Morality Without a Detector
Morality is a construct of sentience. No detector exists, and the universe doesn't care. Every moral statement ever encountered has arisen from a human mind, and the ease with which two people can disagree on almost any action suggests that if a moral code exists outside our heads, we're remarkably bad at listening to it.
It is rarely impossible to commit an immoral act. Outside of physical constraints, the universe does not pick sides. When evaluating an individual's moral character, we find it is founded entirely on the morality of their actions (including thoughts). A method for reliably ascertaining moral status that is not a function operating purely on actions remains elusive.
Formalizing Moral Systems
Some moral systems are binary: an event is either moral or immoral. These systems struggle with complex actions, such as running over little Suzie and Hitler with the same steamroller. Most people, however, assume acts have varying degrees of "goodness," and many would assert goodness is a scalar quantity—saving two lives is twice as good as saving one.
Within these constraints, a moral system maps every action to a scalar value normalized in the range [-1, 1], with positive values corresponding to good actions and negative values to evil. Deontological systems select values based on conformance with rules, and interactions between rules are unconstrained: they may supersede one another or count for more or less. Consequentialist moral functions operate purely on the state of the universe.
Formal Equivalence
As the state of the universe includes its history, a consequentialist framework can distinguish between outcomes ("Suzie is dead") and any action ever taken ("And Nefertiti was inadvertently responsible"). Because deontological rules can take into account the universe state in addition to the action itself, they too are free to distinguish between any configuration of the universe.
This means that, in general, deontological models are formally equivalent to consequentialist ones. For any deontological model, we can construct a consequentialist function that yields identical moral values for all possible situations—and vice versa.
Parsimony as a Deciding Factor
Given functionally identical models, the choice comes down to ideology. Parsimony is the most compelling constraint. Consider two ways to express the same function:
f(x) = (|x|)^2
g(x) = x^2
Both produce identical results for real x, yet f(x) introduced an absolute value operation which has no effect but retains the illusion of meaning.
Now consider canonical moral functions for the Ticking Time Bomb scenario—torturing one person to disarm a bomb and save five:
d(x) = +1/5 for saving a life. -1/9 for torturing a man, unless doing so to save human life.
c(x) = (+2/9 for each life extended) + (-1/9 for a man experiencing torture).
The two are functionally identical, but the deontological expression is more complex, requiring a qualifying clause to account for a situation where a simpler rule system could produce non-optimal results.
Expand the action space: save two of five by throwing a bucket of adorable puppies onto the tracks. The consequentialist model needs a single new term—say, -1/9 per puppy flattened. The deontological model requires more complex rules: animal cruelty is bad except when saving human lives; torture is OK when animal lives are at stake; and so forth.
Imagine a global moral function mapping every possible action to a moral value. Consider the exceptions and special cases needed to formulate a complete rule-based system that yields "reasonable" results. What drives the creation of those rules? What do you do with a situation lacking codified exceptions? Consequentialist frameworks satisfy the reasonability constraint with fewer exceptions by carefully weighting components of outcome states—and those weights are what underly moral rules in the first place.
A Consequentialist Stance
One personal moral function is the sum of the correspondence of every living being's objectives with their subjective experience, integrated over their lifetimes, nonlinearly weighted by sentience. This function is maximized when an individual is self-actualized, living a full life in concordance with personal objectives—eudaimonia. It is nonlinear: it is strongly negative if an individual lacks basic needs or self-determination. It considers the full lifecycle: temporary unpleasantness for long-term fulfillment is acceptable. It gives the strongest consideration to humans and similar sentience, and less to cats or individual insects.
This framing leads to surprising differences with traditional deontological systems. Gay marriage is preferable where it maximizes fulfillment of some couples while imposing minimal constraints on others. Imprisonment and murder are morally correct where doing so has significant positive consequences, such as preventing greater damage or death. Assisted suicide is acceptable when continued life would produce more unhappiness than the loss induced by death. Forced population constraints are acceptable to prevent future resource exhaustion and associated wars or famines. Lying is correct to prevent greater injustice, balanced against creating a cultural norm of distrust.
Evaluation of such a function is in practice impossible; many terms are undefined or unmeasurable. But actions are largely local, and the framework works as a usable heuristic. Its terms are subject to inquiry—you can ask parties what they really want, whether they are happy, and how things could change. Framing problems in terms of a desired world state allows creative escapes from rule gridlock.
Probabilistic Consequentialism
A classic critique of consequentialism runs: Johnny Dangerously speeds on the highway but hits nobody and causes no accidents. Since no harm occurred (and Johnny saved time), his decision to speed was moral.
Consequentialism operates on the full state of the universe. Under strict determinism, there is only one moral value for the universe we're in. All "decisions" are equally moral—if they can even be said to exist. Johnny couldn't have hit anybody if he tried.
But if we believe we can change the universe, decision-making lets us choose between moral outcomes. Johnny's decision to speed selected a universe with different consequences. He didn't know which universe he was choosing—some involved accidents, some not. A moral value for his decision integrates over all possible universes, weighted by their probability of occurring. Since some universes include fatal accidents, the integral for speeding is lower, and the decision less moral.
The universe is fundamentally probabilistic, meaning these integrals over possible universes must be taken into account in every situation. Schrödinger's cat experiment, with the cat in half-and-half superposition of dead and alive, is half as bad as killing it absolutely. The same method evaluates subjective moralities: if an individual has a 50% likelihood of immunity to Iocane powder, feeding it to him is not as bad subjectively as when immunity is inconceivable. Brandishing a knife in a darkened kitchen is worse than waving it about where you can see.
Consequentialism. It works.
