amazing amount of hubris.”
The comment was in reference to “humanity” attempting to uncover a “description of ultimate reality.”
[The episode is called “A Serious Conversation about UFOs” (6/10/20) and features an excellent discussion with scholar Diana Walsh Pasulka, author of what looks like a fascinating book, American Cosmic: UFOs, Religion, Technology (2019). The episode was inspired by the recent confirmation from the U.S. Department of Defense that three videos featuring “unidentified aerial phenomena” (a less-loaded term for UFO) are real; here’s the DoD’s 4/27/20 press release: “Statement by the Department of Defense on the Release of Historical Navy Videos.”]
I frequently hear people say such things about humans. But I single out Klein only because it popped into my head as a nice, recent example of the point I’m trying to make. Here’s what I mean.
I love some of Frank Zappa’s music. Call me a Zappa fan. Klein’s comments, unfortunately, imply that Zappa fans are barely above monkeys. (I’ll use “entail” and “imply” more or less interchangeably.)
That is, were we to ask Klein, under a compelled yes-or-no paradigm: “Did you imply on your podcast that the humans known as Zappa fans are barely above monkeys?,” he would have to answer “yes.”
What’s really relevant here is whether he also implied that only Zappa fans are barely above monkeys, or some similarly distasteful proposition. He obviously implied no such thing. But an appropriately exploited yes-or-no paradigm doesn’t capture such nuances.
Also not captured is that Klein might have never even heard of Frank Zappa, and thus have no beliefs at all about Zappa fans as such. This doesn’t change the fact that his statement, strictly speaking, implies that Zappa fans are barely above monkeys.
Similarly, were we to compel (someone I just invented named) Teri to answer, “Do you have any friends who are Zappa fans?,” she would have to answer “no.” Why? Because Teri has no friends, and having no friends entails having no friends who are Zappa fans. But such nuance, again, is not captured by the question.
This only scratches the surface of the frames (let’s call them) you can put around people with yes-or-no questions.
Of course, there are examples where exploiting entailment would be, I hope, transparently silly: “Is it true that you either dislike dogs or are a human being?” The worry, of course, is that the “or” operator is vulnerable to subtler exploitation.
Are yes-or-no questions sometimes useful? I suppose. “Is this your dog?” being answered with “It depends on what your definition of ‘is’ is” makes reformulating the inquiry as a yes-or-no question attractive.
To be safe, though, we should always also allow for qualification, clarification, context, or there may be an implication we’ll otherwise miss. Notice, though, that in the case of entailment, the worry isn’t about taking something out of context, but rather about bringing out something that is unambiguously, or at least strictly speaking, true, yet supports an arbitrarily assembled falsity.
I wonder what it takes to make an interesting yes-or-no question. “Does 1 + 1 = 2?” Not in binary, it doesn’t.
What I mean by “interesting” is a problem for another day. Perhaps I really mean “minimally robustly informative” or “sufficiently specific.” “Is there something rather than nothing?” is the most general question I can think of. The answer is “yes.” Uninteresting. More interesting: “Why is there something rather than nothing?”
Finally, there are certainly other ways yes-or-no questions are unhelpful or dangerous. For example, two people might answer “yes” to “Are you against capital punishment?,” while one gives that answer for humane reasons, and the other gives it for inhumane ones (i.e., believes instead that wrongdoers deserve many years of physical torture, or should be forced into servitude).
Here’s another example. Two people answer “yes” to being in favor of blifgarb, where blifgarb means something entirely different to each of them, and would result in drastically differing policies. (Better, perhaps, to discuss the policies themselves, then, before worrying too much about defining blifgarb.)
Surely there are many, many yes-and-no modes, but I’m particularly concerned about entailment because its underlying mechanism runs on formally valid logic, while its worrisome effects rely on informal goings-on.
Enjoy or find this post useful? Please consider pitching in a dollar or three to help me do a better job of populating this website with worthwhile words and music. Let me know what you'd like to see more of while you're at it. Transaction handled by PayPal.
Or click the banner to shop at Amazon (at no extra cost: it just gives me some of what would have gone to Amazon).