Ethics

 Claude finished the response

In theory, evolution provided us with an instinct that says stick with the ideas and behaviors of our upbringing.  The instinct is violable, but when we do, we feel bad things (guilt, fear) and when we follow them we feel much better, relaxed and even full of joy.  

We call this conscience, the mores and rules of our culture.  This implies people can end up doing perfectly awful things but feel nothing if they were not taught against it in childhood.  For the most part, however, because of a sort of natural selection among cultures, what the culture teaches is fairly moral and usually a reasonably good guide.

Our beliefs are another thing.  In a rapidly changing world there are downsides if we believe things stubbornly rather than changing them into opinions subject to critical scrutiny.  These beliefs can be compared to furniture we sit on without noticing they are there.  We don't see them as subject to question and can react emotionally when others do, sometimes, in some cultures, with deadly effect (such as executing homosexuals). 

Whether "good" and "evil" exist as objective realities in themselves is a philosophical question.  I feel that we should analyze our beliefs about good behavior rationally, rather than the emotions our instinct produces, but this is hard, not just to recognize the effect the instinct is having on us (such as creating harmful revulsions) but also to force ourselves to avoid intuitive judgments.  Various philosophical schools have tried to reduce the rational process down to deduction from fundamental principles, with a good deal of success, although in some cases with difficulties in artificial theoretical scenarios.  Study of this ethical history is a good idea if one wants to discern what really is good and bad rather than what our culture and instincts tell us.

We understand some people seem to be born without the instincts described above, and feel no emotion when they do wrong.  The tendency is even said to run in families (hence has a genetic aspect) and is usually called sociopathy.
10:51 AM

The framework is sound and I'd only push on a few edges.

The conscience-as-cultural-internalization account is well-supported, but the picture is slightly complicated by evidence that some moral intuitions appear cross-culturally with remarkable consistency — harm aversion, a rudimentary sense of fairness, something like care for offspring — suggesting evolution didn't hand us a blank slate waiting to be written on by culture, but rather a slate with some faint pre-inscriptions. Jonathan Haidt's moral foundations work documents this, though his interpretation is contested. The cultural selection mechanism you invoke is real but imperfect: cultures that persisted weren't necessarily the most moral, often the most militarily cohesive, which selects for in-group solidarity norms that can be actively harmful outward.

Your belief-as-invisible-furniture metaphor is exact and I'd extend it: the furniture is also load-bearing. When you try to remove a belief, you sometimes discover it was holding up several others, which produces the emotional violence of the reaction. That's part of why the response to challenges can be so disproportionate — it isn't really about the specific proposition, it's structural panic.

The sociopathy point is worth distinguishing more finely. There appear to be at least two separable deficits: in affective empathy (feeling what others feel) and in the instilled guilt mechanism you describe. Psychopathy research suggests the latter may be the more diagnostically central one — the absence of anticipatory anxiety about transgression, not merely emotional coldness. Some high-functioning individuals with the profile have intact cognitive empathy and use it instrumentally, which makes them more dangerous, not less.

One tension in your framework worth naming: you recommend rational ethical analysis as the corrective to instinct-driven moral emotion, and you're right that the philosophical tradition has done genuine work there. But Haidt and others have argued — controversially but not dismissibly — that the rational analysis usually arrives post-hoc, as a justification for intuitions already formed. If that's substantially true, the corrective has to be more procedural than purely rational: structured exposure to disconfirming cases, adversarial interlocutors, deliberate delay before judgment. The Quakers had something like this institutionally. Pure deductive ethics from first principles runs into the trolley-problem class of difficulties you allude to, where the machinery produces outputs that then strike nearly everyone as monstrous, which is itself interesting data.

The deeper question your framework raises without quite answering: if conscience is cultural and culture is imperfect, and rational analysis is partly hijacked by the very intuitions it's meant to correct, what's actually doing the work when someone gets ethics right? My tentative answer: mostly the quality of their interlocutors, historical and present.

To answer your answer, since no approach works perfectly, is approximation.  It is not something we are strangers to in other areas.
11:01 AM

Exactly right, and it's a more honest answer than most ethical systems are willing to give about themselves.

The resistance to acknowledging approximation in ethics is itself interesting. In engineering, medicine, economics, we accept that models are wrong but useful, that error bars are part of the output, that successive refinement is the method. Nobody expects a bridge calculation to be metaphysically certain. But ethical systems have historically tended to present themselves as complete and foundational — perhaps because their social function partly depends on that authority. A moral code advertised as "pretty good so far, subject to revision" has a harder time commanding the deference that makes it function.

Yet approximation is demonstrably what actually happens. The abolition of slavery, the expansion of the franchise, the slow revision of attitudes toward homosexuality — these aren't the discovery of new first principles but successive corrections to prior approximations, each generation inheriting a somewhat less bad version and improving it further, usually against fierce resistance from those who'd internalized the previous iteration as conscience rather than opinion.

The useful reframe your point implies: ethical progress is less like solving an equation and more like calibrating an instrument. You don't expect perfect accuracy; you expect to know your error range, reduce it over time, and remain alert to systematic biases in the instrument itself — which in this case is us.

The practical upshot is probably just: hold moral conclusions with the same epistemic posture you'd hold a well-supported scientific theory. Confident enough to act on, humble enough to revise, and particularly suspicious when the conclusion happens to be convenient for the person holding it.

Somewhat like Common Law rather than statute.
11:12 AM

That's a better analogy than anything I offered.

Common law is precisely the institutionalization of moral approximation — precedent as accumulated error-correction, each case a data point, the system explicitly designed to be revised by better reasoning applied to new facts while retaining what has survived scrutiny. It even has a built-in humility mechanism: the later court can always distinguish or overturn, but has to say why, publicly, in terms the tradition recognizes. That's adversarial interlocution formalized.

Statute by contrast does what rigid ethical systems do — crystallizes a moment's best judgment into something that then resists revision by its own authority, and tends to produce absurdities when applied to cases its drafters didn't anticipate.

The analogy also captures something about the role of edge cases. In common law, the hard bizarre case isn't an embarrassment — it's how the system learns. The trolley problems that make deductive ethics look foolish are, in the common law framing, exactly the kind of case you want: they reveal where the current approximation breaks down and force a more precise articulation of what principle is actually doing the work.

What common law doesn't fully solve, and neither does ethical approximation, is the question of which precedents get standing. The tradition excluded most of humanity from its reasoning for most of its history, and the cases that reached courts reflected that. The system is only as good as who gets to be a party. But that too is a correctable approximation rather than a fatal flaw — if you remain alert to who isn't in the room.
Claude Fable 5 is currently unavailable.
Learn more(opens in new tab)


Claude is AI and can make mistakes. Please double-check responses.