Plumb Lines - Episode #1

Moral Truth and the Limits of Preference

Welcome to Plumb Lines. Before I tell you what this series is about, let me set the scene.

You take your truck to a mechanic who knows you can't check his work. The problem is a loose battery terminal. He tells you it's the alternator, charges you for one, and sends you home. You never find out. He makes a few hundred dollars and sleeps fine.

What would you need to know before you called that wrong?

Whether he feels guilty? We already covered that; he doesn't. Whether the other guys in the shop approve? Say they do. Whether he broke a law? That matters, but it just moves the question: if you'd driven across a state line to somewhere the law allowed it, would the same lie be fine?

I don't think so. You were owed an honest diagnosis. The fact that you couldn't check it made you easier to cheat; it didn't make the cheating acceptable.

That is the problem this series works on. Most of us make moral judgments all day without much trouble. We are a lot less comfortable saying what those judgments answer to, particularly when the honest answer would cost us something.

Needless to say, this topic has been done to death. Philosophers have had a few thousand years at it, and the internet has more than enough people explaining how everyone else should live. What I'm interested in is the step before anybody prescribes anything: how do you tell whether the principle you're using is any good? If a decision framework can only confirm what you fed into it, you shouldn't be surprised when it agrees with you.

What are we checking against?

The name comes from a builder's tool. A plumb line is a weight on a string. Hang it next to a wall and gravity gives you vertical whether you like it or not. You can prefer the wall the way it is; you can have good reasons for not tearing it down. Neither of those changes the reading.

Morality is harder than a wall, obviously. There's no string you can hang next to a decision, and the analogy doesn't prove anything about whether moral truth exists. I won't belabor it (it comes back later, but only when it earns its keep). It does give us the question for the series: can a judgment be corrected by something we didn't choose?

I think it can. The philosophical name for the position I'm defending is moral realism: some moral claims are true as a function of moral facts and reasons whose authority doesn't depend on our approval. We can understand those reasons poorly. We can ignore them. But deciding we would rather not deal with an obligation doesn't, by itself, make it go away.

That definition needs a little care, so let me pause there. An obligation can depend on human choices. If I make you a promise, my choice created an obligation I didn't have five minutes ago. The realist claim is narrower: once the circumstances exist, whether the obligation holds isn't settled by whether I still like it. The mechanic's customer has a claim to an honest diagnosis. I can agree with that today and find it inconvenient tomorrow when I'm the one holding the wrench. Realism says the claim survives the change in my enthusiasm.

Worth noting up front, because I want to be fair to the other side: the strongest alternatives to realism can say that too. A careful expressivist (more on them shortly) will also tell you the lie stays wrong when the whole shop approves of it. The disagreement is about what explains that. Realism starts with the moral facts and says our attitudes ought to answer to them. The rivals start with the attitudes and try to show how moral thought can still have truth and correction in it. That's the actual dispute, and it's a better one than "principle versus whatever I feel like."

What preferences can tell us

Preferences are inputs to a lot of perfectly good decisions. You like one kind of music and I like another; we listen to different things, and nobody needs to convene a tribunal about the playlist.

Feelings also carry information. Anger might be pointing at an injury. Sympathy might make somebody's trouble visible when you'd otherwise walk past it. An account of morality that leaves feeling out has left out a big piece of the subject.

But what if my anger isn't justified? What if I feel guilty about turning down a demand I had no duty to meet? What if I'm perfectly comfortable taking advantage of someone? At that point, reporting the feeling doesn't settle anything. The feeling is the thing that needs checking.

There is a practical difference between saying I don't like being cheated and saying you shouldn't cheat me. The first is a report about me. The second hands you a reason to reconsider what you're doing, even if you like the extra money more than you like me.

Promises make this especially clear. Suppose I agree to do something, benefit from your relying on it, and then change my mind. If my obligation lasts exactly as long as keeping it suits me, then whatever you received when I made the promise, it wasn't something you could build on.

This is why correction matters to me. Changing my taste in music doesn't mean my old taste was a mistake. Changing my mind about cheating an uninformed customer might mean I learned something about what he was owed. The question for the series is whether that difference is real, or whether I just swapped one attitude for another.

A fair process, conveniently interpreted

We don't have to leave the gap between principle and behavior hypothetical. People have measured it.

In 1997, C. Daniel Batson and his colleagues asked participants to divide two tasks between themselves and another person, one desirable and one dull. The experimenters supplied a coin and mentioned that flipping it would be the fairest way to decide. Ten of twenty participants flipped. Nine of those ten then assigned themselves the desirable task.^{1414. C. Daniel Batson, Diane Kobrynowicz, Jessica L. Dinnerstein, Hannah C. Kampf, and Angela D. Wilson, "In a Very Different Voice: Unmasking Moral Hypocrisy," Journal of Personality and Social Psychology 72, no. 6 (1997): 1335–1348. doi:10.1037/0022-3514.72.6.1335. [verify before publishing: whether the n = 20 coin result is Study 1 or Study 2, and name it here]}

The authors called that moral hypocrisy. They couldn't tell which individuals had fudged the result, and twenty undergraduates in a lab is not a census of the human race. The finding still gives us a concrete problem: adopting a fair procedure can provide cover for a decision that serves you anyway.

The trick with any check is what you let it correct. If I'm willing to call the process fair but not willing to accept an outcome that costs me, I've kept the appearance of a check and thrown away its function. It's a plumb line I only consult when the wall is already where I want it.

That doesn't establish moral realism. It does make the problem harder to dismiss as something only a cartoonishly crooked mechanic would face.

The alternatives aren't "do what feels good"

David Hume gave sentiment the central job in moral judgment. In Book III of A Treatise of Human Nature he argues that moral distinctions are "not derived from reason": reason can tell us about circumstances and consequences, but it is sentiment that marks the difference between virtue and vice.^{11. David Hume, A Treatise of Human Nature (1739–40), Book III, Part I, Sections I–II. The is–ought passage closes Section I. davidhume.org · Project Gutenberg.} In other words, you can know every fact about a situation and still need something that isn't a fact in order to call it wrong.

That's a serious problem, not a slogan. If someone understood every detail of another person's suffering and felt no concern at all, we'd question his moral understanding and not just his manners. Knowing the facts and caring what they mean are related, but they aren't the same thing. And Hume's account is more demanding than following whatever impulse shows up: sentiment can be examined and disciplined. Starting with feeling doesn't commit you to treating every feeling as authoritative.

A. J. Ayer went further. In Language, Truth and Logic he treats moral utterances as expressions of feeling rather than statements that can be true or false.^{22. A. J. Ayer, Language, Truth and Logic, 2nd ed. (1946), Chapter 6, "Critique of Ethics and Theology." Dover reprint, ISBN 9780486200101.} On his account, "cheating is wrong" works something like a tone of voice. It isn't even a report that the speaker disapproves of cheating; expressing an attitude and describing it are different things, and Ayer is explicit that his view is not that kind of subjectivism.

Contemporary expressivists have built much more elaborate machinery. Allan Gibbard ties judgments about what we ought to do to planning how to live.^{33. Allan Gibbard, Thinking How to Live (Harvard University Press, 2003).} Simon Blackburn calls his program quasi-realism: start from attitudes and earn, step by step, the right to talk about moral truth, objectivity, and getting things wrong.^{44. Simon Blackburn, Essays in Quasi-Realism (Oxford University Press, 1993).} In On Truth, a short book about truth in general rather than metaethics, Blackburn describes moral truth as the axis that practical inquiry about how to live revolves around.^{55. Simon Blackburn, On Truth (Oxford University Press, 2018). Relevant passage excerpted at OUPblog, Sept. 2018.} That passage is about reasoning broadly, but it shows why giving up realism doesn't mean giving up on moral inquiry.

So none of these views stops anyone from condemning the mechanic or criticizing a whole community's standards. It would be convenient for my argument if rejecting realism meant accepting whatever society approves of. It doesn't.

The disagreement is about the explanation. Do our judgments respond to reasons with authority independent of our attitudes, or can that authority be explained starting from the attitudes themselves? Gibbard's later work argues that the strongest forms of expressivism and nonnaturalist realism end up affirming the same theses and differ only in how they explain them.^{66. Allan Gibbard, Meaning and Normativity (Oxford University Press, 2012). The convergence claim is summarized in the publisher's description.} (Nonnaturalist realism holds that moral truths aren't reducible to the facts natural science describes.) Two theories can agree on every verdict and still disagree about what makes the verdicts true.

Which means finding an obvious injustice won't do the work by itself. I owe an argument that realism explains it better.

What would change the judgment?

Let's get back to the mechanic. The argument against him is short:

1. Knowingly deceiving a customer into paying for work the vehicle doesn't need is wrong.
2. The mechanic knowingly did that.
3. Therefore, the mechanic acted wrongly.

There are two places to push. Did he do what we're accusing him of? And why accept the first premise?

Maybe the diagnosis was an honest mistake. Maybe the alternator really was going and we're missing part of the story. If those facts change, the judgment has to change with them. Being pleased with an accusation doesn't qualify it as a good one.

The moral premise needs support too. The customer needed accurate information to make a decision about his own money. He paid for expertise because he didn't have it. The mechanic used that dependence to collect a payment the customer would never have authorized if he'd understood the problem. Those details explain the objection better than "most people dislike fraud."

Ayer would allow this much. Where two people already share a value, factual argument can move them: if we both condemn fraud, finding out whether this transaction involved fraud settles our dispute. What his account won't allow is an argument that establishes the value itself. If we agree on every fact and still differ at the level of basic values, Ayer has nothing further to offer. Later expressivists take on the bigger job of explaining moral reasoning as reasoning.

They also face a technical problem worth naming, though I don't have time today to go deep on philosophy of language. The Frege–Geach problem, associated with Peter Geach's "Assertion," asks how a moral claim keeps its meaning inside an argument where nobody is asserting it.^{77. Peter T. Geach, "Assertion," The Philosophical Review 74, no. 4 (1965): 449–465. doi:10.2307/2183123.} I can consider "if the deception was wrong, then encouraging it was wrong" without asserting that the deception was wrong; I haven't expressed disapproval of anything. An account of moral language has to explain how that reasoning works.

Expressivists have worked on this at length, and valid moral arguments don't establish realism on their own; somebody who thinks every moral belief is false can still recognize the logical form. The question I care about is why the customer's interests constrain the judgment at all. Realism says his dependence and the lie supplied reasons against the mechanic's conduct whether or not either man noticed them. An expressivist owes a different account of how those interests bind. The form of the argument doesn't pick between them.

What happens when the institution approves

The mechanic is one man. Institutional approval is the larger problem, because the people responsible have authority, procedure, and professional standing behind them.

From 1932 to 1972 the U.S. Public Health Service ran its study of untreated syphilis at Tuskegee. It began with 600 Black men, 399 with syphilis and 201 without. The researchers did not obtain informed consent. The men were told they were being treated for "bad blood," and they were not offered penicillin after it became the standard treatment.^{88. Centers for Disease Control and Prevention, "The Untreated Syphilis Study at Tuskegee Timeline."}

Those men depended on people who had medical knowledge and controlled the study. They were denied information and treatment they needed. Institutional authorization did not make either failure acceptable.

When I say the participants were wronged, I mean they were wronged during the study. Public exposure in 1972 changed what other people knew and were willing to say out loud. It didn't create the men's claim to honest treatment.

An expressivist can affirm every word of that, so the case doesn't settle the philosophical dispute. What it does is set the bar for any adequate account: explain how an institution's accepted standards could themselves be wrong. Realism clears it by locating the reasons in the men (their vulnerability, their need for care, the lie told to them), which were reasons for the researchers to act differently long before anyone admitted it. I find that the more natural explanation. Showing it's the BETTER explanation means comparing how each theory accounts for the need to change, and that's work for later episodes, not a sentence here.

Then turn the same demand on ourselves. Condemning an injustice that's already been exposed is easy. Examining a practice the people around us currently defend is not, and if moral reform is possible at all, some of what we defend is probably on the list.

Your conscience has operating conditions

Conscience deserves attention. It does not get an exemption from inspection.

I can feel certain about a judgment because I understand the circumstances well. I can also feel certain because I've heard the same explanation for years, because I like the person involved, or because questioning it would complicate my week. The certainty feels the same from the inside either way.

And even when I recognize an obligation, acting on it is a separate task.

In 1973, John Darley and Daniel Batson sent seminary students, one at a time, between two buildings past a man slumped in a doorway who appeared to need help. They varied how hurried the students believed they were. Of the forty students in the analysis, 63% stopped under low hurry, 45% under intermediate hurry, and 10% under high hurry. Whether the student was on his way to give a talk on the Good Samaritan made no reliable difference.^{1515. John M. Darley and C. Daniel Batson, "'From Jerusalem to Jericho': A Study of Situational and Dispositional Variables in Helping Behavior," Journal of Personality and Social Psychology 27, no. 1 (1973): 100–108. doi:10.1037/h0034449.}

The authors discussed competing obligations and narrowed attention as explanations; they didn't conclude that every student who kept walking was callous. Forty people split three ways is a small study. It still makes a useful distinction visible: professing a commitment and acting on it under pressure are two different measurements.

If I want to understand my own moral conduct, I should look at the conditions I'm making decisions in. What am I rushing past? Which obligation gets crowded out first? Am I treating a problem as irrelevant because it is, or because acknowledging it would blow up my schedule?

Those questions don't let me calculate moral truth. They help me see why I might fail to respond to a reason that's already there.

This is where the plumb line comes back. A good reference doesn't flatter the person using it. If I believe moral truth exists, that should make me more willing to check the wall I built, not less.

Sometimes the missing information is a choice

The mechanic's customer couldn't check the diagnosis. We aren't always in that position. Sometimes the information is right there and we'd rather leave it alone.

A 2023 meta-analysis by Linh Vu and colleagues pooled 33,603 decisions from 6,531 participants across 56 treatment effects to study willful ignorance.^{1616. Linh Vu, Ivan Soraperra, Margarita Leib, Joël van der Weele, and Shaul Shalvi, "Ignorance by Choice: A Meta-Analytic Review of the Underlying Motives of Willful Ignorance and Its Consequences," Psychological Bulletin 149, no. 9–10 (2023): 611–635. doi:10.1037/bul0000398.} Across the included experiments, about 40% of people avoided easily available information about what their choice would do to someone else. Where avoidance was allowed, altruistic behavior was 15.6 percentage points lower than when the information was simply provided.

That's no universal selfishness rate, but it is evidence that not knowing can be part of the decision.

Which makes "I didn't know" a more complicated sentence. Did I lack access, misread what I saw, or decline to ask a question I suspected would be inconvenient? Those are different failures and they call for different fixes.

So the practical job has at least three parts. First, the quality of the judgment itself. Second, the conditions I'm acting under. Third, whether I'm willing to find out what my choices do to other people. Endorsing a principle leaves all three undone.

What the evidence doesn't settle

None of those studies proves moral facts exist. A realist and an expressivist can both read them with interest. The findings are about human conduct; the philosophical argument is about what justifies judging that conduct.

It's worth keeping the two connected without confusing them. If moral reasons exist independently of our preferences, the behavioral evidence explains how we fail to see or act on them. If one of the alternatives is right, it still needs an account of the same failures and of what correcting them amounts to.

Realism has work left to do, and I'd rather name it than pretend otherwise. The first problem is knowledge: how do we tell a reliable moral judgment from a prejudice that happens to feel authoritative? Disagreement doesn't help. Some conflicts come from different beliefs about what happened. Others survive after every fact is on the table. Saying there IS a truth doesn't tell you which of us found it.

Hume's warning about moving from "is" to "ought" sits in the same section of the Treatise as his argument that moral distinctions don't come from reason.^{11. David Hume, A Treatise of Human Nature (1739–40), Book III, Part I, Sections I–II. The is–ought passage closes Section I. davidhume.org · Project Gutenberg.} Discovering that a practice is common, profitable, or useful to an institution does not show it's justified. You have to explain the connection between the facts in the premises and the obligation in the conclusion. Different realists handle that differently: a nonnaturalist can say moral conclusions need moral premises and leave it there, while a naturalistic realist, who takes moral facts to be part of the natural world, has to say why the natural facts she picked out carry moral weight. Citing Hume raises the problem. It doesn't resolve it.

Evolution raises another. If our evaluative dispositions were shaped by pressures that favored survival and reproduction, why expect them to track independent moral truth? Sharon Street presses that question in "A Darwinian Dilemma for Realist Theories of Value."^{99. Sharon Street, "A Darwinian Dilemma for Realist Theories of Value," Philosophical Studies 127, no. 1 (2006): 109–166. doi:10.1007/s11098-005-1726-6.} A realist needs an account of why the relevant capacities are reliable, or of how reflection corrects them. Street isn't a nihilist about any of this, by the way; she goes on to develop a constructivism about reasons, where reasons depend on an agent's evaluative standpoint but particular judgments stay open to correction.^{1010. Sharon Street, "Constructivism about Reasons," in Oxford Studies in Metaethics, vol. 3, ed. Russ Shafer-Landau (Oxford University Press, 2008), 207–245.} Realism isn't the only theory that can make sense of changing your mind.

Those are substantial objections, and I haven't answered them by describing a crooked mechanic, a federal study, and three lab results. What I've tried to establish is the problem a fuller account has to address: another person's circumstances seem to give us reasons to act differently even when we'd prefer not to.

Where my faith enters

I'm a Latter-day Saint, and that shapes how I approach moral knowledge. You should be able to see exactly where it enters the argument.

Moroni 7:16 says the Spirit of Christ is given to every man "that he may know good from evil."^{1111. Moroni 7:16–19.} I read that as a reason to expect moral insight outside my own community. Someone doesn't need my vocabulary or my church membership to catch a wrong I've missed. The verses right after it put the burden back on the person judging: verse 18 warns, "see that ye do not judge wrongfully," and verse 19 says to "search diligently in the light of Christ."^{1111. Moroni 7:16–19.} In other words, the gift arrives in fallible hands, and the habits and culture of those hands can bend what gets done with it. Having moral light doesn't make every strong feeling I have a reliable conclusion.

That matters in light of the studies above. Religious conviction doesn't remove the need to examine what I actually do, what I overlook, or what I'd rather not find out.

Paul makes a similar move in Romans 2:14–15, describing Gentiles who "do by nature the things contained in the law" and whose conscience bears witness.^{1212. Romans 2:14–15.} Doctrine and Covenants 93:24 defines truth as "knowledge of things as they are, and as they were, and as they are to come," and verse 30 adds that truth is "independent in that sphere in which God has placed it."^{1313. Doctrine and Covenants 93:24–30.} I take those to support the view that truth is something we learn to recognize. They don't settle every philosophical question about God, moral law, and obligation, and I'm not going to pretend they do.

If you don't accept those scriptures, you have no reason to accept my conclusion on their authority, and the philosophical argument has to stand on its own reasons. This section is here so you can see my starting point and what I think it requires of me.

Repentance is what that requirement looks like in practice. I can sincerely believe I acted well and later find out I wronged someone. Explaining my intentions might help him understand what happened; it doesn't repair the damage. I may have to admit the wrong, make it right, and act differently afterward.

Starting with our own judgments

"Is this right, or do I just want it?" is a decent place to start. It's incomplete, though. Sometimes our wants are well formed; we want a thing because we recognize it's good. And sometimes what we call principle is a preference wearing better clothes.

The task is to examine the reasons connecting the judgment to the situation. T. M. Scanlon's What We Owe to Each Other is helpful here.^{1717. T. M. Scanlon, What We Owe to Each Other (Belknap Press of Harvard University Press, 1998).} His contractualism asks whether an action would be permitted by principles that no one could reasonably reject as a basis for informed, unforced general agreement. The test runs on the reasons people have for objecting, not on whether they happen to go along with the outcome.

Borrowing from that, plus the questions about factual error and self-interest from earlier, here is what I want to ask when a moral question gets hard:

1. Which facts would change my judgment if I learned I had them wrong?
2. What reason could I give the person who bears the cost of my decision?
3. Could they reasonably reject the principle I'm using to justify it?
4. What am I overlooking because the present arrangement benefits me?

The third one is about the other person's reasons, which is more than imagining whether I'd accept the decision in their place. These four borrow from Scanlon without reproducing his theory, and they don't resolve every disagreement. They do make it harder to use my own comfort as the measure of what I owe.

To recap: the mechanic's satisfaction can't erase the customer's claim to honesty; an institution's approval can't excuse its failure to treat people justly; and my confidence can't establish that I've understood either case correctly. I think moral realism gives the most persuasive account of those three facts, because it treats our judgments as answerable to reasons we didn't create.

That's the work I want this series to do, and the next episode takes up the most natural reply to everything above. Fine, moral claims can be true or false; so how do we find out which ones are? The modern answer is to set up a fair process and let the process decide: a vote, a hearing, a review board, a code. I work in an industry that runs on procedure, and I'm not against any of that. But we just watched twenty people adopt a fair coin and route around it, and the Tuskegee study ran for forty years with every approval it needed. A building can be code-compliant and still unsafe.

So Episode 2 asks whether a fair procedure guarantees a just outcome, or only a legitimate one, and why those two words don't mean the same thing.

Sources

1. David Hume, A Treatise of Human Nature (1739–40), Book III, Part I, Sections I–II. The is–ought passage closes Section I. davidhume.org · Project Gutenberg.↩

2. A. J. Ayer, Language, Truth and Logic, 2nd ed. (1946), Chapter 6, "Critique of Ethics and Theology." Dover reprint, ISBN 9780486200101.↩

3. Allan Gibbard, Thinking How to Live (Harvard University Press, 2003).↩

4. Simon Blackburn, Essays in Quasi-Realism (Oxford University Press, 1993).↩

5. Simon Blackburn, On Truth (Oxford University Press, 2018). Relevant passage excerpted at OUPblog, Sept. 2018.↩

6. Allan Gibbard, Meaning and Normativity (Oxford University Press, 2012). The convergence claim is summarized in the publisher's description.↩

7. Peter T. Geach, "Assertion," The Philosophical Review 74, no. 4 (1965): 449–465. doi:10.2307/2183123.↩

8. Centers for Disease Control and Prevention, "The Untreated Syphilis Study at Tuskegee Timeline."↩

9. Sharon Street, "A Darwinian Dilemma for Realist Theories of Value," Philosophical Studies 127, no. 1 (2006): 109–166. doi:10.1007/s11098-005-1726-6.↩

10. Sharon Street, "Constructivism about Reasons," in Oxford Studies in Metaethics, vol. 3, ed. Russ Shafer-Landau (Oxford University Press, 2008), 207–245.↩

11. Moroni 7:16–19.↩

12. Romans 2:14–15.↩

13. Doctrine and Covenants 93:24–30.↩

14. C. Daniel Batson, Diane Kobrynowicz, Jessica L. Dinnerstein, Hannah C. Kampf, and Angela D. Wilson, "In a Very Different Voice: Unmasking Moral Hypocrisy," Journal of Personality and Social Psychology 72, no. 6 (1997): 1335–1348. doi:10.1037/0022-3514.72.6.1335. [verify before publishing: whether the n = 20 coin result is Study 1 or Study 2, and name it here]↩

15. John M. Darley and C. Daniel Batson, "'From Jerusalem to Jericho': A Study of Situational and Dispositional Variables in Helping Behavior," Journal of Personality and Social Psychology 27, no. 1 (1973): 100–108. doi:10.1037/h0034449.↩

16. Linh Vu, Ivan Soraperra, Margarita Leib, Joël van der Weele, and Shaul Shalvi, "Ignorance by Choice: A Meta-Analytic Review of the Underlying Motives of Willful Ignorance and Its Consequences," Psychological Bulletin 149, no. 9–10 (2023): 611–635. doi:10.1037/bul0000398.↩

17. T. M. Scanlon, What We Owe to Each Other (Belknap Press of Harvard University Press, 1998).↩