“Omniscient Prole” Dilemmas

In pure consequentialism, an act is morally right if it produces results that represent an optimization of what is valued (whatever that is).

All statements of “should” or “ought” are built from two inputs:

• “What is valued?,” i.e., “What is the goal, interest, or desire (or set thereof) to which I’m acting in service?” and
• “For each prospective course of action, what will happen as a result?”

So we have our circles (input values or goals) and our squares (how things work). These always resolve into rounded rectangles (moral statements under consequentialism, e.g., “It is right to do X,” “I should do X,” “X is justified,” “X is correct,” “X is praiseworthy,” “Y is wrong,” “Y is suboptimal,” etc.).

Our Problem: We Are Numbskulls

The thing is, our squares are ill-defined. We may never have perfect squares because observations about the mechanics of the world, in general and in specific instances, may never be completely accurate.

And because the world is chaotic (that is, rather orderly, but nonetheless incalculably complicated), our faculties of foresight are dramatically limited.

We can’t even predict the weather accurately more than 3 days out. How much more terrible are we at predicting the effects of our actions on human behavior, including our own behavior?

We not only recognize that we don’t know that much about how the world works, but we simultaneously recognize that this ignorance leads to unintended consequences all the time.

We respect this intuitively, and we account for it in our decisionmaking.

Our Solution: Humility and Recognition of Uncertainty

Just as businesses account for risk in their decisionmaking, we individuals make intuitive attempts at accounting for risk. We recognize that our squares are severely crippled, and so many of our “shoulds” are loaded with qualifiers like “probably” and “maybe.”

For this reason, we say that pure consequentialism is impractical. We adopt some diluted form of consequentialism that recognizes our imperfect understanding of the world and the dramatic impact that imperfection has on our ability to predict the full consequences of our actions.

This includes even moral templates that allow for rules; it allows us to say that blanket, imperfect social laws may sometimes and in some places be required to correct for subjective error and individual stupidity.

So even as consequentialists, we reject pure consequentialism on the grounds that we are dummies. We humble ourselves below the status of omniscient beings because we know we aren’t.

See this post for an introduction to this concept, called “The Angelic Ladder.” We are “Near-Proles”; we are not completely stupid and blind to consequences, but we are pretty stupid, and pretty blind to consequences.

Loaded Moral Dilemmas

So, here’s the trick that many thought experiments pull:

• They set up a situation that tests your decisionmaking in a consequential context.
• They then give you some measure of implausible omniscience, e.g., “You know for absolute certain that pushing the fat man in front of the train will derail it, and that the train is certainly empty, and that the derailed train will not hurt the group of people you’re trying to save.”
• They then watch as you squirm with the anxiety of being an “omniscient prole,” where you’re struggling to reconcile your learned, intuitive, hammered-in humility with the new God-power you’ve been granted by the situation.

The thought experiment wants you to think that it’s demonstrating that consequentialism is an incomplete description of morality and needs additional deontological (“morality is all about rules”) magic. But all it’s actually doing is showing that we are intuitively averse to pure consequentialism because we know we’re so limited.

In other words, morality isn’t some same-level hybrid of consequentialism and deontology. That’s something a lot of people think and something with which a lot of people struggle.

Rather, morality is consequential, but due to our limited faculties of foresight and understanding, we find it useful to employ bits of subordinate deontology.

I’ll again link you to the previous post on the subject, “The Angelic Ladder,” which serves as a primer to “rules under consequentialism.”

Dealing with Loadedness

If someone asks you, “Are you a piece of garbage, or do you just smell like one?,” how should you respond? No, don’t punch them in the face. Rather, choose one of the following options:

• Indict the loadedness. “That’s a loaded question so I refuse to respond.”
• Unload the question. “I am neither a piece of garbage, nor do I smell like one.”

Similarly, if someone gives you a moral dilemma loaded with implausible omniscience, either are fine:

• Indict the loadedness. “That situation gives me implausible omniscience which ungrounds us completely, so I’m not going to respond as if my answer would reveal anything morally significant.”
• Unload the situation. “I wouldn’t fathom that the man’s body would be able to derail the train, would worry that the train would contain passengers I’d put at risk, and would be concerned that a derailed train might hurt the very group I’d like to save. Sorry. Real situations are addled with uncertainty.”

(It so happens that if we’re dealing with a moral situation with low uncertainty and non-competing interests, then the more grounded that moral dilemma is, the easier it is to answer confidently.)