How do you pick a teacher?

The Dunning-Kruger effect is a term for people who are too ignorant to recognize that they are mistaken. If you work in politics or public health, it’s easy to feel as though these people are everywhere, but it is possible for most of us to spot people or systems which are better at making certain judgements than we are. And if we find one, it’s rational to defer to its decisions.

This is essentially the kind of system G.K. Chesterton claimed Catholicism to be when he declared it to be a ‘truth-telling thing.’ He was claiming to have found a system that was so much more reliable than his own judgement that he felt it prudent to defer to the wisdom of the church not only when he was undecided, but even if his moral intuitions were in stark opposition to church teaching.

This sounds like a textbook case of organized religion asking believers to doff their brains as well as their hats when they stand before the altar, but I don’t find Chesterton’s position to be irrational in theory. When I read books on quantum theory written for laypeople, I usually have to try to reform my expectations and ideas. This isn’t because quantum theory is irrational, it’s because my intuitions and common sense are wrong. My normal modes of thought are a good-enough approximation of the world for day-to-day life, but they’re no more accurate than Ptolemy’s epicycles.

When Eliezer Yudkowsky talks about rationally deferring to the judgement of another, he uses the example of a computer programmer who is not very good at chess but is able to design a piece of software that plays very well. The programmer still doesn’t know how to play high level chess. He built a system that produces correct results in a way that he can’t replicate in his own mind. If he wants to win matches, he ought always defer to the computer’s suggestion.

It’s a lot easier for the programmer to feel confident that the software is a better chess player than he. It’s harder to figure out what kind of evidence could justifiably give Chesterton that kind of confidence in the church’s moral judgments.

It’s hard to look at any particular move the chess computer makes and judge whether or not it’s a good move, but we don’t have to evaluate the process. We can just look at how many wins the program racks up against different levels of opponents. It’s even easier if we’re trying to build programs to categorize data (as I am in my machine learning class). Every time the algorithm is tweaked, we check it for accuracy.

But if you try to apply this kind of test to a moral truth-telling thing, you run into two big problems. First: there’s the problem of overfitting. In computer science, it’s bad practice to train your prediction program on all the data you have. It’s easy to assume that more data must be better, but using everything you’ve got is going to make your program really good at categorizing the data you already have and really bad at predicting anything that wasn’t in your original data set. After all, the most accurate program is one that just stores all the data you gave it in a big lookup table and returns the numbers you inputted. Not a big improvement. In computer science, you avoid this problem by withholding a certain subset of the data from the computer when it’s learning. When it’s generated a model, you test that formula for accuracy by applying it to the datapoints you held in reserve.

You can’t pull this off with moral quandaries. A church or a philosophy isn’t isolated from the world, so you can’t hold some conundrums back and then try to use them as test cases for the first principles the group you’re evaluating has on offer. So, instead you’re left suspicious that any teaching, especially one sourced from a long, complicated book might have more to do with having an intuition about an answer and hunting up something in your canon that supports it. That hardly likely to give you the confidence to throw over your intuitions for the dogma of a new teacher.

That’s a technical issue, but there’s an ever bigger stumbling block for an aspiring pupil. Unlike the chess example, where it’s easy to keep score by number of wins, it’s hard to figure out how you judge one moral system as more accurate than another.

I’m not preaching relativism, some moral systems take themselves out of the race. There are enough commonly-held moral intuitions which I assign a high level of confidence that I feel comfortable disqualifying any system that doesn’t preach them. To name a few names: solepcism, objectivism, and dark kantianism. But if I consider only those systems that can coexist with my list of unshakeable moral precepts, I haven’t found a useful moral system, I’ve just found the ethical equivalent of a generative set — a summary of my list. To put it formally, I’ve come up with a set of first principles/axioms that I know I can derive some true theorems from, but I have no idea whether this set of axioms only generates true theorems. When two systems that clear my initial bar diverge, I don’t have a good way to pick the winner.

The best schema I have is to look for systems that usually turn out to be right even when I think they’re wrong. That’s the kind of evidence that Chesterton and more modern-day converts like Jennifer Fulwiler of Conversion Diary claim to have found. That’s not been my experience of Catholicism, and most of the predictions it offers me are stuck behind the firewall of faith, impossible to test until you’ve assented at least to the point of theism.

"Well, I would love to know if you now believe that homosexuality is intrinsically disordered."

Go Ahead, Tell Me What’s Wrong ..."
"Any chance of you ever addressing the evidence that led you to accept the truth ..."

Letting Go of the Goal of ..."
""Wow, an unevidenced assertion from a religious dipshite. "Your quotes are the evidence and reason ..."

This is my last post for ..."
""Congrats on leaving your brain behind!"Comments like yours are why lots of atheists leave atheism. ..."

This is my last post for ..."

Browse Our Archives

What Are Your Thoughts?leave a comment
  • Lukas Halim

    “A church or a philosophy isn’t isolated from the world, so you can’t hold some conundrums back and then try to use them as test cases for the first principles the group you’re evaluating has on offer.”
    For me, when I learn about Church teaching I come away thinking, “this is sort of like what I’ve always understood to be right, only it’s clearer and more balanced then what I would have come up with.” That’s not always the case, but it’s often the case. Or, think of immature Luke Skywalker meeting Yoda – through various circumstances, Luke comes to recongize Yoda’s wisdom.

  • Well, I’d suggest looking into the Church’s teaching on human sexuality. Everything our society has told you for pretty much all your life completely contradicts what the Church teaches. Look at outcomes when you compare sex lives lived the way the Church teaches, compared to how the world says they should be lived (particularly when you extrapolate from premises to logical conclusions).

    That would, of course, only be a starting example. You’d need to go through the other issues of morality where you disagree with the Church.

  • Gilbert

    The problem is real in principle but there are ways to rule out a lot more theories than one might think.

    One way is simply checking for logical consistency. For example, though many famous academic philosophers seem to have missed the memo, Arrow’s Theorem kills off all variants of utilitarianism.

    Another way is to look at predictions the various theories make in areas you would not normally examine. The kind of predictions that would disqualify a moral theory are very likely to be predictions we wouldn’t deduce from a moral system in everyday life, simply because the correct answer is already obvious almost by definition.

    For example twelve year old me was a very liberal Catholic. So liberal I used to think abortion was acceptable. I thought so, because I thought moral personhood depended on intelligence. This is about the only question on which I recall changing my opinion on something that central that quickly: Someone pointed out that I did accord moral rights to cognitively disabled people. That contradiction probably isn’t news to anyone present, but very young me had never thought of it before. I rejected the intelligence theory of personhood on the spot, and, unable to find any other viable criterion except biological humanity, became pro-life a few days later. Maybe it was the first step on the slippery slope to my present reactionary ways.

    I think I would have had only two other ways out: One would have been biting the bullet and go all Peter Singer. In other words cutting of my nose to spite my face. The other option would have been to declare personhood all vague and mushy and then look for a “pragmatic” rule. That unfortunately seems to be the most popular solution. It’s a general approach too: One can always declare a question gray and then be pragmatic, it’s relativism on a case by case basis. This kind of cowardice, of course, makes it impossible to throw out any ethical theories because they can all be saved that way.

    Whatever you think of the concrete example, this is a second way of ruling out ethical theories: Think of the most absurd consequences you can possibly derive from them and then refuse to be pragmatic. This is a large part of why people deal in far-fetched moral thought experiments: They are practically relevant not because the imagined situations could happen but because they help rule out theories that also have practically relevant consequences. For example many people’s reactions to the Trolley problem rule out consequentialism. Also many a slippery slope argument goes into basically this direction.

    • I disagree with this on a fundamental level:

      It is not “relativism on a case-by-case basis” to recognise that sometimes, two principles can conflict. At that point, unless you operate a strict hierarchy of principles, which I think unwise* you must decide for that individual case which principle will be made “vague and mushy”, and which one stands.

      That’s part of living in the real world. If you hold more than one principle they will eventually conflict, and where now is your lovely black-and-white moral world? Better to recognise and accept that there will be shades of grey; then you’re more likely to deal with them appropriately.

      *I believe this to be unwise because you could theoretically deny all people liberty to save a single life, because you hold life to be a more important value than liberty. I would not feel that to be a good trade, even if it were my life being saved.

      • Gilbert

        The way you use it seems to limit “moral principle” to mean “good to be sought” or perhaps even more limitedly “interest”.

        Now of course there can be conflicts between goods to be sought in individual cases. And of course these cases require decisions on which good will take precedence. But then there are only two possibilities: Either we use some rule to make that decision or we don’t.

        The second possibility is moral relativism on a case-by-case basis, pretty much by definition. If we are not using any rule we are surely not using an objective one.

        In the first case however, whatever rule we use to make the decision is, also pretty much by definition, itself a moral principle*.

        As an analogy, if two forces would move an object in conflicting directions we are not satisfied wondering which force will win out in that individual case. Rather we look for a rule applying to the presence of both forces. If the forces are impressed to the same point of the object, that rule would be the principle of superposition. And that principle is, of course, itself a physical principle, thus making the system of physical principles (i.e. physics) non-conflicting.

        So no, moral principles are not limited to goods to be sought. And no, in their correct application they should not conflict. Because if such a conflict exists that’s just another way to say the system is contradictory and a contradictory system can’t be true.

        Of course this doesn’t prevent us from being morally conflicted about a practical problem, but that’s a different meaning of “conflict”.

        * Actually I don’t like the word principle here, because some of them may actually be derived from others. But you know what I mean.

    • I like the relativism. The grey area of protected human life isn’t a moral question at all, it’s an economic one. We should do things that have the most consistently productive results. Sometimes that means you terminate the fetus of an Einstein. But undeveloped babies can’t be worth more than adults that have had years of education and development already added to their base value.

      And since it’s grey, no it doesn’t apply to relatively valuing adults in a death camp fashion. It only applies to situations in which we as a state would take an individual’s decision away from them and say “you must care for this extra person”. It’s about assigning care and responsibility since the desire to save all life is the unlimited want and the energy of parents and the society to solve all their respective delimas is the limited means. And that means, it’s an economic question.

    • anon atheist

      So Catholicism would not be a viable moral system because of for example the inconsistency of forbidding condoms but allowing so-called natural methods for birth control.

      • Gilbert

        I don’t see how that should follow. A system doesn’t become logically inconsistent because it is inconsistent with some other system. So unless you can derive NFP=contraception from Catholicism you can’t use it to construct a contradiction in Catholicism.

    • Andrew G.

      Given the limited context of Arrow’s theorem, it’s pretty silly to regard it as “killing off” utilitarianism.

      In real life, people don’t just order their preferences, they have preferences of different weights. Also, there is nothing inherent in utilitarianism that requires the use of the kind of social choice function addressed by Arrow’s theorem and its weaker generalizations.

      Your abortion example is almost a caricature of binary thinking; surely it has occurred to you that (a) the criterion to use might not be “intelligence”, but “consciousness”; and (b) that even a fetus (or person) with full moral personhood does not necessarily have the right to the use of another person’s body in order to survive? (see the “Famous violinist” thought experiment)

      Your example of the trolley problem doesn’t work as a defeat of consequentialism either for the very simple reason that, in the usual formulation, the problem as posed fails to consider all the consequences. Specifically, in all the various cases of killing one person to save many, one has to consider not only the simple number of people killed but also the total effect on society of condoning, or condemning, the action. Nobody wants to live in a society in which random people get harvested for organs, regardless of how many lives get saved; the net effect of loss of trust, safety and justice easily accounts for the fact that people reject that scenario in contrast to accepting the scenario of the simple trolley switch, in spite of the fact that that kills and saves the same number of people.

      • Gilbert

        On Arrow’s theorem “preferences of different weights” don’t help, because they still imply an ordering, whatever system is used to aggregate them will still imply an aggregate ordering and the same problem will still apply between the individual and aggregate orderings. Adding further properties simply cannot fix the contradictions between the properties already present. As to doing utilitarianism without “use of the kind of social choice function addressed by Arrow’s theorem and its weaker generalizations” I would like some examples of that. Not explicitly mentioning a concept doesn’t necessarily mean it’s not used, after all kids use fractions long before they know what a field is.

        The abortion example was an illustration of a general method of discarding moral theories and not the main point. It isn’t worked out in as much detail as it would be if it where what I was actually talking about. Now since I gave the example and actually do stand by it I will respond to your objections this once but I don’t want to turn this into a fully general abortion discussion so I will not have further replies to your response on this comment. So briefly: (a) doesn’t help, because the part of consciousness that is unique to humans is dependent on intelligence. (b) doesn’t help because the famous violinist example actually isn’t analogous to abortion. To make it actually analogous we could change it thus: Foreseeing I may cut it the violin appreciation society covered the tube connecting us in invincible armor. The armor also covers the places at which the tube is connected to each of us. But for my safety they included an automatic release activating if the famous violinist dies. There is a hammer nearby. The actually analogous question is “Can I use the hammer to smash the violinists head?” and the actually analogous answer is “no”.

        In the trolley problem I was thinking of the difference between flipping the switch and throwing the fat man, to which your objections don’t apply. Anyway they don’t work in the organ harvesting context either. Trust and safety are only negatively affected if one cause of harm is counted as worse than others, so they reduce to justice. And including justice in the consequences for purposes of constructing a conventionalist morality is just question-begging.

        • Andrew G.

          Arrow’s theorem only shows that there isn’t a good way to condense a set of ordered preferences into a single preference ordering. Once you include preference strengths, it does not apply because whatever function you use to produce the result is expected to take those strengths into account, thus producing a different result from the ordinal preference function.

          Or to put it another way, if person A prefers to have shrimp for dinner rather than pollock, and person B prefers to not die of starvation, we don’t expect a utilitarian ethical system to treat these equally (as an ordinal preference system does), but to take into account the relative importance.

          Regarding the trolley problem, my objection does in fact still apply; but to explain why, I first need to explain why introducing the concept of justice is not question-begging. By “justice” in this context I am referring not to any abstract concept of justice that might emerge from an ethical system, but the actually existing moral emotion present and measurable in the vast majority of humans. People demonstrably do not wish to live in a society they perceive as being unjust or unfair beyond certain limits, and are less happy if forced to do so; thus any factually grounded moral system must take these concepts into account. (That’s not to say that the moral system must always conform to people’s moral emotions – that would be the naturalistic fallacy, and in any event provably results in suboptimal outcomes.)

          With that in place, we can see that while the fat man has the choice to sacrifice himself to save others, people naturally resist the idea that someone not already involved in a dangerous situation can be forced to make that sacrifice at the decision of another. Furthermore, people naturally recognize that the consequences of an action are wider than are typically taken into account in the description of the problem; that no moral issue can be isolated from the larger society in which people live.

          • Gilbert

            On utilitarianism, I heard you the first time, but your objection still doesn’t work.

            So OK, in addition to a preference order your individuals also have some kind of structure accounting for preference strength. And OK, your social welfare function is parametrized on those structures. Still, restricting your range of functions can’t help you escape from properties they all share. So your aggregate order (or its ordinal projection) will still violate some of the principles incompatible according to Arrow’s theorem. And you say it will do so because the principle it overrode would have favored a weaker preference over a stronger one. Let’s look at how that could work. If the independence of irrelevant alternatives was violated your out won’t help, because whatever irrelevant preference affected your result, by definition, didn’t lead to it’s preferred state being chosen. So you have the problem regardless of how week or strong it was. If there was a dictator the dictator’s strongest preference sure won out, but so did his weakest one and the preferences of everyone else were irrelevant. Unless all of everyone else’s preferences where very weak you can’t use the excuse of having overridden weak preferences for strong ones. And if the Pareto criterion was violated, the preference overridden was by definition unanimous except for indifference, so there is nobody whose stronger preference could have been honored in this.

            The one out you have is restricting the domain by assuming people’s strongest preferences never conflict. But then the utilitarian processing step added nothing to the result. Then you are a true-strong-preferentian rather than an utilitarian and don’t need any aggregation to produce the result.

            On consequentialism and the trolley problem it’s basically the same as with the domain restriction for utilitarianism:
            It lets you formally remain a consequentialist. But only if in weighting your consequences the remote chance of someone discovering something has been done in violation of their moral emotions outweighs multiple lifes. In that case every other consequence is irrelevant to the result and you are for all practical purposes a moral intuitionist. Whatever consequentialist rule gets applied is just a superfluous step blessing the result already obtained by other means.

          • Andrew G.

            Regarding Arrow’s theorem, I’m not restricting the range of functions but the reverse.

            Rather than argue the point, I will simply provide a counterexample: range voting, where preferences are indicated by simply assigning a weight to each choice, satisfies all of Arrow’s criteria and thus proves that the theorem does not apply to choice functions not based on ordinal ranking.

          • Gilbert

            While increasingly off-topic, this is also getting increasingly fun, so I’ll go into more depth than most people will be interested in:

            Range voting, like all cardinal methods, only seems to avoid the problem because the redefinition of “independence of irrelevant alternatives” its proponents like to rely on weakens the criterion by sweeping the scaling problem under the carpet. In fact Arrow’s definition of a social welfare function wasn’t, as some zelous advocates of range voting claim, “silly” but rather influenced by the fact that by then economists had mostly already given up on cardinal utility.

            To briefly explain it to the audience, the problem is this:
            Suppose everyone has an individual cardinal utility function mapping all possible world-states to the real numbers as a measure of how much they like it. Now you might think you can get a total utility by adding these numbers up.

            The problem is, my preferences and their relative strengths remain unchanged under any linear transformation of that function. If, for example, you just half my utility function both my preferences and all compromises I would make remain exactly the same. If I previously thought having pollock rather then shrimps for dinner was 1/50th as bad as giving up dinner entirely*, I would still think that, because the utilities of both choices have been halfed. And since my utility function is different from yours, we can’t just agree to use the same factor and let in cancel out, because that operation isn’t even defined.

            So before we can aggregate the utility functions we have to fix one point of everyones utility function to a given real number**. By the way, good luck doing that in any objective way.

            Of course this is a problem strictly in addition to those already raised by Arrow’s theorem. But here comes the trick:

            In range voting we do impose such a scaling by defining the range in which scores can fall. Unless they decide for a situation mathematically strictly equivalent to their vote counting less then someone else’s, the voters will rank their favorite candidate at the top of the range and their least favorite candidate at the bottom. For example on a scale from 0 to 100 the worst candidate gets 0 and the best one 100, thus effecting the scaling for us.*** Now for the clinger: Then we redefine independence of irrelevant alternatives to mean independence from the scores given to irrelevant alternatives, thus ignoring the rescaling we have just caused to be done. Yes, if someone isolatedly changes the score they assigned to a non-winning candidate and that doesn’t change that candidate winning or not winning it doesn’t change the result at all. But suppose a candidate one of the voters finds better or worse than all the old ones turns up. Then the reference points of our scaling have changed and all scores assigned by that voter will be affected. And, even if the new candidate doesn’t win, that can include the score of the until then winning candidate, thus causing them to loose, making our voter worse of because of the appearance of the irrelevant alternative.

            Now we can say the voting system is only the part transforming the actual ballots into the result and the necessary scaling shouldn’t count because the system is designed not to record it. Note how that would also make FPTP into a perfectly fair voting system. But even that doesn’t help for utilitarianism purposes, because then the actual preference aggregation scheme includes voting and another step, and while the voting system might, the complete preference aggregation system doesn’t satisfy the independence of irrelevant alternatives and therefore doesn’t violate Arrow’s theorem. Which comes at no surprise, because no preference aggregation system can.

            * actually I’d prefer the pollock, but let’s stay in the spirit…
            ** actually a second point, because the point of utility 0 isn’t affected by rescalings
            *** this actually fixes two points of the utility function, thus respecting it only up to affine transformations. This might raise additional complications, because, if the differences are extreme, a scrupulously honest voter might actually be unable to preserve their actual relative preference strengths. But in voting if not for utilitarianism purposes that problem is moot, because (a) there is no other way to do it and (b) in reality people wouldn’t vote according to their actual preference strengths anyway, but rather so as to optimize the result of the election.

          • Gilbert

            Arghh, that was more italics than I had wanted.

            Any chance of bringing back previews like there where at the old site? I’m pretty sure WordPress has plugins offering that function.

          • Andrew G.

            You’ve fallen into the common trap of interpreting IIA in terms of the English description rather than in the actual formal definition. (Even Arrow contributed to this misunderstanding.)

            IIA doesn’t tell you anything about what happens if you add or remove options from the agenda. What it tells you is that given two profiles on the same agenda, which differ only in terms of the ranking of choice C, then the relative order of A and B in the result must be the same between both profiles.

            This formal definition is used in exactly that way as part of the proof of the theorem, so you can’t argue in terms of loose English definitions and assume that they still apply.

            Arrow’s formal definition of IIA does not apply when considering non-ordinal systems, so you have to replace it with some other axiom. Of the possible alternatives, range voting passes some and fails others, while an honest utility function (which wouldn’t have the upper bound restriction of range voting) passes all of them.

            As for ordinal vs. cardinal preferences, the fact that economists reject cardinal preferences is if anything a recommendation rather than the reverse – economics is a discipline founded on fiction (both as to the nature of money and markets, and the nature of the people involved), and only in very recent times has it made any significant progress in shedding the falsehoods and engaging with reality. Also significant in this context is the fact that economics as currently taught is an inherently unethical discipline; people who study economics become less personally ethical. (This has been measured in controlled trials with before-and-after testing, not just by after-the-fact surveys that might have been subject to selection bias.)

          • Gilbert

            Neither of us is going to convince each other and the audience left days ago. So I’ll offer this as my last take, and then you can have the last word.

            On economics: The study of this kind of stuff is called welfare economics, and, as the name would imply, a sub-discipline of economics. So while I have my methodological doubts about the proofs of economists being evil that question is also completely irrelevant, I’m offering them as experts rather than examples. Basically the paradigm of cardinal utility had already failed, because its paradoxies were known even before Arrow’s theorem. Now as to the general critique of economics, to be very blunt I think it is the left wing version of the right wing rejection of climatology, both basically motivated by a desire to eliminate those results of science that might conflict with the respective side’s politics. Yes, economics deals in fictions. So does every science. Yes, in economics 101 you will learn about a homo oeconomicus who isn’t real. Likewise in physics 101 you will learn about a point mass which isn’t real either. But both are approximations good enough to make some useful predictions.

            As for my use of the wrong formulation of IIA: Yep, you caught me there. But it isn’t a real problem, because my point works in exactly the same way if the “new” candidate is not new but previously rated somewhere in the middle of pack. Since that kind of construction is basically always possible the common trap just isn’t that trappy.

            On extending IIA to cardinal methods:
            Well, there may be different versions, but in this context we are bound to one. Remember I talked about an order with preference strengths inducing one without such strengths and how restricting the functions mapping that induced orders to the aggregated one couldn’t remove properties they all shared. You offered range voting specifically as a counterexample to that claim. In that context the extension is perfectly clear: The cardinal social welfare function is IIA iff its ordinal projection is. And range voting fails.

            Unboundedness won’t save your honest utility function, because the scaling isn’t necessary only to establish a bound, but also to render different people’s cardinal utility functions comparable at all.

  • Something else that should be mentioned (and that underscores your point, Leah, about what could possibly give Chesterton such confidence in the church’s teachings) is that religious morality is expressly divorced from any observable real-world consequences. The success of the chess-playing program is visible, measurable, and brooks no argument; but the Catholic church, like other faiths, claims that most of the benefits of following its teachings accrue solely in the afterlife.

    One example of this is the Catholic teaching that, if a woman’s life is endangered by complications arising from pregnancy, she should prefer dying to getting an abortion, even if this means the fetus will die also. How could one judge this to be a good or moral position without certain knowledge that Catholic beliefs about the afterlife are true?

    • Agreed. Catholicism cannot be thought of as merely an earthly ethical system – it’s a system of everything seen and unseen. So what the Church says about how to behave is only authoritative in the context of trusting other Catholic teachings about the unseen. Crucially that the Vatican has privileged access to the thoughts of God, depicted in the Old Testament (considered dubious by non-Catholics).

    • I have to say that you are wrong here. First on the Olmsted piece – he diverged from a long line of Catholic medical teaching on that subject. The death of a fetus is allowed under certain circumstances if it is foreseeable but not intended, as per the principle of double effect. He effectively usurped the tradition.

      Second, much of Church argumentation about moral subjects relies more on natural law than appeal to an afterlife. In other words, there will be effects in this life, as in the idea that “the truth of the natural law is found in the wounds of those who do not follow it.” IOW disobey at your own peril. You can eat fiberglass insulation if you like, but it won’t be good for you. Lie, cheat, and steal, and you will not fare well in this life.

      • Andrew G.

        On the published facts of the case, the doctrine of double effect doesn’t apply – the woman wasn’t given other treatment that had the side effect of causing abortion, she was intentionally given an abortion in order to eliminate the strain that pregnancy puts on the body. There was no question of the fetus’ death being “unintended”.

        As for your argument about natural law, that is already contradicted by the facts, especially as regards homosexuality.

        • Hi Andrew,

          Do you know a place where the facts of the case are published? As I mentioned to Joe below I was under the impression that they were not public, but I may be wrong. As for natural law, I do not understand what you wrote. Say more.

          • Andrew G.

            Very few details are public due to privacy law. The most detailed public report I know of is this one from AZcentral.

            But in any event this is just one case amongst many; there are enough other cases where no medical condition other than simple pregnancy was involved to make the point. The 9-year-old girl in Brazil for example.

          • Andrew G.

            Regarding natural law, the Church position is that homosexuality is disordered. Your claim was that this would result in “effects in this life”. Evidence disagrees: homosexuals who accept their sexuality have better lives than ones who attempt to follow the Church’s teachings.

      • Even under the most elastic definition of “natural law”, I see no way to use that concept to justify the church’s position, which is this: If a pregnant woman’s life can be saved by abortion, but if both the woman and the fetus will die otherwise, it’s better to let both die than to save one.

        • Patrick

          What you have to understand about natural law is that its a very, very, shallow system. Its not deep. Its nothing like deep. Its a half dried up puddle.

          Moral systems tend to involve axiomatic reasoning. That is, beginning with some set of premises, and then reasoning towards conclusions that follow from those premises. For example, you might begin with the premise “I care about human well being” and reason to a sort of rough utilitarianism. This works pretty well, and allows moral communication, because if you have two people who roughly share premises, they can discuss what does or doesn’t follow from them.

          So the structure in that context is a “deep” belief (something like, “human suffering and happiness matter to me”), that generates surface beliefs (“we should have a tax system that this this progressive, but no more”).

          But in natural law, you haven’t got any of that. Everything’s right on the surface. Why is it better to let two people die rather than to save one? Because. SHUT UP. Just because. Why is the death of a fertilized egg due to birth control a big deal, but the death of a fertilized egg due to NFP a total non-issue? Just because. SHUT UP. Its the natural order of the universe. Its a brute fact. Quit asking questions, there are no questions beyond this point.

          Nothing flows from anything else. Nothing is linked. There are no underlying values. Everything is on the surface. As a result, it can’t contradict itself.

          • Gilbert

            You not having bothered to study a question doesn’t prove there is nothing to study.

          • Yeah, you have no idea what you are talking about. Talk to Aristotle and Cicero about it.

        • Construed purely rationalistically (i.e. Kantian) natural law states in this case (and which is also a biblical principle from Romans 3:8) “evil cannot be done that good may come of it.” Accepting evil to do good is an exception which destroys principled ethics. Slopes become slippery. One irrationality ruins the system.

          Why is that “natural law”? Because humans are rational animals. There are as many kinds of “natural law” as there are ideas of human nature. “Natural law” merely states that something about human nature is normative, it says nothing about exactly what that normative aspect is until you get into a specific school of natural law.

          With regards to the pregnancy, is it better to let one die rather than both? Of course. The question is HOW it is done. It cannot be ends justifying the means, it must be a double effect. The difference is not in end result, but in the way it is done – it has to be done in a way that does not introduce an irrationality into the system.

    • Gilbert

      Suppose you and a friend are riding a hot air balloon over a body of very cold water. Through some accident both some of your fuel and your radio get dropped. While it would be enough for one, the remaining fuel is not enough to carry the weight of both of you to the shore. There is no realistic chance of rescue. It looks like both of you will fall into the freezing water and die there.

      Are you morally allowed to throw your friend into the water thus saving yourself? If yes, are you sure no atheist could disagree with you? Does the answer depend on the existence of an afterlife?

      • Your analogy fails, and here’s why: In the case of ectopic pregnancy or other life-threatening complications, there is one party that poses a lethal threat to the other and that cannot survive, regardless of what actions are or are not taken. The only question is whether the other party will die as well.

        It’s hard to rework your analogy to accommodate this, but let me try: Let’s say that my friend in the balloon gondola has a ticking time bomb implanted in his body that can’t be defused. If I throw him overboard (or if he volunteers to jump), he’ll die, but my life will be saved. If I don’t, he’ll die anyway when the bomb goes off and he’ll take me with him.

        Am I morally allowed to throw him overboard, as an act of self-defense? Yes, I think so.

        • Joe

          The Church I believe would say that you are within your rights to toss the man overboard, however you do not have the right to crush his skull and dismember him before hand. You do not have the right to directly kill an innocent human being. In the case of a pregnancy were the fetus is developing in the fallopian tube(I can’t recall the medical term for this) the pregnancy can be terminated. This is done by removing the fallopian tube with the baby in it. The child is not directly killed but allowed to die a natural death just as it would have with in the mothers body and at the same time saving the life of the mother. I have talked to several doctors and they have assured me that their is no know medical reason for an abortion(direct killing of and innocent) The problem is most doctors see abortion as a quick fix as was probably the case in Arizona. Unfortunately the modern medical mentality is cost effectiveness and risk management and sometimes that means more abortions most doctors don’t want to put forth the effort to save both mother and baby

          • Joe

            Here is a link that will help you

          • Andrew G.

            There is at least one excellent medical reason for abortion, as seen in the Brazilian case: when the mother is otherwise healthy but is too young (9!) to safely carry twins to viability, much less through a complete pregnancy.

            You also seem to have forgotten that the Arizona case occurred in a Catholic hospital, and the abortion was only performed after approval by an ethics committee. Your assertion that doctors “don’t want to put forth the effort” is nothing more than a base libel.

        • Gilbert

          So your objection is that my example is symmetrical, in that the friend could also throw you out, right?

          The bomb brings in further complications which would lead us too far away, but I can remove that difference from my analogy:

          Your friend happens to be very fat. The remaining fuel would be enough to carry you to safety if he wasn’t there but not to carry him to safety even alone.

          Also the question of volunteering is a distraction, so I will specify you where the designated pilot and he’s too drunk to appreciate the situation or volunteer for anything.

          So can you throw him overboard? And again, if yes, are you sure only a Catholic could answer no to that question?

          That second question is actually much more interesting to me. Because the funny thing is on a very theoretical level I actually agree with you, because I think no moral principle including this one is consistent with the non-existence of God. But we are talking real people here and real people luckily aren’t that consistent.

          So talking about real people, are you sure non-Catholics could not judge it wrong to throw the fat guy out of the gondola?
          I haven’t tested this particular modification on any non-Catholic yet, but my intuition says many of them would agree you can’t. That’s because I’m much surer about this than I am about Catholicism. So if you actually can establish this you’ll be handling me an argument for Catholicism much more dispositive than those I have so far been relying on. But actually I would put up money against you being able to show that.

  • Hibernia86

    When it comes down to trusting authority in science versus religion, the big difference is that in science, you could theoretically redo all of the tests that were done in the past to see if they give the same answer. You can’t do that with faith. That is why science is more trustworthy.

    • Yes, the testing and self-correcting of the physical sciences can be easily trusted to produce knowledge that advances technology. There should be no controversy with teaching the contemporary concensus view in such disaplines.

      Even without repeatable experiments, the scientific method is producing more accurate explanations and predictions in cosmology, the social sciences, psychology, education, philosophy, etc. Again accepting the views of modern academia seems uncontroversial.

      There might be 21st century religions (theologian departments) producing trustworthy knowledge too. Or they might be in possession of trustworthy knowledge. The rest of modernity is all ears.

    • A 2000 year religious tradition is an experiment, or at least the observation of one. With 2000 years inductive strength. It is still alive, therefore it cannot be too far in error. Many other traditions have died; they were failed experiments. Sticking with the succeeding experiment and avoiding the failing ones is a good idea.

      • David

        A 2500 year religious tradition is an experiment, or at least the observation of one. With 2500 years inductive strength. It is still alive, therefor it cannot be too far in error. Many other traditions have died; they were failed experiments. Sticking with the succeeding experiment and avoiding the failing ones is a good idea.

        The religion I’m referring to is Buddhism, which makes numerous claims that are incompatible with those of Christianity. If a religion must be relatively close to the truth simply because it has survived for a long time, then how have Buddhism and Christianity both survived for so long? They can’t both be not “too far in error.”

        Look, I’ll agree with you that we shouldn’t choose to believe the “failing ones.” But I doubt you’ll find any atheist advocating a return to Greek paganism. In any case, the longest tradition of belief is almost certainly non-belief (I doubt that my Australopithecine ancestors had religion). So we win, right?

        • Missed responding to you.

          Unbelief is a lack, right? A disagreement with whatever is the standard belief of the day. Therefor unbelief always changes relative to whatever it is opposed to. It has no consistency, therefore it cannot be said to be continuous, or the same thing. Besides, if you have ever studied anthropology you know that nonbelief is a ridiculous idea in most cultures. Only in very fancy cultures can the idea of nonbelief be entertained at all, and even then, you are still believing a lot.

          You may have been joking, but I have a thing about atheists equivocating between nonbelief being a thing or a lack of a thing. It makes a big difference in how one can address it, and equivocating the point is an annoying rhetorical ploy. That’s all.

          • David

            To begin, I have no idea what you’re talking about when you say “nonbelief is a ridiculous idea in most cultures.” Some level of rejection of the dominant religious practice seems to appear in every society I’m aware of, and usually that includes people who are functionally non believers. That’s true in Europe dating back to at least the Ancient Greeks and in China dating back to at least Confucius (those being the areas I’m most familiar with). I guess you could call those “fancy cultures,” but by that definition, most of the world has lived in “fancy cultures” for all of recorded human history, so it’s not that meaningful a concept.

            As for your second point, yes I was joking (so I won’t respond in any more detail – I think your notion that the length of a religion surviving says anything about its worth is ridiculous, and was simply poking fun at it), but I actually do think it’s reasonable to consider atheism to be between a thing and a lack of a thing. It usually requires an active decision to reject a dominant set of cultural beliefs (making it a thing, though this is less true now, as more and more people are raised as atheists) but remains a lack of a thing in the sense that it does not necessitate any specific set of beliefs (other than not believing in God) and doesn’t have any institutions that are particularly devoted to it.

      • Hibernia86

        You’ve proven that Christianity can get a lot of followers, but you haven’t proven that it is true. The idea that the sun goes around the earth was common belief for thousands of years because that is what it seemed like to people, but it wasn’t true. Religion survives because it gives people hope and community. But that is not the same thing as being true.

        • Hibernia I already said exactly that below.

  • keddaw

    …it’s hard to figure out how you judge one moral system as more accurate than another.

    Which presupposes there is an objective moral code that our moral systems can be accurate about in the first place. I completely reject this and suggest moral error theory as an alternative.

  • Do Catholics subscribe to Divine Command Theory? If so, I could never see the Catholic system of morality as actual morality.

    • A small minority of Catholic theologians are DCT, but the mainstream is Thomistic (rationalist) natural law.

  • All in all, morality is tough because it’s got a lot of retrofitting of data and limited abilities to perform identical tests. Extrapolating from that, I guess I would prize moral systems which encouraged individuals to come forward with lots of data so that correlations can be continually updated. If I’m not careful, I’d fall off into the opposite of justice (proudly) because I think the educational effects of a crime are way more important than what happens to the criminal.

    So I’m looking for a system that has lots of forgiveness, lots of after action reviews, and lots of data analysis. Good arguments is a plus.

  • “It’s harder to figure out what kind of evidence could justifiably give Chesterton that kind of confidence in the church’s moral judgments.”

    Leah this is a great post and the above quote is really worth addressing. As I mentioned above, it is really about endurance. Pick a tradition that succeeds. Catholicism has been enduring the natural and human environment for 2000 years and has not only survived but grown quite well. It must be doing something right. That deserves respect and a modicum of confidence for future, although as in finance “past performance is no indication of future returns.” This is a problem of a cultural tradition engaging a changing environment. Will it withstand selective pressure? Who knows. All we know it s that it has successfully resisted all comers so far.

    All that proves nothing of its theory (theology), only its practice (ethics). The practices are in accord with what succeeds in nature. But insofar as the theory preceded the practice then there is the indication that there is intelligence in the practice-selection process, whether human or divine. Insofar as practice comes first and theory post hocs a rationalization, then it is just natural selection with a human cultural varnish.

    • David

      Why does truth have any relation to success? Do you give the same credence to Islam, Buddhism, or Hinduism, all of which are very old religions with huge numbers of followers?

      Also, your claim that the theology of Catholicism comes first and the practice follows from it is highly suspect. Catholic theology isn’t some unchanging doctrine that was put into place 2,000 years ago – it has undergone substantial revision due to arguments made by theologians, decisions made by popes and church councils and the like. Belief and practice are in dialogue and influence each other. Why did the church go from Dum Diversas, which explicitly allowed enslaving “Saracens and pagans” to its modern position that condemns slavery? The most reasonable explanation is that involved conforming to practice unconnected to fundamental doctrine, and that as that practice changed, church doctrine changed. So perhaps the Catholic Church is an adaptable social institution that has survived due to its adaptability, rather than its adherence to an unchanging truth about which only it is correct.

      • Well said, David. One could also add that the Catholic church used to be explicitly and totally opposed to religious freedom (as Pope Gregory XVI said in 1832 in “Mirari Vos”: “that absurd and erroneous proposition which claims that liberty of conscience must be maintained for everyone”), whereas they’ve changed their tune in recent decades, as with Dignitatis Humanae.

        What we’re really seeing isn’t a single moral standard that’s been unchanged through time, but a large social institution adapting and changing its ways in response to cultural trends and societal pressure. The same is true for pretty much every major denomination.

        • Andrew G.

          Not to mention the rather unedifying history of the medieval and early popes.

      • “Do you give the same credence to Islam, Buddhism, or Hinduism, all of which are very old religions with huge numbers of followers?”

        Yes, in matters of practical truth, I do. The practices of those religions are sufficiently in accord with natural law, therefore they endure and grow. I disagree with their theory for why their practices work.

        The matter of ethics is long settled (excepting social and environmental change, which cannot be excepted, I know), human groups with “bad” ethics do not endure. Theory is another matter entirely – it only affect endurance and growth insofar as it affects practice.

        Your point on slavery is not an overturning of doctrine, it was removing a tension from the tradition. Slavery was never a doctrine, it was just a fact of life from time immemorial, that people eventually realized was not in accord with Christian ideas of human dignity. Human dignity and slavery do not fit well together, and since human dignity is the more important of the two ideas slavery had to go.

        Same with religious freedom, in response to Adam. Human dignity is a trump card. With human dignity, the idea is unchanging, the ramifications are not. As time passes, the idea is more and more understood and practice must change to fit theory, as theory is worked out.

        Lastly, the adaptability of the Church is built in to the Church’s doctrine. Rome settles disputes. Not all religions have central authorities, and so tend to split or not adapt.

  • Joe

    Brian can you give me some reading material on your understanding of the Olmsted problem and the principle of double effect? I don’t think you’re wrong I just want to understand were you are coming from. Thanks

    • Hi Joe,
      First I would say that because of patient confidentiality the specifics of the case are not completely clear. My understanding of the case if that an 11 week old fetus was delivered prematurely, but not dismembered. I do not know the means of inducing delivery, which is relevant here because delivery of an unviable fetus is specifically ruled out according the _The Ethical and Religious Directives for Catholic Health Services_.
      The ERD 45-49 gives the specifics, it is avaliable here:
      And double effect is the context for the ERD:
      I have not seen a good description of the case online that answers the necessary questions of direct vs. indirect. I only have information from some of my medical ethics professors (who are on ethics committees at Catholic hospitals and need to know about these things) who have said this was an incorrect application of the principles. I acknowledge that they may be wrong, in which case I am wrong as well, and Olmsted is right. The devil is in the details… and since they are not publicly known I possibly overstated my point. So there is a much more measured answer.

  • J. Quinton

    Catholicism has been around for 2,000 years? Ok, I raise you Judaism, Zoroastrianism, Hinduism, and Buddhism. If your metric for reliability is longevity, then Catholicism comes in second to last among the major world religions. It would mean that all of those other religious traditions are doing something better than Catholicism to have lasted longer.

    • Not just longevity but also growth. Zoroastrianism is dying because conversion in is not allowed. Judaism is not into proselytization. One cannot convert to Hinduism either, one can only become a Hare Krishna (a sorry approximation!). Buddhism, sure, no problem with that. The real challenger is Islam, because despite being younger, it has grown very fast. I dod not know if it has grown faster than Christianity, and figuring it out would require a lot of historical analysis. Maybe someone, somewhere has done that.

      My point is, if you are looking for a “teacher,” then look for one with a good record. In this case the teacher is a tradition. Choose one with a long record of not going extinct (unless you want that). No guarantees of future success, but a good track record ought to be a first starting point. No guarantees about it being theoretically true either, only practically wise. Theory and practice are not the same.

  • Patrick

    Catholics usually get really angry when they think I’ve implied that Catholicism today is exactly the same as historical Catholicism. They give me lectures on how much things have changed.

    • Some things change and some things stay the same. Depends what you are looking at. Also the context and explanation. “Change” as in natural development or growth of doctrine is not considered to be bad in Catholicism. It’s just a working out of what is already there (like a growing organism), not a destruction and replacement (like a transhumanist replacement of organic with electronic!). Hey, I even got transhumanism in there.

  • Interesting post. Although differing to some authority might be a sign that someone’s not thinking for himself, it’s something everyone does numerous times in various areas of life. I think it can be a sign of humility and having a realistic view of one’s own understanding and capabilities. We have to learn to say we don’t have all the answers, even when it comes to religion and politics. Those who think they do may believe they’re free from cult-think, but in reality, they’re probably more likely to fall prey to some silly beliefs.

  • Joe

    Learning about the Dunning-Kruger effect makes me feel neurotic. Am I the only one?

    • 🙂

      Good observation! Seek the virtuous mean. Not neurotic, but neither un-self-reflective.

  • I don’t think that attacking this on the personal level only won’t get you much everywhere. “Being right where you were wrong,” does not seem like the reason for climbing the hill but a reflection looking back after the journey has already been made.

    Look at perhaps the best example: Who stood against eugenics for the best reasons? The Smart Set was all for it until Germany came along and actually did it. Yet the reason I hear from materialists — quite a different thing from atheists, indeed far worse — is that eugenics is wrong mostly for some blather about diversity in the gene pool. This is not a proper denunciation of eugenics, merely a dismissal as being impractical. Such reasons are not insoluble given time: Suppose through genetic engineering we invent a gene randomizer which as safely creates the same effect. What reason now do we have to oppose eugenics? Is eugenics still wrong? Who will stand against eugenics when it comes back for the second pass? Only we who have better reasons.

    I don’t say you approve of eugenics. I say, merely: To properly denounce such a thing as bad as eugenics we must have very strong reasons transcendent to the material universe, which is always at our fingertips and may be manipulated. The Catholic sort of normative denunciation, and in this case the first time around, is a powerful witness for being a truth-telling thing, I think.

    Simply put: We mustn’t give moral horror — including, I’d posit, abortion — an acquittal; the Church issued conviction.

    • Andrew G.

      So why did the Catholic church engage in its own analogue of eugenics? While opposing sterilization and genocide-by-extermination, it has widely participated in baby-stealing and forced adoption from parents who it considered morally substandard (such as unwed mothers), and cultural genocide via coercive institutionalized education.

      In the case of the Canadian residential schools, for example, the Catholic church was the last organization involved to acknowledge its culpability and apologize.

  • Joe

    In the case of the nine year old girl you simply perform an emergency hesterrectomy. Allowing the twins to die a natural death they would have suffered any way and at the same time save the girl. Always remember you can not kill an innocent directly.

    • Andrew G.

      I sincerely hope that response was merely medical ignorance on your part.

      Hysterectomy is removal of the uterus (rendering the patient sterile). I sincerely hope it is not necessary for me to elaborate on all the details of why, exactly, it would be unethical to perform such a procedure on an underage patient in the circumstances described.

      Perhaps you were confusing it with a hysterotomy, which is a term applied to any procedure in which the uterus wall is opened surgically. Hysterotomy is used as abortion procedure in relatively rare cases either when other methods have failed or the condition of the placenta makes them impossible or unwise; it is one of the most dangerous (to the mother) of all abortion procedures.

      Reportedly the girl was around 4 months pregnant at the time, so we’re talking about an early second-trimester abortion, for which the preferred procedure would normally be dilation and evacuation. An intact removal of the fetuses via hysterotomy would not be possible until somewhat later in the pregnancy, as I understand it, and the mortality rate for hysterotomy is more than 10 times higher than for D&E.

  • Joe

    I certainly ment hysterectomy. While fertility is held in high regard by Catholics it is not a god we worship. When faced with rendering someone infertile or directly murdering two innocent children fertility should be sacrificed. The operation would be risky for sure but don’t you think trying to save one of the three children is worth it? If the poor girl was suffering from uterine cancer or a sever infection would you hesitate to perform the procedure? I hate to correlate the twins to cancer or an infection but unfortunately the situation calls for it. There is simply no way around it you can not directly kill kill innocent children there is just no other moral option. Of course to the athiest murder is simply a matter of preference. “I don’t want to go to jail so I won’t murder someone inconvenient to me.” To you guys human beings have no intrinsic value

    • Andrew G.

      But it’s not a choice between killing two people or rendering one person infertile. The choice is between killing two people or killing two people AND rendering one person infertile.

      The idea that a hysterectomy performed for no medical reason other than to terminate a pregnancy is somehow morally superior to any other method of abortion is nothing short of insane.

      Are you actually capable of making a comment without including a slur on some group of people? First it was doctors, now atheists – doesn’t your Catholic morality have anything to say on the subject of lying?

  • Joe

    Sorry about the poor spelling I’m using my iPhone

  • Joe

    You are right I apologize for the unnecessary cheap shots. One of the things I enjoy about this blog is the civil tone of the discussions. I’m sorry it was me that went too far and made unfair generalizations.

  • Joe

    I think distinction in the case of the hysterectomy is that you are letting the twins die a natural death. In an abortion you are directly killing them then removing them part by part. It simply is not ethical in a catholic moral universe to directly kill innocent children no matter what the situation. Even if it means rendering the girl infertile.

    • Andrew G.

      There’s nothing “natural” about it. The sole purpose of the operation is to remove the twins to an environment in which their death is certain.

      If I kidnap you and lock you up in my basement with no food and water until you die, did I kill you?

      The sick part is the way that Catholic “ethics” promote the idea that any amount of risk, pain, and mutilation of the mother is justified in order to have the fetus die one way rather than another. This abominable attitude was, fortunately, rejected by the doctors in the case (and the judge); proper medical ethics requires that once you have determined that you can only save one person, you must act in the best interests of that person.

      • The rationale from the Catholic perspective is different. The choice must be made with the best interest of the whole community in mind, not just the person. If, as I said above, “the ends justify the means” logic is allowed here, that introduces an irrationality into the entire ethical system, leaving it vulnerable to unraveling into innumerable slippery slopes. Which would be bad for everyone, including the patient. The exception to the rule cannot be allowed. It must be a double effect, not an ends justify the means.

  • Joe

    Im sorry C-section would probably be the best and safest option after she was a little further along. Apparently the girls life was not in danger at the time of the abortions. and apparently extremely young mothers are not as uncommon as they should be.

    I apologize for wasting your time with all this. I should have looked into this before shooting off my mouth. The Dunning-Kruger effect!!! Sorry

    • Andrew G.

      See my response above discussing hysterotomy for the issues with that.

      As for young people giving birth, you’ll find that people have also carried ectopic pregnancies to term, survived rabies without vaccination, fallen 33000 feet and survived, and various other exceptional events. That doesn’t make it OK to impose those kinds of risks on another person for any reason, not even to prevent a death.