Originally I was just going to write the data crunching part of this series.

But as I worked on the first part, I found that it was increasingly impossible to prevent myself from noticing a few things about what I was reading.

So this post is where I take off the programmer hat, put on the philosopher hat, and explore my puzzlement with Leah’s conversion, while ocassionally referring back to the results of the prior post.


In preparing to write the first post, I looked through all of the posts Leah made between her conversion announcement and her baptism. I also read a fair number of other posts over the course of trying to make the machine learning algorithm converge into some kind of meaningful clusters.

Over the course of this reading, I came to the conclusion that Leah has two distinct arguments for Catholicism. She never explicitly separates them, unless I’m mistaken. But whenever I try to work out her arguments in my head, they always fall into one of two boxes.

The first argument that Leah makes starts with how she has many particular moral commitments. Moreover, over time she has revised and changed her moral commitments. But she noticed that many of her particular moral commitments were the same as those in Catholicism—and that, as she revised and changed them, they became more like the moral commitments of Catholicism. Thus, just as finding cases where the Copernican model correctly makes predictions that the geocentric model does not is evidence for Copernicanism, so also finding cases where the Catholic moral model correctly makes predictions that an atheistic moral model does not is evidence for Catholicism.

Leah gives an analogy for this—it’s like deciding that a map is right, after you find repeatedly that your exploration of the territory agrees with the map. She also affirms that she did not convert because of any observable empirical predictions that Catholicism makes, but because of the kind of reasoning that I’ve just outlined:

I did not convert because of empirically testable claims about Catholicism. There may be miraculous cures that ought to convince atheists, but I haven’t encountered them. Starting from an atheist prior, a spontaneous remission or other ideopathic healing might be surprising, but not surprising enough to outweigh the prior estimate of the improbability of God. But there are other kinds of claims to examine.

Some are the more abstract philosophical claims for the necessity of God, which are interesting, but not accessible or urgent feeling unless you have a strong scholastic bent. But religions also make moral claims, and that’s where, a la Chesterton, we can try and see whether a theology is a truth-telling thing. Is it self-consistent? Does it cover the things you already know to be true? Where it makes different moral claims, do they turn out, on further inspection, to be right after all?

This kind of investigation is more akin to how a chess novice would recognize Kasparov as proficient. The novice doesn’t have the skill to discern whether each move is clever or foolish, but she can see that Kasparov keeps winning. Correct moral judgement is less obvious than checkmate, but the general idea is the same. It’s the problem of finding a teacher when you know that you’re deficient in the subject of instruction. And these moral claims do require more than lip service.

As the map has revealed itself to be a truth-telling kind of thing, and as Kasparov has revealed himself to be a truth-telling thing about what chess moves to make, so also Catholicism has revealed itself to be a truth-telling thing.

So much for one argument.

The second kind of argument Leah alludes to more shyly. She hasn’t spelled it out with the kind of detail that she’s spelled out the first argument, as far as I know. So my reconstruction of it is tentative. But the basic idea behind this argument is that God is necessary, either for man’s knowledge of moral truths, or for moral truths to exist at all.

So for instance, in Leah’s account of her conversion, she finds that she already believes that “the Moral Law wasn’t just a Platonic truth, abstract and distant” but “some kind of Person, as well as Truth.” In the argument leading up to this realization she is discussing both “how humans bootstrap up to get even a partial understanding of objective moral law”—that is, man’s knowledge of morality—and where “moral law came from in [her] metaphysics”—that is, the existence of morality. God is apparently necessary for one or for both of these. This is a notably different kind of argument from the above.

I want to summarize this argument more clearly, but I can’t. I don’t have a good enough grasp of the moving parts and internal logic of the argument, and I don’t think Leah has given them. I could fill it in with what I would surmise MacIntyre would say; or I could fill it in with what someone like Sartre or Dostoevsky would say; but all this would be guessing. It probably has something to do with an Aristotelian nature and teleology.

So instead, let me turn to problems with both of these arguments.


Actually, forget that. I don’t understand the second argument well enough to poke holes in it—I just don’t get the whole thing, although I’d like to. So I’ll just explain some things that doesn’t make sense about the first argument.

There are two things that puzzle me about this first argument. The problem is that if you want to argue that X predicted the truth about A, B, and C, and therefore X is likely also telling the truth about P, Q, R, and so on, you must be sure of two other things.

First, you need to be sure that it really is X telling the truth about A, B, and C, and not some small and accidental subsection of X which says A, B, and C, while other subsections say ~A, ~B, and ~C.

Second, you need to be sure that X has actually always predicted the truth about A,B, and C, and hasn’t just switched to saying those things because everyone else found that A, B, and C were true.

The former problem bothers me the most, although it isn’t the most salient for everyone. Let me explain it at a bit more length.

If Leah wishes to use this argument to support belief in the Church, then she needs to show that the Church as a whole is a truth-telling thing, not merely the parts within the Church that most appeal to Leah.

Now, there are parts within the historical Church that sound like the kind of virtue-ethicist Leah is.

But there are also all sorts of ethicists who write like heaven and not being an alter Christi is man’s telos. Catholicism itself has an extremely ambiguous historical relation with slavery. And if we leave the field of morality, Catholicism also makes some really dubious empirical claims. I’d be happy to discuss any of these issues at greater length—but the point is that Catholicism is enormous, and if finding that parts of Catholicism say things I agree with is sufficient reason to convert to Catholicism, then everyone, including people who disagree with each other about almost everything, should become Catholic.

In short, if the truth-telling argument is to have any robustness, it must show that the unexpected predictions Leah speaks about were really being made by truly core aspects of the institution to which she credits them—that is, by aspects of Catholicism more central to it than the parts of Catholicism, alluded to above, which have ruled obviously and clearly falsely. It would be unreasonable for me to convert to Islam, merely because Averroes was a virtue ethicist and a major Muslim thinker, and because I like what Rumi has to say about human natue. Or, to switch to the map analogy—it would be unreasonable to trust a box full of often contradictory maps, merely because I think that the maps which I prefer within it are right.

The second problem is straightforward, and Leah already seems to be aware of it. Yvain, commenting on her blog, remarks.

Make sure that a theory which claims to predict the data hasn’t been massaged into the correct predictions after the data were discovered. This seems to me a notorious problem with moral theories.

Or, in short, we need to be sure that Catholicism’s current support for things like religious freedom isn’t a retcon of past decisions.

To Yvain’s comment, Leah responds “this is my biggest concern.” This sounds like a prelude to an examination of the issue; but as far as I know she never brings it up again. If there’s a series on her blog where she goes into history to make sure she’s avoiding such post-hoc fitting, it certainly didn’t happen between her announcement of conversion and baptism—and I haven’t found it anywhere else either.

I don’t see a way for her argument to have anything but very little to no force without some kind of doctrinal or historical study showing that the above concerns do not apply. So far as I can tell Leah has never even gestured towards such a study.

(Another weird thing about this is that Leah’s intutions about LGBT issues diverged from the Church’s before conversion—but after converson, as the charts in the prior post indicate, she seems to have pulled back from discussing them. The pre-conversion Leah often seemed to look at atheism, seek the weakest points in it, and then relentlessly drill into those points. I don’t know of times or ways that the post-conversion Leah does the same thing for Catholicism—the data about LGBT issues seems to indicate that she may have done the opposite.)


That’s all I have to say. I haven’t presented proof of anything—as I said, I hope to start an argument, not to end one.

(Epistemological status: This is perhaps a quarter as rigorous as I wish it was. Corrections welcome and encouraged.)


Leah Libesco is a convert from atheism to Catholicism. I’m a convert from Catholicism to atheism. We both are interested in MacIntyre’s virtue ethics, empiricism, and statistics. So I thought it would be interesting for me to examine how her blog has changed since her conversion.

In this post, I apply some basic algorithms to scraped blog data to examine how Leah’s blog has changed since she has converted. In a follow-up post, I’ll talk a little bit about what I think of the arguments Leah gave as reasons for her conversion.

In neither post do I think I’ve come to a really definitive result about anything. The only thing I’m willing to say about the first section is that I think there’s a decent chance that it presents the kind of data that should promote a particular hypothesis to conscious attention—it by no means presents data proving any hypothesis. If you want the end of all debate, you’ll have to look elsewhere. I do hope that this can at least start a debate, however.


All of the code I used to scrape / parse / cluster the data, as well as the raw data itself, can be found in a github repo. The cluster index numbers I use below relate to this uploaded data, if you want to look at it yourself.

Methodological paragraph follows. Note that the following describes the final way the data was analyzed, and not the process by which I converged on this final way—there was of course an exploratory / tinkering period where I was just trying to get meaningful data from the scrape.

I scraped all of the data from Unequally Yoked on October 25th. I converted the text content of each post (excluding title, tags, and comments, and attempting to exclude the ideological Turing Test posts which largely consist of content from other people) into a bag of words, after stemming each word. I then got rid of every word that occured in more than 30% of the articles (as likely too generic to help cluster the data) or less than 0.75% of the articles (as likely too rare to help cluster the data.) Using each of the remaining words as a dimension, I then used the article-length-normalized frequency counts for each of the words in each post to locate each post as a point in the resulting vector space. I then performed k-means clustering on the points which remained, initializing with 30 clusters, and iterating through 2000 different initializations of Loyd’s algorithm so that the resulting clusters would be as compact and repeatable as they could reasonably be expected to be. End of methodological paragraph.

Clustering data, in theory, allows you to pick out groups of related things. So a good test of whether the above has had any success is to look at whether it has picked out as grouped any text that we’d naturally think of as grouped.

So for instance, one cluster (#26) includes about 50 different posts discussing the Ideological Turing Test—as I said, I tried to remove the entries themselves, but this still leaves the peneumbral posts. When we look at the centroid which defines the cluster, it is centered around posts with words like “answer,” “test,” “Turing,” and “ideological.” So the algorithm picked out at least one cluster which seems fairly natural.

That a cluster appears can tell us two things. First of all, clusters express correlations between words. Assuming that using particular words generally reflects concern for the ideas expressed by these words, words which are are common to a particular cluster probably express ideas which are related or which are frequently discussed together in the blog.

Second, by looking at how instances of posts belonging to particular clusters increase or decrease in frequency with time, we can gain insight to how the blog’s concern with the topics denominated by those clusters increases or decreases with time.

Excluding degenerate “clusters” of two or less posts (numbers 1, 2, 4, 9, 10, 11, 13, 15, 16, 17, 19, 20, 21, 22, 23, 24, 25, 27, 29) and an also excluding an apparently degenerate “everything I couldn’t classify” cluster (number 12) which has fully half the posts thrown into it, that leaves ten reasonably classified clusters.

I’m quite worried about that “junk drawer” cluster, but I’m not sure how to elliminate it—I have tweaked the above procedure a fair bit, trying to find clusters without a junk cluster, but so far haven’t been able to do so in a way which seems to preserve meaning in the clusters which remain. And I’m not terribly sure that this “junk cluster” doesn’t reflect the data in some real way. Unequally Yoked does leap around a lot in the kind of things it discusses, and if I were to manually classify the posts it contained, I don’t know how much success I would have in classifying everything without finding a number of posts which seem to be edge cases without clear classificaition. Nearly all the quick takes fall into this category, which seems correct. So I’m not sure how much it should bother me.

Anyhow, here are some names that I would give to the clusters that remain.

  • #0: The Moral Law, God, and Rationality

  • #3: Marriage and Friendship

  • #5: Gay Marriage, Civil and Otherwise

  • #6: Sex and How We Talk About Sex

  • #7: Catholic Stuff and Prayer

  • #8: Atheism, Evidence, and Reasoning

  • #14: Women and Religion and Friendship (?)

  • #18: Announcing Winners of Ideological Turing Tests

  • #26: Discussion About Ideological Turing Tests

  • #28: Humanist Morality

Let’s go through a few of these, starting with those which don’t seem to trend upwards or downwards in frequency as time goes on.

Numbers #18 and #26 are probably the least interesting; they’re clearly about the Ideological Turning Test; their frequency remain reasonably constant over time. I don’t think there’s much to learn by looking into them, given that they correspond very clearly to what we already know.

I didn’t anticipate category #14, which makes it very interesting; my name for it is tentative. Keywords that documents in this category are likely to have include “she,” “share,” “women,” “relationship,” and “religion.” It includes posts discussing Ron and Hermione’s relationship, whether Lydia Bennett is to be admired, and attractiveness and nudity in art and narrative.

This category also includes discussion of Pope Francis’ discussion of Saint Judith and discussion of Beguine Convents. I guess I’d tentatively identify this category as looking at ethical issues—in the sense of “how can I live a good life,” not in the sense of “what do I do with a trolley”—from a female perspective.

In any event, this category remains fairly constant over time, with a probably-insignificant slight increase in frequency.

So that covers the largest clusters which don’t have any particular trend over time. What about clusters which shift as time passes?

One thing I’d like to remark on before I go over these is that, of the clusters which do shift, all seem to shift more or less when Leah converts. This is what you would expect if you think that (1) these categories are significant, and reflect natural ways of cutting up topics in the world, and if (2) you think conversion is a serious thing. So that’s a little more evidence that these clusters reflect some way that reality is really cut up.

Number #7, “Catholic Stuff and Prayer” has posts like the pre-conversion Living on a Prayer and Adoration and Abramovic, as well as the post-conversion So Small a Cloud of Witnesses. The cluster is centered around posts with many words like “Catholic,” “prayer,” “church,” “heart,” and so on and so forth.

This is the frequency of ocurrence of posts in this category over time.

This looks pretty much like you would expect. After Leah converts, posts about prayer, Catholicism, one’s heart, and things like that, tend to increase in frequency. I don’t see anything else particularly interesting in this category, so let’s move on to some other categories.

Categories #0 and #8—“The Moral Law, God and Rationality” and “Atheism, Evidence, and Reasoning” are interesting, and seem like they should be approached together. Why together? Well, for one thing, both of these are centered around some of the same keywords. Both of these are centered around the kind of articles that use the word “evidence” more than any of the other non-degenerate clusters; they are also apt to include words like “atheist,” “exist,” “believe,” “philosophy,” and “religion.” Number #8 is more apt to use the word “fight” than any other non-degenerate category. So, generally speaking, both are related to trying to come to the truth about atheism versus Christianity.

As the name I picked indicates, 0# differs from #8 in that it is more about the moral law and its relation to atheism. It includes posts from the series discussing morality in terms of math, as well as discussion of when God orders genocide. There are discussions of the four causes and map-territory problems. This category also notably (and reasonably) includes the post where Leah announces her conversion, as well as attempts to come to a taxonomy of religion.

Cluster #8 is similar, but seems to focus more on direct discussion of atheism versus Christianity, and on methodological questions of how to conduct such discussions. A number of Ideological Turing Test posts also fall into this category. This category holds posts of how best to talk to people you plan to convert. It also includes suggestions on how to profitably learn from people you disagree with. So I think it would be fair to summarize it as about reasoning about theism and atheism in particular, but also about how to reason well while disagreeing with people in general.

Here are graphs describing the frequency of posts in these categories.

I’d like to note that these categories and those like it are moderately robust; clusters like these were among the first to emerge while I was fiddling with the data.

So, if it is fair to say that these posts are chiefly about sussing out the truth of atheism versus Catholicism, it seems as if this is (some kind of) evidence that Leah’s concern with sussing out the truth of atheism versus Catholicism decreased fairly rapidly after her conversion.

Of course, you could maintain that perhaps Leah’s concern with sussing out the truth of Catholicism versus atheism didn’t decline at all—it’s just that she expressed it with such different terms, such that they don’t show up in these groups. Perhaps all these posts fall into the unfortunate and massive miscellaneous cluster, mentioned above. This is very mildly improbable—this isn’t want you’d anticipate, going in to this clustering problem—but it is a very good point. This is part of the reason I’m really reluctant to draw conclusions from these; and it’s also why I’m going to talk about that possibility a little bit later.

Let’s switch to topics which have to do with sex.

Cluster #5 is about gay marriage and covenant marriage. It includes posts about the Prop 8 play, discussion of whether civil marriage is a right, and towards the end, a declaration that Leah is still in favor of gay marriage.

Cluster 6# also discusses gay marriage ocassionally, although more in the context of friendship. It starts off with a postdeclaring that natural moral theory about gay sex is bunk and moves on to discussing gay marriage and male friendship. It also discusses terrible ways to have sex, which caused a bit of discussion in the comments, and avoiding rape-adjacent sex. This cluster also decreases in frequency after the conversion; but it too has very few members.

Here’s a chart showing the frequency of posts falling within either of these two clusters.

These two clusters also seem to decrease in frequency after her conversion. So the — again very tentative — conclusion is that Leah’s expressed interest in LGBT issues decreased after her conversion.

So there are obvious clusters of posts about atheism, Catholicism, morality, and God, where Leah “picks fights in good faith.” Although the computer happened to pick them out, in my opinion these clusters are sufficiently obvious that if you want to spend a few hours browsing the Unequally Yoked archives you’ll see them as a natural group as well. Both of these clusters drop precipitously in frequency after Leah’s conversion.

One danger when doing this kind of clustering is motivated stopping. Over the course of surfing through the data, I’ve seen other ways of clustering—so for instance, the Pope Francis bookclub was also a fairly common cluster for data, although it didn’t really emerge in the above. When I acted in a way that tended to produce smaller clusters, I managed to make some emerge with Christmas-themed posts or posts about Les Miserables. In any event, if I happened to stop refining the clusters when I found a set of clusters that fit with a narrative I had in mind, then I could tilt the clusters to support that narrative; I could do this easily, of course, even if I was trying to be objective. And, of course, anyone who reads Leah’s blog on a regular basis is surely aware of this possibility.

You’re welcome to search through other clusters—as I said, some of the above tends to appear through a fair number of different ways of grouping. But, leaving aside hypothetical alternate clusters, what would be another way of looking at this data? What would be another way to see if the kind of trends suggested by the above clusters are real?


One way—also fallible, of course—would be to perform a month-to-month wordcount of words connected with concern for rationality and evidence and disputation—things like “prediction,” “evidence,” “debate,” “argument,” and so on. We could also perform a month-to-month wordcount of words connected with concern for LGBT issues, to test for the same.

(Note that, of course, using words like “predict,” and “evidence,” and “logic” in no way guarantees that you’re actually reasoning well. Obviously. What I do take using such words to indicate is at least concern for these matters. Scholastic philosophers, in my current opinion, reason poorly—but they use the word “reason” all the time, and this at least means that they are concerned with reasoning. Similarly, a scientist could use the word “predict” while HARKing all over the place. But it at least shows that he’s attempting to adhere to a particular standard, and probably has the standard present in his mind.)

I came up with a list of reason-connected words only once—I haven’t altered the list to try to produce a smooth curve that matches any narrative. (Although, of course, I could have come up with a list of words that I suspected would match a narrative.) The list I came up with (visible in the github repo) gives a count of uses of reason-connected words as follows.

The break isn’t quite as abrupt in the examples above, but it’s hard to look at the picture and refrain from seeing a quite definite effect following conversion. Here’s a chart detailing words signalling concern with LGBT issues.

Again, this appears to tell us something about the revealed preferences of Leah since her conversion.


So the above indicates that there’s probably been some kind of change in content since Leah’s conversion, and also tells us a little bit about that change.

What if we could try to use the data to ask a more Leah-like question, though? What if we could ask how Leah’s conversion has changed Leah’s view of herself, and of the actions she habitually does?

One way to attempt this would be to look at the verbs following “I” in Leah’s writings—one could examine whether she habitually says things like “I wonder” or “I doubt” or “I think” and so on and so forth. A methodologically rigorous way of doing this would involve parsing out helping words—“I have wondered” and “I would doubt”—but I don’t have the time for that.

So instead I looked at the words which followed “I” in the data I scraped. I also dropped all the words I judged boring—words like “was” or “wasn’t” and so on. All the cases before the month of her conversion announcement I put in the atheist box, and all the cases afterwards I put in the theist box.

Here are the ten most frequent interesting words to follow “I” when Leah was an atheist.

Word Frequencies
think 5.41%
know 1.76%
thought 1.13%
found 0.96%
read 0.90%
believe 0.86%
find 0.82%
wrote 0.67%
disagree 0.59%
need 0.57%

And here are the ten most frequent interesting words to follow “I” after Leah became a Catholic.

Word Frequencies
think 4.75%
found 1.27%
know 1.27%
read 1.18%
thought 1.12%
love 0.95%
feel 0.84%
saw 0.80%
find 0.72%
need 0.62%

This probably tells us more about the constants in Leah’s personality than about how she has changed: in each case, the top five words remain the same, although they do switch their order around a bit. Leah thinks about things, recounts what she has found, says she knows things, reads books, and so on and so forth.

There are a few potentially notable aspects. I was surprised by the prominence of “believe” in the atheist group, until I realized that in conversational English it is less commonly used in a credo and more often used in a concessive phrase: “I believe that this is so, although…”

“Love” and “feel” both increased in frequency after conversion, love moving from 0.46% to 0.95% and “feel” moving from 0.46% to 0.84%.

Similarly, “disagree” is over twice as common when Leah was an atheist as when she was a theist.

Frankly, though, I expect this data is sufficiently noisy that I’m unwilling to trust any of it. If there were some obvious pattern running through it, I would follow it—but I don’t think there is much of anything.


It would be interesting to train a supervised-learning algorithm to recognize the difference between pre-conversion and post-conversion posts. Looking at the parameters for the resulting model might be enlightening. But I don’t really have time to do this at the moment.

In summing up the data-crunching section of these two posts, I’d like to emphasize that I haven’t proven much of anything.

I’m pretty confident that we can say that the content of Leah’s blog shifted when she converted—but we already had really strong priors that would be the case. What kind of narrative you support regarding the shift will probably depend on whether you’re already atheist or Catholic. There are both atheist and Catholic narratives that fit it reasonably well.

I think that the above results should at least spur a bit of curiosity.

(This post is the first of a two-part series.)

(Epistemological status: Seems likely enough)


Evidence supports the broad outlines of modern physics pretty well. So a great deal of physics is rather certain.

Because of this certainty, we’re able to infer something about anyone who confidently claims to have produced a device or theory overthrowing the structure of all of modern physics. Specifically, we can infer that the claimmant is very likely either stupid or deliberately ignoring evidence — or perhaps both. If a reasonably rational person were actually seeking evidence relevant to their physical claims, after all, they would very probably soon stumble across some of the mounds of evidence contradicting their beliefs. That they strongly hold the beliefs they have, then, serves as evidence that they are either foolish or blind.

The same could be said just as strongly of people who say that they can square the circle in math, and so on and so forth. Those who undertake to argue with those who claim mathematical impossibilities soon find their interlocutors intractable.

Because of the mass of evidence in favor of physics, people who do find potentially groundbreaking results in physics always present their results as very tentatively, both because (1) they know that such results are very probably a result of error on their part, and (2) they don’t want to appear crazy. The EmDrive has experimental evidence for it from a NASA research group, a research group at the Dresden University of Technology, and a university in China. Most scientists[citation needed] still expect it to turn out to be some kind of error in measurement, though, just like the superluminal neutrinos of 2011. And that’s entirely reasonable of them, because the EmDrive would likely overturn big chunks of physics if it did turn out to be real.

So generally, when there is overwhelming evidence for a particular position, the people who confidently disagree with it do so because they have some kind of personal flaws. These flaws may not be things for which they can be blamed—bad education can do quite a lot—but they are nevertheless flaws.

Note that this inference has to happen only when you think the evidence for your position is both strong and public. If you’re less certain that you are right, or if you have access to information the public lacks, then the people who disagree with you can be correspondingly more reasonable. People who disagree over various interpretations of quantum mechanics, say, don’t have to think that the people who disagree with them are bad or stupid people. And that’s because this is an open question, with evidence on both sides.


People are generally very overconfident.

There are a lot of studies that confirm this. Kahneman and Lovallo say that typically when people say they are 99% sure, they are right about 80% of the time. At least in particular tasks, the odds can get far worse—people who give themselves literal million to one odds of being wrong can be wrong one in sixteen times.

So people are generally extremely overconfident—they tend to think that they have more evidence for what they believe than they in fact have. (Moreover, a fair number of the above studies used facts about fairly boring, not-politically-charged topics. People tend to go a little bit crazy about political or religous matters, or anything pertaining to tribal affiliation, so there’s every reason to think that overconfidence would be far, far worse in these fields.)

(We should greet this news about overconfidence, of course, as news about you and about me, and not as about everyone else. In another study, when people were told about cognitive biases, they generally decided “Yeah, I bet everyone else suffers from those biases. Nah, I don’t really, especially now that I’ve heard of them,” even though tests went on to show that the people in question suffered from the biases just as much as everyone else.)

Being overconfident feels from the inside like “having a lot of evidence.” There’s no infallible introspection-visible sign that shows you that you are being overconfident—even discovering that you’re often overconfident is a slowly-learned skill. So overconfident people, obviously, are those who do not identify themselves as such. They are just those who think that their position is obvious, and that there’s tons of evidence for it.


I’ve written a little bit before about how some sets of beliefs have canonical or nearly-canonical ways of dealing with people who don’t think that the beliefs are true. Namely, some sets of beliefs accuse those who disagree with these beliefs of necessarily having particular vices.

So, for instance, Christians often deal with people who don’t think Christianity is true by accusing them of blindness induced by the desire to sin. Atheists deal with non-Christians by saying that they have some kind of tenacious meme infecting their minds. This can get ugly and I don’t want to go into it deeply. For now I just want to say that this is common and everywhere.

What I do want to go into deeply is this: Whenever you think the evidence for your position is overwhelming, the logic of the zeroth part of this post will tend to force you to think that everyone who disagrees with you is bad or stupid. After all, if there is publically-accessible, strong evidence for your belief, than willful blindness or stupidity are the most probable explanation for people who disagree with it. So because people are systematically overconfident, and people veeerrry often think that the evidence for their position is much stronger than it actually is, people will also systematically and erroneously think that people who disagree with them are bad or stupid.

I want to emphasize that the implicit reasoning going on here is basically right. When you suspect something negative about someone’s character because they deny something for which there is actually overwhelming evidence, your suspicions are based on sound reasoning: that someone denies something for which there is overwhelming evidence gives you (probable) evidence about someone’s character, as in the case of physics or math. (And in physics or math, you can often find your suspicions about their character confirmed.) The problem comes when one reasons correctly from an incorrect premise—when one decides that there is a great deal of public evidence for one’s own position, when there actually isn’t.

So when I read the hundreds of people on Facebook or Reddit arguing that people who agree / disagree with the morality of abortion are peculiarly self-decieved, these people are in fact reasoning correctly from the premise that their position is obvious. The thing is that their premise might be wrong.

(I had some example headlines here, but I decided to kill them. It shouldn’t be hard to think of your own examples.)

In short, I think a lack of charity towards others will pretty much always accompany overconfidence in one’s own opinion, not merely because they are psychologically akin, but because this is how the logic tilts you. And so you won’t get rid of the former till you get rid of the latter.

The kind of general shouting and hatred that occurs in the public sphere won’t disappear until systematic overconfidence in one’s own opinions also disappears. Just as I will not be able to get rid of any of my own unreasonable anger and or disappointment with the irrationality of others, until I get rid of my own irrational overconfidence.


A few interesting things actually follow from this, which can be useful for critiquing your own ideas.

First point: If you hang out with math cranks or read their arguments, my understanding is that you generally don’t find your evaluation of their character shifting that much.

On the other hand, if you hang out with people who disagree with you on any number of other things, often you’ll find that they seem really reasonable and coherent. This is sometimes surprising. That you are surprised by this is evidence that the evidence for your own position is probably not as firm as you think it is. This doesn’t always mean that your own position is wrong; it just means there’s a good chance that you’re wrong in ascribing to your position evidence as good as the evidence for, say, physics broadly conceived. Which means that, correspondingly, you should lower your level of confidence in the belief, because generally your level of confidence should correspond to actual evidence that exists for it.

Second point: This is only glancingly related.

Everyone gets their news from sources which mostly agree with them, nowadays. So everyone tends to read news that points out horrible things that SJWs have done, if they dislike SJWs; or news that points out horrible things the police have done, if they dislike police; or horrible things that particular kinds of protestors have done, if they dislike some kind of protestor; and so on.

Every time a new item appears in this feed, the natural thing to think is “Wow, look at a horrible thing X did. I can’t believe people are in favor of X. How can they be so evil?” So the news—which we carefully select to support only our own views—becomes an oportunity to see how willfully evil other people are. This is problematic.

This is not about how to perform experiments without doing bad things—that is, this post is not about the Tuskegee syphilis experiment and the Milgram experiment and things like that.

Instead, this is about trying to perform experiments to determine which things are right or wrong. It was somewhat inspired by a friend who said that he didn’t see how repeated, methodical observation of the world according to particular rules could ever increase your knowledge of ethics. I found I disagreed with this pretty strongly.

What I’m going to say really only applies to virtue ethics.

If you’re a virtue ethicist, good actions are the ones that promote good habits; good habits are in turn the kind of thing that help you flourish in a community that seeks the truth and which help that community itself flourish. (That’s MacIntyre’s account of virtue ethics, which has the benefit of being (1) probably the canonical modern account of virtue ethics and (2) the one I’m most familiar with. You could probably work what follows into slightly different accounts as well, but I’m not going to try and deal with that.) There are other meanings of “good” which are quite useful—the utilitarian account, for instance—but they’re not what I’m talking about right now.

A lot of people say that they are virtue ethicists. Very few people—in fact, basically none that I’ve ever run into—have really talked about performing some kind of social experiment to try to determine if some action is good or bad. So to say that virtue ethics naturally leads to “ethical experiments” would seem to require some argument.

Here’s the basic idea.

There could easily be some habits which, if adapted by many people in a community, could help that community flourish. In some cases, though, it might also be impossible to determine if adopting these habits en-masse would actually help that community flourish without some community actually adopting them. And you can’t just put on habits instantly—you have to try to ingrain them so they work even when you aren’t terribly conscious of them. So knowing whether particular actions are right or wrong might very well depend on actually doing them, to acquire the associated habits.

Dominicans in a monastery; those people in an intentional village trying to live sustainably but also happily; those Christians who come together for weird gatherings every week; or those community-occupying polyamorists who live a sparsely connected graph of relationships: you can conceptualize all these people as if they’re trying to see whether a particular set of actions-cum-habits can constitute a commmunity. They are (arguably) trying out particular habits to see if you can form a flourishing community from them.

These are all higher-risk, higher-stakes ethical experiments, like CERN for ethicists. There could easily be lower-risk experiments: Does being open about disliking someone and therefore not wanting to spend time with them help humans flourish? Is it ever good to cuddle with people you are not romantically attached to? And so on and so forth.

I’m not saying, you’ll note, that you have to personally try out everything before you know if it would be good or bad. The habits associated with psychopathic murder-sprees will not help you flourish in a community or help that community flourish, and you can know that without trying them out. It might be that some of the above could be rejected without trying them out—in fact, it’s almost certainl the case.

But I suspect that you have to try at least some things, before you know if they’ll work. You can’t figure out everything through game theory and introspection. Everyone’s going to disagree about where these lines are drawn, but I think the lines would actually have to exist.

I see this as an advantage for virtue ethics. Some ethical systems don’t seem to imply anything about the world—“ought” doesn’t seem to imply any “is”, in addition to the reverse. It’s only because “ought” seems to imply an “is” for virtue ethics that this proposal makes any sense at all. A deontologist could never propose any communal experiment to test his ideas; his ideas don’t lead you to expect anything anything at all from the world.

I could see that you would disagree with this proposal if you think the space of all possible virtues-cum-accompanying-communities is pretty well known. (I doubt that this space has been thoroughly explored.) I’m not really sure this proposal is entirely accurate. But it seems to me at the moment that something like this should accompany belief in virtue ethics.

I just finished up The Magicians, by Lev Grossman. It features a bunch of students learning magic in an English-style school named Brakebills, who afterwards travel to an also-magical land named Fillory, which has talking animals and tree spirits and is to be ruled over by two queens and two kings.

The story is partly about the kind of stories that people tell themselves about themselves, and how these stories can aid or destroy one’s ability to be happy. But writing against the background of Hogwarts and Narnia, as Grossman has done, allows Grossman make points about storytelling and about human nature far more effectively than he could have if he had been writing in a vacuum.

The contrast makes everything stand out more boldly—you can ask yourself if the image of human nature offered by Lewis or by Grossman is more accurate, or how the two complement each other. If The Magicians had somehow impossibly been written, word for word, in a world without Hogwarts and Narnia, it wouldn’t be nearly as interesting as it is.

Let me look at a similar case where a narrative is enhanced by springing from another narrative.

In After Virtue, MacIntyre describes how the Greek playwright Sophocles took a traditional character from Greek mythology, Philoctetes, and used this well-worn character to make a point about incoherences in Greek standards of morality. That he was taking a traditional story would allow him to make his points more clearly than making a new story, because everyone would know the assumptions made in the original story. That is, because everyone knew exactly how all the characters in Philotectes would normally behave, Sophocles could easily show serious problems with this normal kind of behavior.

What I’m getting at is that derivative narrative is a very useful thing. In some ways, it has enormous advantages over original works. The Greeks made derivative fiction ad nauseam with their characters, after all—there were four Philoctetes made by different playwrights, and Philoctetes is a fairly minor character in the array of stories the Greeks received from their tradition. Hamlet was one entry in a long line of riffs on the same story. And Watchmen loosely bases itself on previous comic-book characters, and certainly bases itself on archetypical comic-book characters.

Given this, why does fanfiction have a bad reputation? Given that many of the myths and plays that scholars scrutinize now are, so far as content goes, fanfiction based off earlier works, why do people sniff at it now?

I’m going to guess that it has to do with intellectual property. If you could make a profit writing fiction featuring characters from other intellectual properties, writers would probably do this—and we would see an evolution of stories like the Greeks saw an evolution of stories. But you cannot do that and make money, generally speaking, because people own characters. So a large number of skilled writers decide not to write fanfiction, which leaves a great deal — although not all — of fanfiction being written by unskilled individuals who like to insert themselves as love interests or just write straight-up porn and who generally give fanfiction its bad name.

But I think fanfiction in itself has no literary problems, and I don’t see any reason that there could not be great works of literature made that were fanfiction. My favorite work is, of course, Harry Potter and the Methods of Rationality. But there are others which are good, and so far as I can tell there’s no reason to sniff at the genre in general.

Once upon a time, in a timeless pseudo-Renaissance kind of era, there were two small kingdoms. Let’s call them Florin and Guilder.

Florin and Guilder for the most part live in peace. Each is ruled by a moderately benign, excessively decorative, and surprisingly sane monarch. Each trades with the other, and each benefits from comparative advantage in a way that would make Ricardo and Smith smile: Florin has a great deal of timber and pasture on its rolling hills and often-painted hills, while Guilder has greater mineral resources in its mountains and desserts, and has much better fisheries. The two countries throw actual Renaissance fairs, complete with enormous quantities of doubtfully-sanitary food and dangerously-violent forms of entertainment. They speak languages closer together than, say, Italian and Spanish are today—their people can converse with each other with only a little difficulty, which smooths the differences between them.

Of course, there are some cultural differences. Florin, for instance, has a fairly strict and hierarchical religion, which spread from a single prophet about five hundred years ago. Most of the important events in life take place beneath the auspices of its hierarchy: One cannot get married, or become king, or become a knight, or get buried properly without one of them present. This is not to say that they are oppressive. On the contrary, their priests are carefully trained to help people with their personal difficulties in psychologically sophisticated ways; their monasteries support and train people without sustenance or trade, and their monasteries also provide homes for children abandoned by their parents.

Guilder, on the other hand, is dominated by an older and looser religion, completely without hierarchy, which perhaps hearkens back to the nomadic period the Guilderians have relatively recently emerged from. The relation of man to the gods is scarcely mediated at all—to speak to them, you merely need go to a high place, a massive rock, or any of innumerable hallowed grounds which have been handed down from time immemorial. The gods speak to each individual as an individual, and not through any other organization. This sort of religious individuality also is present in their attitude towards the other matters in life. Burial is properly handled merely by your own relatives, and not by anyone else. You contract marriage with someone by sleeping with them; there need not even be witnesses to any formal agreement. The king becomes king through acclamation of his lords, without any particular person anointing him.

Of course, the theologians of Florin and the wandering gurus of Guilder like to argue over which of these is better. The theologians, for instance, often say that it is important for one’s king to be formally recognized by the gods, from whom all authority comes. They point out that the Guilderian gurus are very often charletans and frauds, who have no training in helping people and who muck up people’s lives. They say that the attitude of the Guilderians, who lack a ceremony for marriage, weakens the bonds of family; a few portray the Guilderians as sexual deviants. On the other hand, the Guilderian gurus often criticize the people of Florin; the people from Florin, they say, forget the meanings of things by draping them all in ceremony and ritual and gold cloth, as if that were what makes them important. And they scorn the people in Florin for requiring large marriage ceremonies: in Florin, a father is held responsible only for the children of his wife, and can forget and ignore his natural-born sons as he chooses; in Guilder, a man is held responsible any child he fathers, and all children inherit from their father equally. Those from Florin, the gurus say, build and visit brothels as they please—but a man from Guilder requires more self-control.

These differences rarely come to the fore; the theologians and the gurus are pretty much the only ones constantly shouting in each other’s faces about it. After all, the only people who spend a significant amount of time in both Guilder and Florin are merchants, and they have every reason to be polite; and after all, they see far weirder things in far weirder places. So the people of one country don’t actually care, for the most part, how the the people of the other one lives.

So they continue in peace.

One day, a princess from Guilder visits Florin as part of her education. She meets the high prince of Florin, next in line to the throne after the king dies. No one is really sure what happened next, apart from the almost undeniable physical facts. Some people say that the princess seduced the prince to try to gain the throne of Florin for Guilder. Other people say that the prince took shameful advantage of the princess by promising eternal love. Other people say even more nefarious things about plots long in the hatching. In pretty much any story, either the prince or the princess is the villain taking advantage of the other.

In any event, after returning to Guilder, it soon becomes evident she is pregnant. She bears a son with an unmistakable likeness to the prince of Florin. The political situation in the capital of Florin quickly boils over; the prince goes on a hunting trip to escape the mess that he has made. His horse bolts while he is hunting, and he tumbles off a cliff and dies.

By the laws of Florin, the son of the princess is a now bastard, who might receive special treatment from the king of Florin if the king feels nice, but who is owed absolutely nothing by the king.

By the laws of Guilder, the son of the princess is the rightful heir to the throne of Florin.

This is problematic. It might mean war. No one wants it to mean war; the mostly-decorative kings are old and tired, and really would like to ignore all the young knights clamoring for blood and glory, each sure that they can win the war in a fortnight. But the kings need some pretext to ignore them; the people are restless.

So they decide to hold a debate, between one of the foremost theologians in the University of the capital of Florin, and the guru who is acting advisor to the king of Guilder. It is a public debate, to be held before an audience of… as many people as can be in an audience, in the age before microphones. In retrospect, this will appear to be a stupid idea; but it seemed like a good idea to each of the kings when they first heard it.

The stands are packed in the hastily-redecorated theater in Florin. The theologian from Florin stands draped in the richly decorated robes of his office, the guru from Guilder in a plain grey tunic. Each of them thinks the other is trying to make a ridiculous statement by dressing this way. A mediator stands between them, from the land of Amry Esu. He coughs, looks at the resolution, and frowns, but reads it any way.

“Resolved: The bastard of the dead prince is not heir to the throne of Florin.”

The guru from Guilder looks surprised for only a moment, before the corner of his mouth twitches up. Typical ham-handed, pejorative terms to expect from a Florinese theologian. He speaks, slightly sotto voce, to the mediator. “We cannot debate that. To say ‘bastard’ is the same as saying that he is not the rightful heir, which is what we’re here to argue about.”

“Well, what should we say?” says the theologian. The mediator, the guru, and the theologian are all standing close together now; the mediator has ushered them together, so that they are being watched but not heard by the audience.

“We should debate whether the dead prince’s orphan is heir to the throne,” says the guru.

“Absolutely not,” says the theologian. “Orphans are children abandoned by their rightful, legitimate parents; that’s what the word means. Bastards are those conceived outside of a legitimately sanctioned marriage. We should call things what they are, and what the child is is a bastard.”

“That’s what you think,” says the guru. “But other people disagree with you. The language of Guilder doesn’t even have the word ‘bastard’ in it—we don’t judge children arbitrarily by the actions of their father. We just call a child who has lost his father an orphan, and that is what this child is.”

“It would be an offense against truth for me to call the child an orphan,” returns the theologian. “It would prejudge the entire debate.”

“And it would be a lie, and an insult directed at our innocent princess and her son, to call the child a bastard,” says the guru. “Our people”—he gestures to the audience, which has people from Guilder as well as Florin in it—“would not hear of it.”

The theologian from Florin sighs, as if irritated but unsurprised by this irrationality from the guru. “How about this—you can add a little preface, telling the audience that by calling the child a ‘bastard’ you do not mean to concede the matter to me. I’ll give you an extra minute of time if you do that, so you’ll have more than enough time to explain, and even add some more rhetorical flourishes.”

The guru snorts. “An extra sixty seconds, for a word carrying that weight of meaning? Before this audience? So you can use the word the nobles of Florin hurl at each other when they are drunk, or when they want a duel? So you can use that word, that word smelling of the tavern and the whorehouses of this city? The child is an orphan—at least your language has a word for it, even if you don’t use it for everything that you should. You can have an extra thirty seconds to explain what you mean by the word ‘orphan.’ Surely your in-depth academic training equips you to make a little distinction like that. You do that all the time in your academic debating shows.”

“Absolutely not,” the theologian says, also glancing at the audience. “If this were an academic debate, we could redefine the word. But here, we must use words with the meanings they have. Every orphan is a legitimate child, so I will not use the word orphan, nor accept it in the text of the resolution.”

The mediator speaks. “Perhaps we can use some kind of neutral word. Say, the ‘child’ of the dead prince.”

The theologian looks at the mediator like he is insane, or has just ignored everything he has said.

No. ‘Child’ is a word that is only used for the results of a lawfully contracted marriage. This was not a lawfully contracted marriage. ‘Naturalchild’ is the word that is used to indicate a bastard. Sure, under some circumstances, people casually use ‘child’ as if it refers to either, but when one is speaking formally, among nobility, under circumstances like these, everyone would notice.”

“Ok,” says the guru, who actually rolls his eyes, “how about infant. Can we debate using the word ‘infant’? Or does your language literally not have any way of referring to young people without needlessly dragging in what you see as the sins of their parents?”

“Fine,” snaps the theologian.

“Good,” says the guru.

So everyone returns to their places, and the mediator reads out the resolution: “The infant of the dead prince is not the rightful heir to the throne of Florin.”

And the debate starts; the theologian has the floor first. In his opening statement he refers to the child as a bastard every chance he gets. Every time he does so, the guru shakes his head a little. Why is the theologian so perverse, he asks himself. He can tell that the Guilderian audience is becoming incensed: Every time he uses the word, it is as if he is asserting that the Guilderian marriages of many people in the audience are on the same level as fifteen-minute arrangements in a whorehouse; that the Guilderian princess acted like a whore or in any event like an easy woman; and as if the child could somehow be tarnished by the actions of the child’s parents. The debate should be about the nascent field of international law and how it resolves disagreements among the laws of nations, he thinks—but he also just keeps thinking about how damn irritating this proud, self-confident, book-learned Florinese puppy is.

So when he speaks, he gives his carefully thought-through speech on the historical antecedents of this case. But he also makes sure to refer to the infant as an orphan. He notices that the Guilderian audience squirming, and he likes this. It feels really good. He bets that they’re thinking of how it isn’t the case that their marriages are something real and proper, just because they had someone else say a few words around them; he bets that the men are feeling guilty for things that they’ve done, or children they’ve abandoned.

The debate ends, and each of them feels confident that they’ve laid down a tight, waterproof case for their respective nations.

Later, incensed by the debate, the two countries go to war. Thousands die. Neither country wins any battles definitively, and they come to an uneasy peace two years later—which, as far as wars go, is pretty short. The infant dies of the flu when he is only ten months old, which is part of the reason the war is able to come to such a quick end with such few casualties. Until each of their deaths, both the theologian and the guru are confident that they did the best they could in the debate, because they only ever referred to things by words that were true and accurate and reflected reality.

Since I wrote this, I’ve revised my views on the matter. I still think that the physical pain that nearly all theologians believe occurs in Hell makes God a moral monster. But I’m not sure I can support as strong a temporal distinction between two versions of Hell as the below makes out.


I’m going to explain two different versions of Hell, then remark on the tenability of holding that either of them are really part of Catholicism. One of them is, I think. The other isn’t.

(I’m interested in this because I think only one of these versions is really morally tenable—and a while back people criticized me [for which I am thankful! criticism is good] for criticizing an idea of Hell that Catholics do not ostensibly believe in.)

(A lot of what I say could apply to every version of Christianity, I think, but I’m focusing on Catholicism because I know it best.)


There are at least two ways of talking about Hell.

The first is the way that preachers, writers of spiritual works, and the Bible seem to talk about Hell. In this way of speaking, Hell is the place where divinely-ordained punishments inflict pain eternally on those who have offended God. The pain is to a great degree physical, and is not self-inflicted by the sinner; Hell has flames, brimstone, lakes of fire, and wailing and gnashing of teeth. The pain breaks off the pleasure of the damned’s earthly life; the damned may have enjoyed their time on Earth, but now they regret it. And the people in Hell are confined to Hell by God against their will. This is, generally speaking, the popular manner of conceptualizing hell.

St. Francis de Sales’ Introduction to the Devout Life speaks about a “dark city, reeking with the flames of sulphur and brimstone, inhabited by citizens who cannot get forth.” One could easily pile up examples of this vision of Hell from sermons, devotional works, and the popular imagination, but something like this idea of Hell is also easy to find among the theologians. A majority of theologians have held there is pain of sense in Hell; Aquinas held that there was a literal fire inflicting pain on people in Hell; scriptural references to Hell as such a place abound.

The second idea of Hell is a little less obvious. Under this scheme, people in Hell torment themselves through their own interior disorder, because they have turned away from the goodness that is God; so their pain is to at least a great extent self-inflicted and internal. The pain is a continuation of the disorder that the sinner already began to inflict on himself in Earth through his evil actions; those who tore themselves apart through lusts or inordinate passions in life now find themselves forever at the mercy of the same. And finally, the people are in Hell because they want to be there; having fled God, this is the most pleasant place they could possibly be. This is, generally, the more intellectual manner of conceiving of Hell, at least right now.

Examples are a little more important for this way of understanding Hell. C.S. Lewis’ allegory about Hell in The Great Divorce is a preeminent example of this; in it, Napoleon forever paces his gigantic, empty palace, muttering desperately about how his failure was anyone’s fault but his own. In this, his lack of self-knowledge and narcissistic ambition punish himself; he cannot turn to any goodness at all, because he can never admit that he is flawed—and so he also cannot turn to Goodness in God. You can see a character heading towards Hell, so conceived, in Flannery O’Connor’s short story “Revelation”—actually, you can probably see characters heading towards Hell, so conceived, in very nearly all Flannery O’Connor’s short stories. The general characteristic they share is massive self-ignorance: they all have great, public flaws, which are visible to everyone around them—but they refuse to see these flaws and willfully blind themselves. This willful blinding of the self prevents them from following good on Earth, and correspondingly prevents them from taking pleasure in Good after life. (Of course, many of her short stories also feature God’s grace enlightening them to this fact through the bizarre and grotesque—but that’s another subject.)

The difference between these two Hells is important for what follows, so I’m going to draw the contrast between them a bit more explicitly.

People in the former Hell are there because God has forced them there; they wish to leave because they are in pain, and they would leave if they could. People in the latter Hell are there because they shy away from anything that would break their self-conception, and accept no goodness that does this; because of this, any place apart from Goodness is the most pleasant place for them in the entirety of the universe; they want to be there. The doors of the former are locked on the outside; the latter, on the inside.

The torments in the former Hell are (principally) external punishments inflicted on the sinner; the torments in the latter are the sinner tearing at himself.

The punishment of the former Hell stops the joys and pleasures of the prior life (which may have been quite pleasant) and makes them seem pale and insubstantial; the punishment of the latter Hell begins even during one’s life, and is merely a continuation of the misery in which one has already plunged oneself in life.

In short, in the former, God is the keeper of the cell, creator of the rack, and one who has cut off all one’s fun; in the latter, you yourself keep the cell, create of the rack, and destroy your own life. This contrast is certainly a little stark—you could try to soften it, for instance, by talking about how Hell could easily have both physical and psychological pains. But they are irreconcilable at least in aspects—it cannot be the case that you would like to get to Heaven and not like to get to Heaven, for instance. So I’ll let it stand while starting the historical overview.

Which of These Versions of Hell is Christian — History

I’ve tried to dig up every description of Hell I can find in the Old Testament, New Testament, and Church Fathers, and Church Fathers, to see what version of Hell Catholicism endorses.

I would provide the references that I’ve found, but I don’t see much point in it—see the appendix for a few, if you’d like. Every description that I can find lines up with the first idea of Hell, not the second. People in Hell want to leave and get to Heaven, but are walled out from it; the pains of Hell are contrasted with the pleasures they enjoyed on Earth; the pain is inflicted by some external thing. Enumerating these passages would be tedious—if you know of any passage that contradicts me, please tell me, but in the meantime I’m going to move on to issues that seem more interesting.

As far as I can tell, the earliest Christian writer who talks about Hell in a way resembling the second manner is Origen, in the early-third-century-work De Principiis: he interprets a vague passage in the Bible as a metaphor supporting his way of talking about Hell, with little internal textual support. Of course, Origen is one of the most (infamously) Platonist-influenced writers, and what he says has clear antecedents in the Platonic tradition, so this doesn’t seem terribly helpful for someone who wants to say that the more modern version of Hell is actually based in Scripture, Tradition, the Deposit of Faith, or anything Jesus actually said. Having a guy who became a heretic because of his opinion about Hell (albeit because of his opinion about Hell’s eternity) be the first one to advance your opinion about Hell is pretty much a worst-case scenario.

When I’ve told friends that I can’t find anything like this version of Hell in Scripture, friends have said that, well, you need to interpret the Bible. They’re surely right that you have to interpret anything. But you can’t just make interpretation say anything you want—you have to have intra-textual support for your interpretation, or else you’re just making stuff up. Origen has to turn an apparently unrelated passage in Isaiah into a metaphor for his view, to try to make it make sense—but if you allow yourself to (1) turn vague passages into metaphors for your view and (2) dismiss contradictory passages as metaphorical, then you could interpret anything to mean pretty much anything.

Of course, you could have a principled argument for such an interpretation. I don’t think I’ve ever seen one, though. Moreover—the only principled argument I can imagine for this conclusion, given the apparent contents of the Bible, would proceed from the idea that this idea of Hell is incompatible with God’s goodness, and that God would never establish something like that, and that therefore such descriptions of Hell must be metaphorical. It would, in short, turn on arguing that if Hell is what it seems to be described as in the Bible, then God is a monster. I imagine most people would be reluctant to argue this—it’s awkward saying that God as described by the Bible is monstrous, so the descriptions must be metaphors. And this is why, so far as I tell, apologetic discussions of Hell more or less assert Hell is a particular way without showing that teaching about Hell is that particular way.

Another Reason to Think the First Version of Hell is the Catholic Version

If Hell is a place that you go because you will not let goodness break your image of yourself, then the Church seems much less important for salvation.

After all, there are surely plenty of Catholics who die while turned away from goodness and self-knowledge—although they have taken the Sacraments. (This seems pretty necessary given that there seem like plenty of Catholics who live this way—it’s easy enough to find someone who seems ready to turn Catholicism into a tool to attack others and never reflect on his or herself. I’m pretty sure a fair number of sermons worry that Catholics do this all the time.)

It also seems likely there are many non-Catholics who die while ready to follow goodness and truth anywhere, although they haven’t taken the Sacraments. (This seems pretty necessary given that you can find non-Catholics who live this way.)

If the psychological model of Hell is true, then it seems like the Catholic should go to Hell while the non-Catholic should go to Heaven.

The Church has been pretty adamant that generally speaking this isn’t what happens, though—that Baptism and Confession and being a member of the Church are actually really, actually, quite key to salvation. The Council of Trent anathematizes those who say that Baptism is optional for salvation; in the same place it anathematizes those who say water is optional for Baptism, just in case you want to squeeze in salvation through a metaphorical kind of Baptism. (I have no idea how this is consistent with modern Church teaching, but that’s not my point at the moment.) This is in keeping with the Church taking Baptism very seriously, to a rather problematic degree, even in modern times. You could find similar statements about Confession very easily. The Church has generally taught, in many places, many times, and many ways, you can’t make it to Heaven outside of the Church.

Suppose you wanted to trash all the above tradition in favor of a more internal model of who makes it and who doesn’t. If this is the case, then I don’t see what the point of the Church is, other than to be one of the organizations that makes people better so they’re good enough not to go to Hell. And again, this kind of opinion seems pretty well and thoroughly condemned.

So, if the psychological model of Hell is true, then people would seem to go where the Church is pretty adamant they don’t go. So the psychological model of Hell must not be true—or at least a Catholic cannot hold that it is true.

I’m not really sure about the particular argument I just gave. I am quite certain that someone reading this has some kind of a sophisticated theory in mind which slips away from this argument—I’ve tried a few on for size myself, but none of them leave me feeling satisfied. Everything seems inevitably to slip afoul by making the Church superfluous (can’t say that) or making the Church only good for making people morally better in a natural way (can’t say that either). I’d be happen to listen to some such theory if it does manage to evade these problems, but I’m not going to try to imagine them all at the moment—comment away if you think you have a robust one.

Another Consideration

Attempt to hold the whole of Catholic art, song, sermon, and theology on Hell and the Last Judgement before your minds eye.

So we have nightmares in miniature by Bosch; we have The Last Judgement in the Sistine Chapel, where Christ sends the damned into punishment; we have every manner of torment that the damned attempt to escape in Dante’s Inferno; we have thousands of sermons or spiritual works on the eternity and pains of Hell, such as Bellarmine who speaks of the fire “whipped up by the breath of Almighty God” to punish sinners, or of De Sales who urges you to “stir your soul to terror” and then explains why you should be in terror; think of the purported visions such as that of Fatima, which speak at length of human souls like “transparent burning embers, all blackened or burnished bronze” being tossed up and falling into Hell alternately; think of the enormous amount of dread focused in prayer and piety on the hour of death and on judgement, when neither St. Paul himself nor any man sits secure but lives in fear in trembling, when only a few shall enter through the narrow road, when the rex tremendae majestatis comes to judge, and even the just require mercy, and when finally Christ thrusts who he will not save into the eternal fire prepared for Satan and his kind.

If Hell is not a place where people are unwillingly put after a judgement, then all this is scarcely intelligible. If people wish to be there, after all, surely they wouldn’t require a judgement to be put there. So if the psychological model of Hell is right, then it seems like enormous quantities of Catholic piety, closely connected with Catholic doctrine because of how they are tied up with the idea of the Last Judgment, are just wrong.

Final Point

You can actually make a point appear stronger by not arguing for it, and weaker by arguing for it. Giving someone an argument lets them think of counter-arguments, and if you’re sufficiently clever you can think a counter-argument to anything. So after thinking of a counter-argument to a position you dislike, you can think “Hey, this position seems weak. I don’t need to accept it.”

…but really, what we should be asking is not “Do arguments absolutely speaking force me to accept something, or can I wiggle out?” We should ask “What does the preponderance of the evidence point to?” I don’t know any good arguments to the effect that Hell is like the psychological model. So far as I know, the model came from people trying to make a barbaric institution sound sensible. But this model does not appear to be reflected in Scripture; it doesn’t mesh well with Church teaching about sacraments and the necessity of the Church; it doesn’t mesh at all with centuries of spiritual works and reflection on judgment and the last things.

So the preponderance of the evidence seems to be that the psychological model of Hell is wrong: The Church does not seem to teach that people in Hell actually want to be there, nor does she teach that the punishments they suffer there are principally self-inflicted.

Appendix on Scripture and Fathers on Hell, Which You Are Welcome to Skip

The Old Testament doesn’t give you much either way. The Hebrew word sheol seems to mean the subterranean abode of the dead in general, not a place for punishment. There is at least one verse which seems to support the idea that rich people who were happy will be a lot worse off once they get there (Psalm 49), and generally it doesn’t seem to be a nice place, there is no terribly specific information about it.

The New Testament is a great deal less ambiguous. In the New Testament, the people in Hell are held there by chains or chasms; they are thrown alive into Hell; they are like the virgins who wish to be in the wedding feast, but cannot make it (2 Peter 2:4, Mt. 25: 10, 2 Thess. 1:9, Rev. 19, and so on). People are physically tortured in Hell (Mt. 5:29-30, Mk. 9:43-48, Mt. 10:28). Those who were happy on earth are no longest happy in Hell (Lk. 16:19-31). I don’t really feel like going through all of these and other places in detail, because they’re all quite obvious, and probably quite present-to-mind to the reader. In any event, they each support the earlier, less psychological model of Hell: it seems a place where one is unwillingly kept, to be tortured, for the things one enjoyed on Earth.

So I don’t see any support for the psychological model of Hell in the New Testament—if I’ve missed a location, please tell me about it. What of the Church Fathers, then? (Note—my model for digging up references here is performing full-text searches on the CCEL. It is probably leading me to miss things, but it shouldn’t bias me in one direction or another as far as I can tell.)

Early on, the Fathers seem far more likely to say “Hey, you’ll go to Hell for this,” than to explain what Hell is like. The First Apology of Justin (chapter 19) says that Hell is where the wicked and non-believers go. Ignatius’ Letter to the Ephesians says that you can go to Hell for teaching false things. In his Letter to the Philadelphians (chapter 3) he’s even more strict and talks about how you can be condemned to Hell for spending time with unbelievers. The rule is hard to square with the psychological model of Hell—spending time with people doesn’t seem to turn you into a self-deceiving person, and indeed might seem to indicate the opposite—but in any event provides no support for the psychological model.

Tertullian’s discussion’s of hell, unsurprisingly, seems more inclined to the non-psychological side.

In another author in the early 3rd century, though, you can finally find discussion of Hell which is readily recognizable as belonging to the psychological side. This passage and this author disagree with me, so I should quote them at length

We find in the prophet Isaiah, that the fire with which each one is punished is described as his own; for he says, Walk in the light of your own fire, and in the flame which you have kindled. By these words it seems to be indicated that every sinner kindles for himself the flame of his own fire, and is not plunged into some fire which has been already kindled by another, or was in existence before himself. Of this fire the fuel and food are our sins, which are called by the Apostle Paul wood, and hay, and stubble. And I think that, as abundance of food, and provisions of a contrary kind and amount, breed fevers in the body, and fevers, too, of different sorts and duration, according to the proportion in which the collected poison supplies material and fuel for disease (the quality of this material, gathered together from different poisons, proving the causes either of a more acute or more lingering disease); so, when the soul has gathered together a multitude of evil works, and an abundance of sins against itself, at a suitable time all that assembly of evils boils up to punishment, and is set on fire to chastisements; when the mind itself, or conscience, receiving by divine power into the memory all those things of which it had stamped on itself certain signs and forms at the moment of sinning, will see a kind of history, as it were, of all the foul, and shameful, and unholy deeds which it has done, exposed before its eyes: then is the conscience itself harassed, and, pierced by its own goads, becomes an accuser and a witness against itself. And this, I think, was the opinion of the Apostle Paul himself, when he said, Their thoughts mutually accusing or excusing them in the day when God will judge the secrets of men by Jesus Christ, according to my Gospel.
This seems to be evidence for the individual supporting the psychological theory of Hell. The problem, though, is that the author here is Origen; and Origen’s interpretation seems to be more or less made up, as far as I can tell. Isaiah 50:11 could just as easily be interpreted to mean the opposite of what Origen says it means; after all, it goes on to say in it that “this is what you shall have from my hand: you shall lie down in torment.” As far as I can tell the first passage from Paul (1 Cor. 3:12) could as well be talking about crappy good works as it is about other things. And so on and so forth.

I’d like to note that the only conditions under which Origen could manage to advance the psychological view of Hell is as the only view of Hell (which he does not in fact attempt to do) is by taking (1) an ambiguous or vague passage as a metaphor for what he means and by (2) taking passages that disagree with him as metaphorical. I’m pretty sure such a standard of interpretation would allow you to interpret anything to mean anything.

I recently read something by an atheist who claimed that it was obvious that there was no God, shortly after which I read something by a theist who claimed that it was obvious that there was a God. Both of them took seeing the truth of their claim to be an elementary test of rationality or of good faith in seeking the truth.

I think the lesson to be learned from this has nothing to do with God.

First—what you see as true often seems obvious. It’s hard to imagine that other people would not see what you see as true as obvious, without imagining also that these other people are stupid or malicious. I think this is mostly a failure to imagine the other person’s perspective with anything remotely approaching verisimilitude.

Second, and more importantly—I think it would be best if people thought of the word “obvious” as something to be predicated not of states-of-affairs, but of states-of-affairs-as-seen-by-people. Most claims regarding obvious truths will appear non-obvious if you come from a community that sees the claim in question as false. Those who see the claim as obvious will say, of course, that coming from that particular, claim-denying community gives one a particular epistemological or motivational disadvantage. But those who come from that particular community will also say that the one who sees the claim as obvious comes from a similarly disadvantaged background. The exchange is circular—calling something obvious is pointless, and as a whole this again points to the idea that one should think of the word “obvious” as something to be predicated of states-of-affairs-as-seen-by-people.

Third—really, just don’t say that what you’re arguing about is obvious. You’re arguing about it, after all.

I was recently reading a (really excellent) book on Neural Networks and Deep Learning. The author spends a lot of time explaining both how to train neural networks and why why this can be very difficult.

(Start Jargon.) One of the problems that you can run into while training a neural network—while using sigmoid neurons, the mean-squared error loss function, and stochastic gradient descent—is that when the parameters for certain neurons are very far wrong, these neurons learn very slowly because changes to the weights of the neurons only offer small decreases to the loss. (End Jargon.)

Let me rephrase that. Sometimes neurons in neural networks learn slowly, even when they very far are wrong, because the rate at which neurons learn is proportionate to the improvement in network behavior that occurs because this learning. Due to the structure of the sigmoid neurons, when the neuron is far from its ideal value, any improvements made to it only improve the network by a small amount. So, to a certain degree, individual neurons—and thus part of the network—learn slowly precisely when you really want them to be learning fast—when they are most in error.

The author of the book then comments that humans are not like this at all. And he shares the anecdote of when he was playing the piano as a kid, and started on the wrong octave in his recital, and how he resolved never ever ever to make that error again—and indeed never did. So he says that humans seem very unlike sigmoid neural networks in this respect; when we are in error, we learn from it quickly. (He then explains a different loss function which helps resolve this error, but that’s not important right now.)

The thing is, I think humans might actually be really like sigmoid neural nets in this respect. We do often learn most slowly when we are most in error, and for pretty much the same reason that sigmoid neurons often learn slowly—our rate of improvement is often proportionate to the effect of our improvement, and improving when you are very far from correct can have very little influence. Let me explain what I mean.

Imagine that you’re in ancient Rome. There are a bunch of different philosophies to choose from. You are, let us say, an hylomorphist Aristotelian; you think that every living thing is the way it is because it has some immanent form inside it, which is a principle of life and of unity. But then you run into an atomist; the atomist thinks that every living thing is the way it is because it is made of a bunch of little pieces which interlock and hook together and somehow make living things do the things they do.

You argue over this. Maybe you’re convinced by him. Maybe you aren’t. Suppose you are. In fact, your opinion is now much closer to the truth than it was before—atomism has subsequently been shown to be a really useful way of trying to explain reality, although no one has yet figured out what to make of hylomorphism. But, although you’re correct, this knowledge doesn’t actually help you navigate the world better. Applications for atomism are still far away; knowledge of the scientific method, which would help you produce applications from atomism, are also still far away; the social and technological apparatus necessary for atomism to be useful are all far away. So your correct opinion has pretty much no influence on your life.

And indeed, this is one of the reasons that your conversion to atomism is unlikely. If the atomist were able to better navigate reality, in demonstrable ways, then it might be harder to resist conversion. But, even though atomism is closer to the truth, it is still so far from the truth in other respects that it is about as useless as hylomorphism; so it is easy to not convert to it.

Here’s another analogy.

Imagine that you’re learning to golf. A friend of yours is teaching you. He notices that your grip is totally wrong, so he spends 15 minutes showing you exactly how to grip the club. And afterwards, you take a few more swings—and you still suck just as much as you did before, because to swing well you need to have more go right than just your grip.

Now, if you manage to remember how to keep your grip as you learn the rest of the aspects of golf, this knowledge might be useful. But as it is, it will be hard to remember, because this knowledge doesn’t really do anything for you, yet. And generally this is one of the things teachers might be good for—they keep you going down the correct path, even when that path doesn’t result in any visible improvement. But despite teachers, that the grip seems useless is certainly going to slow down the speed at which you learn it; if you had everything else right, and then improved your grip, you can bet that the speed at which you learned the grip would be much greater. So once again, learning is slower when you are far from where you should be than when you are close to it.

Let me speak more generally.

There are fields of knowledge, and fields of error. Fields of knowledge are things that attempt to describe reality, and actually do so—chemistry, for instance. Fields of error are things that attempt to describe reality, and systematically fail—alchemy, or astrology, for instance. I think one of the reasons that people are willing to remain in fields of error is that you don’t see any improvement when you begin to leave such a field—you pretty much stay as good at understanding reality as you did before. Only once you have exited some erroneous field, and begin to enter some accurate field, does your skill-at-life improve. Until then it just seems like you are trudging away from some field, and gaining nothing in return. And, given this, this is why it is easy to remain in a field of error—it takes a long while for you to begin to benefit from leaving it.

You can compound this problem with the fact that humans don’t like to admit that they are massively wrong. Instead, they like to think that they were basically right, to make small concessions to inescapable arguments against their position bit by bit, and generally to reluctantly retreat before an enormous mound of truth. This just guarantees that, as you retreat from a some field of error, you’ll take an eternity to begin to benefit from it. Which—because the speed of human learning often depends on the visible increase in skill that we get from that learning—once more will the speed of learning slow to a crawl.

So yeah, I think humans are pretty cool being in error. I think this is good to know, because, knowing this, one can try to speed up the rate at which one admits that one is in error, and move to a place where the truth benefits one.

Also titled, It is Impossible To Seek Evidence that Confirms Your Hypothesis, only Evidence that Tests It, Except for Maybe in a Few Odd Cases, And I Think Not Even Then

(I have a feeling I’m getting the titling capitalization wrong.)

(Caveat Lector: I have stolen examples freely in the below from friends, LessWrong posts, books, and other locations.)

0. Intro

Sometimes Very often, one is not certain whether a particular statement is true or false. To resolve this uncertainty, one can seek evidence. The point of this post is that it is impossible to seek exclusively confirmatory or exclusively disconfirmatory evidence.

To put this another way, it is impossible to seek out confirmatory evidence without at the same time risking disconfirmatory evidence. To put this yet another way, it is (at least almost always, and probably always) only possible to seek to test your hypothesis—to seek evidence which will move you to think either that a hypothesis is more likely or less likely to be true. You cannot seek evidence that will move you to think that a hypothesis is more likely to be true, without running the risk of finding that your hypothesis is less likely; and you cannot do the reverse either.

This post will also show that any procedure that promises to make you more confident in a hypothesis, without the accompanying risk of making you less confident in your hypothesis, cannot actually provide evidence.

There’s a mathematical way of saying this, which is superior to the non-mathematical way. But I’m going to attempt to describe this first with words, rather than with math. After I’ve tried to describe it non-mathematically, I’ll move on to the math, which will, at that point, hopefully make sense.

1. First Examples

Imagine that a doctor thinks that a particular chemical substance is likely to cure cancer.

So he decides to set up an experiment to test this. He would like to have definitive evidence, after the experiment is over, that it cures cancer. So he sets up the experiment very, very carefully: he draws from several different demographics when assembling his control and experimental groups; he tries to ensure that each group is large enough to detect even small differences in outcome with p < 0.01 significance; he is really careful to try to anticipate and avoid potential confounders; he does all the double-blinding that every good scientist would do; he probably does all sorts of other things that I can’t even name because I don’t know enough about experimental design. He wants the evidence that pops out of this experiment to be weighty.

And he goes through all this, performs the experiments, waits five years for a bunch of data to come in, and eventually finds that he can’t reject the null hypothesis—the drug probably does nothing for cancer. And all the careful design of his experimental procedure now only means that this is a more crushing refutation.

Note that the only possible way he could open himself up to evidence in favour of the hoped-for outcome was by opening himself up to evidence against the hoped-for outcome. If he had screwed up the experimental design, so that it was more likely to show the experimental group getting better than to show the control group getting better, apart from the efficacy of the drug—well then, obviously the inertness of the drug could not have been indicated by the experiment, but equally obviously the experiment would be useless as evidence showing that the drug was effective. In short—because finding evidence for his hypothesis necessarily involves the possibility of finding evidence against it, you ultimately would have to characterize running the experiment as an attempt to test his hypothesis, not an attempt to confirm it, even if that is what he wanted.

Here is another example.

Imagine that someone creates a new aluminum-steel alloy. They want to show the world that it is, per unit weight, stronger than titanium.

Well, in this case the only thing they can do is set up an experimental rig that smashes into it and see if it breaks more slowly and rebounds more effectively than titanium. And of course, in doing this, they also run the risk of finding that it actually breaks faster and rebounds less effectively than titanium. To show that it is stronger—as they hope to do—they also necessarily run the risk of showing that it is weaker.

Again, you can’t seek to confirm that what you wish to believe is true; you can only test it. The reason for this is that, if the test that you perform is not causally entangled with the thing that you wish to examine, then it cannot provide evidence for your favoured hypothesis about the examined thing: if the test apparatus shows that the steel-aluminum alloy does not break, regardless of whether it is actually strongest, then it cannot provide evidence for the alloy being the strongest. But if the test must be causally entangled with the thing that you wish to examine, then it also necessarily involves the risk of providing evidence contrary to your hypothesis: it risks showing that the evidence does not turn out as your favoured hypothesis predicts that it would.

Well, someone could object—maybe this is true of the hard, experimental sciences. What about something else, such as history?

It works the same in history. Suppose I’m interested in knowing whether Roger Bacon anticipated some optical notion heretofore attributed solely to Christiaan Huygens; I think it would be awesome if he did. So I decide that I’m going to go through every text attributed to Bacon, as well as every questionable text and every oblique reference I can find to him and his works in the corpus of 13th century literature. 25 years, 20 scholarly articles, and 1 dissertation later, I complete my survey—and find, to my dissatisfaction, that Bacon did not anticipate Huygens. This is contrary to my wishes, but now far more definitively proven than it would have been if I did nothing. Again, the opportunity of finding evidence in favour of a hypothesis must be balanced by the opportunity of finding evidence contrary to it.

Ok, well, you could say—maybe this is true of academic things. What about people’s personal lives?

Same here. Suppose you would like to find if someone is trustworthy. If that’s the case, then you’ll probably have to trust them with something—whether big or small. And doing this involves running the risk of having them betray you, and finding that they are not trustworthy.

I could go on with examples forever, but I hope my initial point is clear. You cannot seek confirmatory evidence for a hypothesis; you can seek merely to test a hypothesis.

2. An objection, turned into an advantage

Here’s an objection I suspect some people are thinking about: It really feels like you can seek confirmation of a hypothesis. Let’s look at a case when it feels like you’re successfully seeking confirmation of a hypothesis.

Suppose I think that the Earl of Oxford was Shakespeare. I cannot really give any arguments for it, though, if you ask me for arguments. So I go to the library, check out a dozen Oxfordian books on Shakespeare, and read them all. I also browse through a bunch of Oxfordian websites online, and look through all of their arguments. I also look through their dissection of all the anti-Oxfordian arguments, just for completeness’ sake. So now I’m able to give a dozen different arguments and pieces of evidence in support of the Oxfordian cause, whereas before I wasn’t able to; I was in a prior state of ignorance, and deliberately sought confirmatory evidence that allowed me to be in the posterior state. So it seem to me pretty clear that I was able to seek and to find confirmatory evidence in favor of my hypothesis.

And this is a fairly universal experience. I can start by thinking that some proposition is true, and by wishing that I had more evidence that it was true. To find such evidence, I can read books by other people who think that this proposition is true. And then I find that I apparently have attained confirmatory evidence in favour of my hypothesis, which I have sought. So it seems false to say that I cannot seek out confirmatory evidence for a hypothesis.

I’m not going to address this argument directly, at first. Let me give an analogy.

Suppose you’re interested in whether a particular die is loaded or fair. If it is loaded, it is loaded so that the higher triplet of numbers (4,5,6) comes up far more frequently than the lower triplet (1,2,3) of numbers. If it is fair, then members of each triplet will come up with roughly equal frequency. A good way to test this would be to roll the die 100 or 1000 times, and count how many times it comes up high and how many times it comes up low.

We’re not really interested in doing this tedious experimental work ourselves, though, so we ask a friend to do it and record his results on each throw.

After the experiment, we ask him how it went. And he says “Well, it came up high on the 3rd, 4th, 10th-13th, 16th, [insert long sting of text], 89th, 91st, 95th-98th, and 99th rolls.”

We say “Um, so you mean it came up low on the other rolls?”

He says, “Oh, no, I don’t mean to imply that. It is conceivable that it could have come up low or high on them. I’m just not reporting them, right now.”

And we ask to ourselves, “Hmm… so is this evidence in favour of the die being loaded? All of the instances that he reported were instances which were high. The particular series that he gave us is the kind of series you would expect if the die were loaded. But on the other hand, if he has resolved only to report rolls where it came up high, that doesn’t count as evidence pro at all—he’s just tilting his reporting of the experiment so it seems that it came out in favour die being loaded.”

So we ask him, “Had you resolved to report only on those aforementioned throws before you actually made the throws?”

There are two ways he could respond at this point.

Suppose he says “Yeah, before I rolled the die, I decided only to report the 3rd, the 4th, the 10th-13th, [etc], rolls. And then I stuck to that resolution after I rolled the die.” If he is telling the truth, then we actually do have evidence that the die is loaded—it’s as if he had performed exactly the experiment we wanted to perform, merely with a smaller number of rolls. If the die were not loaded, we would have expected this sample to indicate that the die were not loaded; we were open to refutation as well as confirmation, and so we have found that we in fact genuine evidence in favor of the loading.

On the other hand, suppose he says “Nah, I decided just to report on those throws after I had made them,” and then winks. If this is the case, then we have a very strong suspicions that he is only reporting on throws which turned up with members of the higher triplet. And if he’s filtering the evidence in this way, then we obviously do not have real evidence in favour of the dice being loaded. His resolution to only report evidence that seems to favour the point that the dice is loaded makes the so-called evidence not evidence at all.

So in this case, our friend can only offer real evidence to us if he has resolved to offer all the evidence he finds, whether it supports or does not support the hypothesis that the die is loaded. If he is deliberately filtering out evidence in accord with whether he likes it or not, then he isn’t offering evidence: he’s offering a simulacrum of evidence, which is designed to produce assent but which has nothing to do with the truth of the matter.

Let’s turn to books.

Suppose that a Scientologist is uneasy in his Scientology. So he reads a bunch of books on Scientology by Scientologists, and in the books he finds all sorts of bits of evidence in favour of Scientology. He finds many stories about people who went through their auditing treatment, and whose lives then improved in amazing ways; these stories are evidence that auditing works, he thinks, and if auditing works, well, this is evidence for the historical stories about Thetans that the practice of auditing is based on. He also finds stories about how people tried to dig up scandals about Scientologist leadership, and about how these attempts were thwarted and turned out to be based on lies; and if the arguments against Scientology turn out to be based on lies, he thinks, then surely these arguments are not formidable. And he also finds really insightful and interesting things that L. Ron Hubbard said about human nature, which really help him explain himself to himself and navigate the world better; and if some of the things that L. Ron Hubbard seemed really enlightening, he thinks, then maybe he should be more trusting when L. Ron Hubbard says something counterintuitive. All the things that the books says might be true as well—he could verify them all through third sources. And so he seeks bits of information and evidence for Scientology; and so he keeps himself in his delusion.

It’s pretty obvious that the author of the book on Scientology is acting like our deceptive friend in second imagined scenario above. Just as the friend reports only evidence which seems to suggest that the die is loaded, the author of the book is only reporting evidence which suggests that Scientology is true: each is only reporting evidence that appears to favor a particular hypothesis.

And, I hope it is now evident, if the author is doing that, then the evidence that they are reporting simply should not count as evidence, just as our friend’s supposed evidence simply should not count as evidence. You can find pieces of evidence to support any hypothesis you please—the world is very large. And this means that, when arguing for a point, everyone has the opportunity to filter and select like the Scientologist author or the die-rolling friend. But if the die-rolling friend is not providing actual evidence, because he filters—and he clearly isn’t—then neither should the scientologist author or any author who reports only on positive evidence be seen as actually providing evidence. That they’re acting as this kind of filter rules out the possibility that they are providing evidence: instead, they’re providing what I see as a very dangerous, rhetorical simulacrum of evidence.

Of course, this simulacrum of evidence can be improved; both our friend and the scientologist could really work on presentation. Our friend above could include a few cases when the die roll turned out low—that would make it look more like he was being honest. Similarly, the Scientologist author could include a few cases when Scientology was abusive and deceptive, and admit that some unfortunate abuses have occurred—and thereby gain a greater appearance of honesty.

Our friend could even be really devious—he could include all the die rolls, but then mention that his die-rolling technique was bad in certain, particular cases—“Really, I wasn’t rolling the dice so much as setting it down; I wasn’t following the ISO-312 Die-Rolling Manual; and there might have been a magnet near these die in the case of these rolls”—and then, having performed unequal scrutiny on different die-rolls, show how miraculously the really good rolls were only those that turned up higher. The scientologist author could do something similar, and provide a few of the genuine arguments that people make against scientology—but then hold them to much higher standards of evidence than he holds other arguments, or present them in rhetorically unsatisfiying manners.

But doing all this would merely window-dressing which is designed to make the evidence-substitute look like real evidence, as the two examples should make clear.

What one is doing, then, in the case of the Oxfordians should be evident. If the authors that you read are acting like the scientologist author, then what they give is a simulacrum of evidence. If, on the other hand, they present the best evidence in both directions, the most deadly attacks on their own position, as earnestly and rigorously as they present those for the other position; if they have spent equal time searching for flaws in their own arguments and for the flaws in the arguments of their opponents; if they have tried to see the world through the eyes of their opponents, and tried to examine how their supposed evidence would seem in such a case, and how their opponents might effortlessly explain it; if they have done all this, then there is a small chance they are presenting real evidence. Even after reading authors like these, it is still best to read counter-arguing authors, as everyone knows—it’s hard to present evidence for a position you disagree with fairly, even if you’ve resolved to try to do so.

And if the author you read did all this, and if you followed this writing pattern, then reading them might not convince you of the Oxfordian hypothesis—it might convince you of the opposite. So once again, it seems that if one wishes to deal with real evidence in favour of a hypothesis, you cannot ensure that you will not also encounter real evidence against it. You cannot search for confirmatory evidence; you can only search for a test.

3. More objections, which are hopefully fairly stated.

I just laid out a really, really high standard for writers. Let me attempt to meet it.

For a while I wondered whether what I have said applies to the domain of mathematics. The line between valid and invalid arguments seems to be sharper in mathematics than it is everywhere else. A valid mathematical proof, baring you insanity or a mistake made within the proof, seems like fairly absolute evidence for what the proof purports to be a proof of. So, supposing that I sought a proof for a mathematical theorem and found it, it seems like I’ve sought confirmatory evidence without running the risk of finding disconfirmatory evidence.

I think ultimately this is wrong. I could have made a mistake in my proof—this happens all the time. Alternately, while searching for a proof, I could stumble across a reason for thinking that the theorem is false—this kind of thing also happens. And if I search for a proof for a while, and fail to find one, then that fact itself could be evidence against the existence of a proof. So searching for a proof in math does not, as far as I can tell, provide a chance for confirmatory evidence without disconfirmatory evidence.

Suppose, though, that we grant that you can do this in math. I’m ok with that. Mathematicians generally aren’t involved, qua mathematicians, in the kind of disputes and cases of tendentious reasoning that the rest of the world is involved in—or at least that’s what it looks like from the outside. And math is pretty far removed from the rest of a lot of things. So I wouldn’t really mind including an ad hoc exception for mathematics, although I think it is unnecessary.

One could try to extend this (probably unnecessary) ad hoc to things like philosophy, but that would pretty clearly be unneeded. If we are to have an ad-hoc addition to the theory for mathematics, it would only be because it is (nearly) impossible for a conscientious individual to offer unsound arguments in mathematics. (Again, I think this is pretty clearly false.) The notion of proof, of what constitutes acceptable premises, and of what constitutes acceptable work—all of these are far more hazy in philosophy than in mathematics. So the ad-hoc extension does not work, as far as I can tell.

Another counter-argument you can give to this standard is that it would make everything that people write interminably long. I concede that this counter-argument is the case. But sometimes if you are interested in the truth you have to read interminably long things. That’s just how life is.

4. Another counter argument, which isn’t actually a counter-argument at all

Here’s another counter-argument.

Suppose that Kimiko, a high-school student, thinks her boyfriend Bob is cheating on her with Tracey, even though Bob claims that he never even talks with her. Naturally, as the trusting soul she is, Kimiko decides to a install a backdoor in Bob’s cellphone, so she can remotely monitor whether he receives or sends any compromising texts or calls. She installs it early one morning at school, and she checks the backdoor fours hours later at lunchtime. There are two possible results she could find after checking: Bob might have sent / received a message or call, or he might not have sent / received a message or call.

If Bob has received or sent a sexy message, this serves as extremely strong evidence that Bob is cheating: if Bob gets a message from Tracey saying “Beast w. 2 backs 2nite!” then Bob is very probably cheating. On the other hand, if Bob has not received or sent such a message, this is only very weak evidence that Bob is not cheating: even if he is cheating, he might not communicate with her every four hours. Both the world where Bob is cheating and the world where Bob is not cheating would very likely lead to a situation where Bob does not receive any message from Tracey—and for this reason, his not receiving a message seems like only weak evidence that he is not cheating. On the other hand, the world where Bob is cheating is far more likely than the world where Bob is not cheating to involve such a message, so receiving such a message is very strong evidence he is cheating.

So it seems as if Kimiko can seek evidence confirming that Bob is cheating, at least inasmuch as the strength of the possible evidence confirming cheating is different from the strength of possible evidence disconfirming cheating.

The problem is that I actually agree with everything stated above—it’s actually quite vital to my point, and not a counter-argument at all. However, it still would be wrong to characterize Kimiko’s actions as an attempt to confirm cheating, for the following reason.

No messages from Tracey would likely happen in both worlds where Bob is cheating and worlds where Bob is not. So we actually have a relatively low expectation of seeing any message, although a message would be relatively strong evidence. Similarly, because a message from Tracey would almost certainly only happen in a world where Bob is cheating, and perhaps not even in that, we have a strong expectation of seeing no message, which serves as weak evidence that Bob is not cheating. So even though possible evidence is strong in one direction and weak in the other, this is balanced by having a strong expectation of the weak evidence and a weak expectation of the strong evidence.

Let me give another example to try to illustrate the idea.

Suppose I’m going through Roger Bacon’s letters, over the course of going over the entirety of his work. It might be that many of his letters deal mostly, although not exclusively, with matters of a personal nature. So I have very little expectation that I’ll run into any theories previously attributed solely to Huygens in these letters—although, if I run into an apparent instance of such a theory, I’ll have to revise my certainty about Bacon’s opinions drastically. So I have an (extremely) low expectation of (extremely) strong confirmatory evidence while reading these letters. On the other hand, if I don’t run into any such theory, I’ll only revise downwards very slightly the probability that Bacon anticipated such a discovery, because this is not the first place I would expect such a discovery to be written down. So I have an (extremely) high expectation of (extremely) weak disconfirmatory evidence.

One way to say this is that, on average, your anticipated future opinion after running into evidence must be, on average, the same as your current opinion. If you anticipate really strong evidence that might move your opinion a great deal in one direction, but cannot see any way for there to be really strong evidence that moves your opinion in another direction, then you must have only a very weak expectation of very strong evidence in one direction and a very strong expectation of very weak evidence in the other direction. This is why you cannot seek confirmatory evidence, then—your anticipated future average opinion must be the same as your current opinion.

Let me give one more example.

Suppose you are a Catholic, and you hear about some supposed miracle. The Catholic Church tends not to actually endorse miracles; it tends to be reluctant to say “This is miraculous and a sign from God.” Furthermore, you know that there are many fraudulent miracles in all religions; the Catholic Church is no exception. So you think that this particular miracle is very likely a fraud as well, but investigate it nevertheless.

Now, if you investigate extremely carefully, and find no conceivable scientific explanation for the phenomena, then you have pretty good evidence for Catholicism. But you did not expect to find such a miracle when you started to investigate the miracle; you thought it was probably a fraud. So you had a weak expectation of strong evidence, starting off. If, on the other hand, you were to find that the miracle was a fraud; well, that’s what you expected anyhow. You had a very strong expectation of this extremely weak evidence that Catholicism is false, although this evidence is so weak it is probably not even worth keeping track of mentally. (To see that it is nevertheless contrary evidence, imagine what it would be like to investigate a thousand such supposed Catholic miracles and to find every single one of them to be frauds. The thousand is composed of ones; each of the thousand shifts probability by a calculable amount, given a particular amount by which the complete thousand shifts probability.)

5. You’ll skip this part

Now, there’s actually a mathematical way of saying, and proving, what I’ve been trying to say above. I was originally going to go through this really slowly, but then I realized that (a) everyone who doesn’t like math would probably skip this section anyhow and (b) everyone who does like math would prefer that I just give the proof. So let me give the proof.

In the below, “H” stands for the hypothesis under consideration and “E” stands for yet-to-be-found evidence supporting “H.”

1. P(H)
2. P(n) = P(n,m) + P(n,~m)
3. [From 1 and 2]: P(H) = P(H,E) + P(H,~E)
4. P(n,m) = P(n|m)P(m)
5. [From 3 and 4]: P(H) = P(H|E)P(E) + P(H|~E)P(~E)

This formula [P(H) = P(H|E)P(E) + P(H|~E)P(~E)] says most precisely what I am trying to say above.

It means that a weak expectation [low P(E)] of strong evidence in favor of the hypothesis [higher positive P(H|E) - P(H)] must be accompanied by a strong expectation [high P(~E)] of weak evidence contrary to the hypothesis [slightly negative P(H|~E) - P(H)]. This is like the case with Kimiko and Bob.

It also means that a medium-strength expectation [middling P(E)] of strong evidence in favor of the hypothesis [higher positive P(H|E) - P(H)] must be matched by a medium-strength expectation [middling P(~E)] of strong evidence against the hypothesis [lower and negative P(H|~E) - P(H)]. This is like the case with the cancer-curing drug.

6. Finis

Why should anyone care about the above?

Well, I think it is important to show that you cannot actually seek confirmatory evidence in favor of a hypothesis. You can seek to convince yourself that a particular hypothesis is true—but what you seek in so trying to convince yourself is not evidence, but a simulacrum of evidence. That seems important, because it can help one distinguish, in one’s own actions, between when one is performing an exercise in self-directed rhetoric and when one is attempting to find the truth.

In my experience, making this distinction is difficult to do; I still wonder all the time which one I’m doing.