Don't Fight the Hypothetical

If you want to have a productive philosophical conversation, you need to be willing to entertain hypothetical scenarios.

My favorite philosopher Michael Huemer has written a philosophy textbook Knowledge, Reality, and Value, in which he takes a section to describe how to be a good interlocutor when talking about philosophy. His first principle is that you should be cooperative. One important feature of being cooperative is that you should accept hypothetical situations without raising a huge fuss over unimportant details.

If someone gives a hypothetical example, do not raise objections to the example that will only make your interlocutor waste time thinking of other examples, or thinking of a series of increasingly elaborate modifications to the example. Do not start a debate about how realistic the example is or “what would really happen” in a situation of that kind; just accept the example as the other person intended it.

The person providing the hypothetical example is trying to illustrate something. Tangential discussions about plausibility and workarounds aren’t addressing the central issue. If you respond to the trolley problem by suggesting that you would try to move the switch halfway in-between left and right to run the trolley off the track to save all six people, you are missing the point. All that this does is frustrate your interlocutor and make them devise a more complicated scenario to get you to address the central issue.

Scott Alexander recalls a similar example in his article The Least Convenient Possible World. When faced with the organ harvesting thought experiment in which a doctor can harvest the organs from an innocent traveler to save several people’s lives, one of Scott’s friend’s responded by saying:

It wouldn't be moral. After all, people often reject organs from random donors. The traveller would probably be a genetic mismatch for your patients, and the transplantees would have to spend the rest of their lives on immunosuppressants, only to die within a few years when the drugs failed.

While this is an interesting consideration, we can see how it doesn’t address the central issue. The point isn’t to note something about genetic mismatch, it is to see what someone thinks is the right choice in a difficult ethical dilemma. I think that hypotheticals are often avoided when they create a tension between someone’s intuitive sense of the good and someone’s ethical code. For example, it’s likely the friend was a utilitarian who thought the organ harvesting would be utility-maximizing, and yet they felt it was intuitively wrong to non-consensually take someone’s organs.

This happens a lot when I have ethical discussions. Many people adhere to one-sentence ethical theories such as radical libertarianism or utilitarianism. I think that we have ethical intuitions that give us access to moral knowledge, and that these intuitions don’t support single-sentence theories of ethics. If I see someone express the belief that ethics can be reduced to a single sentence, I attempt to find a counter-example that creates tension between their intuitions and their theory. When this happens, people fight the hypothetical.

Imagine that someone said it was always wrong to lie, and I asked them if they would lie to protect someone from an axe murderer. I think intuitively they would believe it was okay to lie in that situation. For that reason, they may say that they don’t need to respond to unlikely scenarios. Hardly anyone is killed via axe murderer and even if that was going to happen, it’s unlikely you would have to lie to protect them. While that is true, it is still worth answering the question. Just because something is unlikely to occur does not mean that it can be dismissed outright.

If we were discussing geopolitics, and I asked someone to consider what would happen if China invaded the Bahamas, I doubt that they would object on grounds that it was a low-probability event. We could have a meaningful conversation about this hypothetical world. Maybe the USA would react by sanctioning China. Maybe Cuba would react by preparing to defend itself against a Chinese invasion. Even if it is a waste of time, it might be worthwhile to entertain the idea to be a cooperative participant in a conversation.

Another objection to the axe murderer situation is that I am trying to normalize lying. I could agree that it is generally not ethical to lie, but is ethical in extreme circumstances. Finding one scenario doesn’t mean I want it to be regarded as generally okay. Besides, I have no problem with the idea that it’s most often bad to lie. I have an issue when people make extreme and absolute moral statements. The reason that I devise hypotheticals is to test someone’s actual commitment to the extreme position. Even if you think that I am trying to normalize lying, you can still answer the hypothetical while also expressing that you don’t think it generalizes to all situations.

I’m skeptical that people would have such a reaction to unusual scenarios if it is not in conflict with their beliefs. For example, I could ask if I should lie to a cashier to get away with stealing alcohol to give to some seven-year-olds outside the local gas station. I think most would easily respond with no. It goes against many ethical intuitions and conflicts with almost everyone’s ethical principles. I am doubtful that people would say the idea of a 7-year-old being unaccompanied and wanting alcohol is highly improbable and so they don't have to entertain this. Even if it’s a waste of time to entertain it, at least respond yes or no before the spiel about how it’s a dumb question.

Analogously, imagine that someone said the function g(x) is larger than the function f(x) everywhere. If I pointed out that it wasn’t true for the number fifty billion, then the statement is falsified. You shouldn’t say that fifty billion is a particularly large number or that it works for most numbers. If that’s the case, then just say that at the outset. What an extreme counter-example suggests is that your theory is false and that the method of reasoning used to derive it is suspect. Even if it is merely an extreme example of a single principle, it should cast doubt on your other moral principles. You’re doing something incorrect to reach your conclusions.

If I was teaching students about probability, I could ask what the probability of me getting two heads after flipping a coin twice is. A clever student could say "Coins are not perfectly balanced and you cannot flip the coin in a perfectly random way and so we can't know what the probability is." If I said "just let the probability be 50-50" the student could respond "Why should I trust you? That sounds extremely suspect to think that a perfect coin could exist." And if I went "Okay, imagine a Bernoulli random variable with p=.50" the student could object "there exists no such random variable in reality." This would all be a very tedious process. Ultimately, we know that probabilistic reasoning is legitimate even if it is not perfectly representative of real-world events. So, what function does fighting the hypothetical serve?

If you have a law of morality, then the law needs to apply everywhere. If a scenario could conceivably occur and your moral theory says in all circumstances do X, then in this scenario you should do X. My position is that you can always find a scenario where X is counter-intuitive. You can either bite the bullet or modify your theory, but I think objecting to hypotheticals generally is not a fair approach. It sounds reasonable to say you only respond to real scenarios, but I can't find a legitimate reason as to why we cannot entertain low-probability or impossible hypotheticals. These are often insightful in a way similar to talking about coinflips, even if you can’t have a perfect coinflip. When you find a counter-example to an absolute ethical theory, it is suggestive that your reasoning process for reaching that ethical theory is wrong.