Universalism and Particularism
(Immigration and Religion, pt. 3)
Controlled Flight Into Terrain

The Bay Area Unitarian Universalist Church (BAUUC) and The Clear Lake Islamic Center (CLIC)

(Immigration and Religion, pt. 4 of 4)

Peter A. Taylor
August, 2017

The actions of radical Islamophobes should not be used to condemn the peaceful Islamophobe majority.
Allum Bokhari

(This was written in dribs and drabs between September, 2016 and August, 2017. This essay continues a long-standing theme of mine that the messiah that I'm looking for needs to be a statistician. It is intended to be read after reading three previous essays on immigration, religion, and universalism)

Is Unitarian Universalism a "good" religion? This is sort of a trick question. The question I'm really trying to get at is whether we can say with anything approaching objectivity that any religion is "good" or "bad". Are there some criteria by which a reasonable person would agree that one religion is better or worse than another? Imagine sitting down with Vernor Vinge and trying to write a science fiction novel with a religion as a plot device that causes a lot of problems. We could start with some real-world examples. Aztec religion? The Thuggee cult? The early Mormons? Mencius Moldbug had a thought experiment (in the context of whether "religion" was a useful philosophical category) in which National Socialism was converted into a "religion" by grafting on a requirement that members believe in the literal existance of Thor.

One plot device a science fiction writer could use for a story set in the US would be to take advantage of the disestablishment and free exercise clauses in the First Amendment. In effect, the US government is constitutionally forbidden from fighting a religious war. So what happens if the US is attacked by a religious movement? Larry Auster has suggested that this "plot device" can't really be addressed effectively without a constitutional amendment.

What are some other ways in which a religion can be problematic?

  1. Religions can have rules like the rules about driving cars. Americans drive on the right side. British drive on the left. Neither approach by itself is inherently better or worse than the other, but they don't play well together. If you mix British and American drivers without either being forced to assimilate to a uniform standard, you get a lot of car wrecks. Imagine if Anglican churches were to call on their members to practice altruistic punishment of people who drive on the right, and Congregationalists were to demand altruistic punishment of people who drive on the left.

  2. A religion can be good in one context, but bad when taken out of context. Early Christianity may have made much better sense in the context of Jews living under Roman occupation than it did post-Constantine, when Christians had to take on the role of the hegemon. Islam may have been an improvement over some of its predecessors (e.g. if you were low caste, it was better than being part of India's caste system), but may not lend itself to multiculturalism and globalization as well as competing religions.

  3. A religion can be good for some people and bad for others. Jim and Tammy Faye Bakker's religion may have been good for the Bakkers, at least financially, but bad for their supporters. A religion (e.g. the Thuggee cult) could be good for its members, but bad for nearby non-members. There could even be a public good problem within the religion. Maybe having a lot of Thuggees around is a net bad thing, even if you're one of them.

  4. Certain aspects of religion may be so subjective that we can't even talk about the criteria "a reasonable person" would choose. Maybe what works for one person is totally different from what works for another. Maybe Buddhist meditation works for some, but not others.

  5. This is not an exhaustive list. A religion can make false factual claims which might matter if taken literally, or only benefit people who died long ago, etc.

I say that whether UU is a good religion is sort of a trick question because I'm a recently lapsed member of BAUUC. Surely I wouldn't want to be a member of a bad religion, would I? But I've actually been pretty critical of UU in several places (e.g. here and here). UU is good for me in that they put up with me, an atheist, where a serious Christian church would not. But if I had to choose between living in a democracy with a majority of UUs vs. a majority of evangelical Christians, I might be better off with most of my neighbors being evangelical Christians.

Another thing that makes religion hard to talk about is the difficulty of finding the right "unit of analysis". It's hard to say where one religion leaves off and another begins. Jason Pullen says, "There's no such thing as a non-hyphenated Christian." That is, the differences between different groups of Christians are large in comparison with the differences between Christians and Buddhists. Is Rowan Williams really the same religion as John Calvin? Is Christianity one religion, or is each denomination its own religion, or each congregation? We could take the view that each independent organization is functionally an independent religion. On the other hand, we could say that everyone who venerates more or less the same holy book (e.g. Protestant and Catholic Bibles being essentially similar) is the same religion, and that the differences between co-religionists are the result of other variables. Statistically, we might say that behavioral variance between Methodists and Presbyterians is variance "within group" and that variance between Christians and Buddhists is variance "between groups".

Larry Iannaccone, in Deregulating Religion (and in the economics of religion literature in general), measures religious diversity (Herfindahl index) based on each denomination being an independent unit. Iannaccone shows a Herfindahl index for Protestant denominations in the modern US of less than 0.05 . But if you look at Protestanism as a single religion, at the time when the 1st Amendment was passed, the Herfindahl index for all religions in the US was something like 0.96, and if you view Christianity (including Catholicism) as a single religion, it was something like 0.9999 . Did religious freedom help build up the nation's moral capital, or did we start off with a huge stash of moral capital that we've been allowing to deteriorate by allowing too much religious freedom? Depending on your unit of analysis, you could argue either way.

Which approach makes more sense, viewing Christianity as a bunch of different statistical groups or as a single group with a bunch of confounding variables? To answer that, I might want to know how loyal Methodists are to Presbyterians vs. Buddists or Santeros. Will they stick together in a fight? I also might want to know how much social mobility there is between the two groups. How likely is the daughter of a Methodist to join a Presbyterian church vs. becoming a Buddhist or Santera? How much do Methodists and Presbyterians mix socially? How likely are their theological arguments to be cogent for one another?

If the behavioral variance within the various denominations is large compared to the variance between denominations, then there might not be much point in treating the denominations as separate groups. For example, if the differences between individual Methodists and between individual Presbyterians are large, the differences between Christians and other religions are large, and the overall difference between Methodists and Presbyterians is small, then I might as well lump the Methodists and Presbyterians together for purposes of understanding Christianity.

We could talk about this in terms of Markov chains. We could define a vector to store the probability distribution of religious affiliations within some population, and a transition matrix that shows how the sub-populations migrate over time. If the transition matrix is an identity matrix, the religions are totally independent. If the off-diagonal terms are large compared to the diagonal terms, the religions are indistinct.

The same question arises in trying to understand Islam. Is "Islam" a religion, or does each imam have essentially his own different religion?

In a previous essay, I said "Bibles and theologians are like cloth and tailors. The tailor is more important than the cloth, but the cloth still makes a difference." I tend to think of Islam as one religion because it is all based on the same holy book. Different imams are different tailors, but they're all working with more or less the same cloth. I also perceive a lot of loyalty between different groups of Muslims. Iranian Shi'ites seem to get worked up pretty easily about conflict between Palestinian Sunnis and Israeli Jews. I'm told that Sunnis and Shi'ites get along fine in CAIR. There also seems to be a great deal of mobility between "moderate" and "radical" versions of Islam. The father of Syed Rizwan Farook, the San Bernardino shooter, is not a troublemaker, but the son is. So I'm inclined to think of Islam as one statistical group with a very large behavioral "variance within group". But your mileage may vary.

It's also not obvious why any of this matters. If I did decide that a particular religion is "bad", that doesn't necessarily mean that I want to repeal the First Amendment. I may think that the nearby church bells are a public nuisance, but not nearly as big a nuisance as a repetition of the Thirty Years War. Nor does thinking that a religion is "bad" necessarily imply that I should be impolite to its members, boycott their businesses, or refuse to offer aid in an emergency. But I would at least like to be able to frame the question clearly.

Unitarian Universalism preaches religious tolerance and "acceptance". We don't want to re-fight the Thirty Years War, and we want to treat people as individuals, but our opinions are varied regarding certain religions. Many of us were raised Christian and emphatically rejected it. We often sing John Lennon's "Imagine" as a hymn. We borrow a lot from Christianity in terms of low church trappings and style, but we approach its theology and moral teachings selectively. We tend to paper over a lot of theological disagreements by keeping things vague. But there is also a joke that the only time Jesus Christ is mentioned at a UU church is when the janitor falls down the steps. A lot of us like it that way.

I would have expected members of my church to be similarly circumspect in expressing their affection for Islam. Islam is generally not known for its acceptance of homosexuality, religious tolerance, women's liberation, and other things that are near and dear to most UU's hearts. So when my church (BAUUC) held a rally in support of our neighbor institution, The Clear Lake Islamic Center (CLIC), I viewed this with a certain degree of puzzlement. In particular, some of the people who were especially enthusiastic about the rally were people whom I know to also be especially enthusiastic about gay marriage, women's liberation, and separation of church and state.

This creates at least the appearance of a double standard: Homophobia is bad, and it's okay to criticize Christianity when Christians don't approve of gay marriage, but it's not okay to criticize Islam when a Muslim shoots up a gay bar in Orlando, etc. (Rod Liddle notes the same phenomenon w.r.t. leftist criticism of mild homophobia in Orthodox Christian Russia, but not severe homophobia in Muslim Bangladesh.) Part of the context for this puzzlement is a debate about immigration in the 2016 election season. Donald Trump has come out in opposition to immigration by Muslims in general. There is also war in Syria, and there has been some debate specifically about immigration by people claiming to be Syrian refugees.

So I find myself scratching my head about the meaning of BAUUC's support for CLIC. Several possibilities come to mind.

  1. We have coherent ideas about what is and is not a good religion, and have come to the reasoned conclusion that Islam is a "good" religion.

  2. "Islam" is too broad a category to analyze. We have to look at a specific version in order to categorize it. CLIC is good even if much of the rest of Islam is bad. For example, ISIS has made death threats against Imam Basyouni, the leader of CLIC, so he must be doing something right. Maybe we're deliberately trying to promote the CLIC version of Islam at the expense of ISIS.

  3. We have enough information about the individual people at CLIC to be able to treat them as individuals. Some individuals will be good people regardless of whether their religion is a help or a hinderance in this regard, and we don't want good people to be punished for the behavior of bad people that they don't have much influence over. We want each member of CLIC to be judged as an individual. (The claim that we have enough information to judge people individually could also be extended to refugees. See paragraph 4.B. in my above border policy essay.)

  4. We have double standards. Islam is a stick with which we wish to poke the Christians. We condemn Christians for "homophobia", but Muslims get a pass. (I also have a chip on my shoulder regarding Christianity, which I explain briefly here. In addition, I recommend Mencius Moldbug's serialized essay, "How Dawkins got pwned". Find the rest here.)

    Update, 8-10-2018: You can say that the double standard is justified because the Christians are an "oppressor" group and Muslims are "oppressed", but in doing so, you're acknowledging that the double standard exists. In order to not have a double standard, you have to have the same standards of evidence for both groups, and in order not to have a circular argument, you have to evaluate the evidence before deciding who the bad guys are.

  5. Maybe peer effects are such that whether Islam is good or bad according to some criterion depends on their numbers. In Western countries where Progressivism is dominant, Muslim leaders generally take pains to support gay marriage (or at least maintain a "don't ask, don't tell" posture regarding homophobia). But in majority Muslim countries like Saudi Arabia, a very different pattern emerges. This needn't be disingenuous (but it could be). It could simply be that religious movements evolve differently over time when subject to different selection pressures. From a gay rights standpoint, maybe there's a critical threshold where more Muslim immigration causes "don't ask, don't tell" to break down, or this immigration goes from being an advantage (in helping the Democratic Party) to being a disadvantage (in changing the character of the Democratic Party or in Muslims becoming too expensive to maintain as a reliable client group).

    Raymond Ibrahim suggests that the behavior of Muslim populations undergoes qualitative transitions when the numbers cross thresholds of 10%, 35%, and 85%. (Hat tip: David Solway.) Peter Hammond takes a similar view, putting thresholds at 2%, 5%, 10%, 20%, 40%, 60%, and 80%.

  6. Maybe our attitude is "Love the sinner, hate the sin." We want to import Muslims with the expectation that they will eventually assimilate to Progressivism. They or their descendants will either abandon Islam or transform it so that it becomes what Rod Dreher called "Unitarianism with prayer rugs and Middle Eastern kitsch". Every religion, properly understood, is really Progressivism.

  7. Maybe we're telling a "noble lie" about how compatible Islam is with Western values in order to generate political support to save the Syrians from the hardship of living in Turkish refugee camps.

  8. I am overthinking this. People at BAUUC see that high status public figures are calling for more or less open borders (specifically in the context of Syrian refugees), and are motivated by an unconscious desire to affiliate with people who have high status. Everything else is retconning.

  9. Certain people at BAUUC have formed personal relationships with certain individuals at CLIC, and are trying to be nice to their friends. We are next-door neighbors, after all. But I am not seeing a corresponding interest in Christians who are pleading for religious freedom and accomodation. UUs' interest in religious freedom seems awfully selective. Are there no Christians living nearby?

BAUUC members' behavior is perfectly reasonable in terms of explanation #2. My problem here is two-fold: First, while it's entirely plausible that most people at BAUUC think of the different versions of Islam as highly independent, I'm not really convinced that they are all that independent. Second, this explanation isn't consistent with the political rhetoric I hear. Instead of saying that there are good and bad versions of Islam, what "politically correct" people say is that "Islam" is good, or that it is at least no worse than Christianity. (George W. Bush famously insisted that Islam was a "religion of peace", and most UUs are way more politically correct than he is.) Then, when the inevitable examples of terrorism are mentioned, shenanigans ensue.

One kind of shenanigan is the no true Scotsman fallacy. Suitably modified, it goes like this:

Alice: Islam is a religion of peace.
Bob: Osama bin Laden was a Muslim, and he was a terrorist.
Alice: Well, Osama bin Laden wasn't a true Muslim.

No true Scotsman is a special case of cherry-picking one's data. When the "data" consist of anecdotes and news reports, it can be hard to say who is doing the cherry-picking, and accusations fly both ways.

There was an article in The Bay Area Citizen about a recent Ramadan celebration at CLIC with numerous BAUUC participants, which reports,

But any reference to radical violent fundamentalists following an act of terror inevitably presents a distorted picture of Islam to many, said Dr. Mohamed Shalaby, a member of MAS Houston.

My interpretation of this statement is that Dr. Shalaby is accusing Islam's critics of cherry-picking the data, and he is using this as an excuse for doing his own cherry-picking.

Another type of shenanigan is the fallacy of the excluded middle. Pope Francis, unfortunately, provides a wealth of material here. Rod Dreher wrote a recent article entitled, "What Is Wrong With Pope Francis?" I strongly recommend that you read the whole thing, but here is a quote (from Crux):

Pope Francis on Sunday defended his avoidance of the term "Islamic violence" by suggesting the potential for violence lies in every religion, including Catholicism.

"I don't like to talk about Islamic violence, because every day, when I read the newspaper, I see violence," Francis said, when asked about why he never speaks of Islamic terrorism or fundamentalism when condemning attacks such as the murder of a French priest last week, who had his throat slit by an Islamic terrorist as he was celebrating Mass.

The pope said that when he reads the newspaper, he reads about an Italian who kills his fiancé or his mother in law.

"They are baptized Catholics. They are violent Catholics," Francis said, adding that if he speaks of "Islamic violence," then he has to speak of "Catholic violence" too.

As Dreher says, "Again, how very odd for a world religious leader to deny the power of religion to mold the minds of men and to motivate their behavior." But my point here is the fallacy of the excluded middle. It goes like this:

The correlation coefficient between Islam and bad behavior is either zero or one.
The correlation coefficient between Islam and bad behavior is not one.
Therefore, the correlation coefficient between Islam and bad behavior is zero.

The Pope treats Islam as a single religion, even though his unwillingness to criticize this single religion forces him to engage in shenanigans. Now, it's entirely possible that most BAUUC members regard CLIC and various problematic versions of Islam as being different religions that are almost completely unrelated. But that's not how BAUUC members, and people in general, normally talk about Islam. Again, I'm having trouble seeing a bright line distinction between groups of people who venerate the same book, and so I tend to think of Islam instead as a single statistical group with a large behavioral "variance within group".

Part of my motivation here is that I don't trust the US Government to be able to tell the sheep from the goats. The USG couldn't stop the Tsarnaev brothers from blowing up the Boston marathon even with Russian intelligence playing a supporting role as The Clue Fairy. Part of the problem is that American elites are too much in bed with Saudi Arabia. But it appears to me that a bigger problem is that political elites, and people with affectations of being similar to the political elites, are too busy using Islam as an pretext for feeling superior to the Hillbillys. Maybe I'm wrong and the other BAUUC members are right in thinking that the Markovian transition matrix is sparse. But if the US Government (and the "politically correct" in general) are untrustworthy, it doesn't matter.

Proposition 1: There is a rational basis for believing that the US Government is either unable or unwilling to separate the sheep from the goats w.r.t. Islam.

So I'm still interested in whether "Islam" is a "good" religion.

In order to say that something is "good", I need several things:

First, I need criteria for what is "good". The context for this is immigration policy. For me, a good criterion would be, "Is it good from the standpoint of making life more pleasant for my descendents?" Does it tend to make people under its influence good neighbors for people like me?

Second, I need a baseline for comparison. Something is only good or bad relative to some alternative. This is somewhat arbitrary, but I need to be realistic and consistent. There are a lot of different versions of Protestantism, so if I have to choose a specific version of Christianity for comparison, I'm going to take Catholicism as my baseline.

I am also going to need honest statistics. Is thalidomide a good thing? To answer this question, I need to look at all of the possible outcomes, weigh their desirabilities, and consider the probabilities of all of them. A lawyer representing Chemie Grünenthal might want to throw out the data on thalidomide babies because it harms his client's interests, but a judge would have to be breathtakingly stupid or corrupt to allow that. Any reasonable person would weigh the harm and the probability of a thalidomide baby, along with the benefits of taking thalidomide, and conclude that no one should use thalidomide unless she were very certain that she was not pregnant. This is true even if the probability of a thalidomide baby is very low; the benefits of thalidomide aren't that great, and the potential harm is awful. If you want to be honest, you have to sum over all of the possible outcomes, not just the ones that help your case, and not just the most common ones.

A good social scientific answer from Pope Francis would have been something like, "Yes, some percentages of Catholics perform acts X, Y, and Z; and some percentages of Muslims perform these acts. Here are the numbers and here is how I weight them."

I don't have satisfactory data at hand to do an analysis of this sort. Most of what I have to go on are news stories, which is to say, anecdotes. But some of these acts are enumerated at the Religion of Peace website. Here's another anecdote. "Islamic extremism" certainly punches above its weight in this Wikipedia list. (Hat tip: Robert Mariani.)

Okay, let's take a closer look at that Wikipedia terrorist list. I count 406 rows and 276 citations of "Islamic extremism". Let's generously attribute all of the other 130 motivating "ideology" entries to religions other than Islam, treat both Islamic and non-Islamic terrorism as Poisson processes, and try the Przyborowski and Wilenski "C-test" to see if Islam produces the same or less terrorism as other religions. The null hypothesis is that the terrorism event rate for Muslims ("Islamic extremism") is no greater than the terrorism event rate for non-Muslims. I get:

k = 406 number of terrorism events
k1 = 276 Islamic extremism citations
k2 = 130 other citations
world Muslim % = 23.18% Wikipedia, religions by country
n1/n2 = 23.18%/(100%-23.18%) = 0.30174 population ratio
H0: λ12 <= c = 1 null hypothesis (Eq. 1.4)
p(c) = 0.30174*1/(1+0.30174*1) = 0.23180 Eq. 2.1
P = sum (i=k1:k) ((k choose i) * p(c)^i * (1-p(c))^(k-i)) Eq. 2.2
P = 1.3633*10-81 < α = 0.05 probability of a false negative

Therefore: Reject the null hypothesis. It is way more than 95% certain that Islam is more prone to terrorism than other religions.

Proposition 2: There is a rational basis for believing that Islam, or something that correlates with Islam, is problematic.

What the "Religion of Peace" and the Wikipedia terrorism websites describe are spectacular but infrequent events like the murder of Father Haber. But I am more concerned with large numbers of mundane acts like voting for illiberal (in the classical sense) politicians. This is why the numbers are so important in discussing immigration. The probability of being killed by one Chechen immigrant is low. The probability of large numbers of Chechen immigrants helping to produce bad government strikes me as very high. This is the main sort of danger that I think I and my descendents face.

One example of what I think is bad government is Turkey. Turkey's Prime Minister Recep Tayyip Erdogan is famous for saying, "Democracy is like a streetcar. You ride it until you arrive at your destination and then you step off." It's not that I think that democracy is a panacea, but if we are going to make democracy work in the US, this kind of thinking is not helpful. And if we're ready to give up on democracy, I don't want to get off at the same stop as Erdogan. Nor do I want decisions about freedom of religion to be made by politicians selected by people from Saudi Arabia, Iran, or Iraq.

I would like a good proxy for what it means to support bad government. Let's try a Mann-Whitney-Wilcoxon U-test on the data at the Wikipedia page on the Transparency International corruption perceptions index, combined with data from the "religions by country" page I mentioned earlier. Population 1 are the Muslim plurality countries and population 2 are everybody else.

The null hypothesis is that the countries in both populations are equally corrupt. I get:

n1 = 45 number of Muslim majority countries
n2 = 124 number of other countries
W1 = 4887.5 sum of ranks for population 1
W2 = 9477.5 sum of ranks for population 2
avg_rank1 = 108.61 W1 / n1
avg_rank2 = 76.431 W2 / n2
U1 = 3852.5 W1 - n1*(n1+1) / 2
U2 = 1727.5 W2 - n2*(n2+1) / 2
μu = 2790 expected mean, n1*n2 / 2
σu2 = 79050 expected variance, n1*n2*(n1+n2+1) / 12
Z = 3.7790 deviation ratio, (U1 - μu) / sqrt(σu2)
P = 1.5745*10-4 < α = 0.05 probability of a false negative

Therefore: Reject the null hypothesis. It is more than 95% certain that Muslim plurality countries differ from other countries in terms of corruption.

I re-did this analysis as a one-tail test and found that Muslim plurality countries had significantly higher corruption than the others (P = 7.8727*10-5). I did it a third time and found that Christian plurality countries had significantly lower corruption than the others (P = 0.0036173).

Next I tried doing an analysis of variance (ANOVA) on the Aggregate Score data in the 2016 Freedom House Freedom in the World report as a function of the plurality religion from the "religions by country" page. I discarded column 7 in the latter page because there were zero countries with a plurality of "Other Religion".

The null hypothesis is that all of the group means are equal (i.e. all religions are the same in terms of freedom). I get:

k = 7 number of groups
N = 195 total number of observations
n = [128 47 6 3 9 1 1] observations in each group
GT = 11803 grand total
SStotal = 1.7037*105 total sum of squares
SSbetween = 5.2962*104 sum of squares between groups
SSwithin = 1.1741*105 sum of squares within groups
DFbetween = 6 degrees of freedom between groups
DFwithin = 188 degrees of freedom within groups
MSbetween = 8827.0 mean square between groups
MSwithin = 624.51 mean square within groups
F = 14.134 test statistic
P = 2.9177*10-13 < α = 0.05 probability of a false negative

Therefore: Reject the null hypothesis. It is way more than 95% certain that different religions are associated with different levels of freedom.

I re-did this analysis, also discarding groups 6 and 8 ("Folk Religion" and "Jewish") because there was only one country with a plurality in either of these categories (Taiwan and Israel), and got a P value of 3.6637*10-14 .

Proposition 3: The dominant religion in a country is correlated with that country's production of woe (e.g. aggregate score in the Freedom House report).

One thing that struck me, looking over the corruption by country table, was that there are an awful lot of third world countries that I don't want to live in, that are majority Christian. There's clearly more going on here than just religion. Specifically, my idea of good government is pretty much one that respects "the traditional rights of an Englishman". What I think I should be trying to do here is to show statistically that this is an English phenomenon, or maybe English and Dutch. My perception is that, to a lesser extent, it is a European cultural phenomenon. It seems almost tautological to say that England and certain of its former colonies are more English than Saudi Arabia is. But I'm interested in how portable this "Englishness" is. Are immigrants reprogrammable androids, or does the political culture from the old country tend to stick with them and their descendents for many generations?

David Friedman reports in ch. 44 of The Machinery of Freedom, 2nd ed., that medieval Iceland didn't have a king...until there was an influx of immigrants from Norway, which did have a king. The immigrants didn't assimilate to Icelandic political ways. The native Icelanders were forced to accept the king of Norway.

Nick Rowe reports that, according to Arthurian legend, England is English rather than Cornish because King Vortigern "invited some Saxons to settle in Britain, and they invited other Saxons to join them."

Another story of assimilation gone bad is the Norman invasion of Ireland. In 1167, King Diarmait was losing a war, internal to Ireland, and asked King Henry II of England for some help. Nearly a millenium later, the British are still in Northern Ireland, and people there still hate each other along lines that are nominally religious, but really have more to do with ancestry.

Here in the US, we have first-hand experience with Tammany Hall and The Curley Effect. And then, of course, there are the Balkans....

Proposition 4: There is a rational basis for skepticism towards the importation of woe from the third world (i.e. loose immigration policy).

So what should BAUUC members' attitude be towards (1) foreign Muslims who want to immigrate, (2) American Muslims in general and (3) CLIC?

I would like to be able to use Bayesian analysis on prospective individual immigrants, but I don't think we normally have enough information on individual people to do it properly. The best I think we can do is to generally only admit exceptionally gifted people. The prospective immigrant's cultural background also has to be a consideration, and religion is definitely part of what I mean by "culture". I don't want problem people, or likely parents of problem people, to immigrate. I don't care whether Islam causes problems or is merely correlated with problems. Either way, I don't want problem people to immigrate. Donald Trump's call for a ban on Muslim immigration thus strikes me as perfectly reasonable.

As an aside, American Protestants used to view Catholics with considerable trepidation. The Thirty Years War was much on the Framers' minds when the First Amendment was passed. Religious war between Protestants and Catholics was a very real concern. Why is it not a concern now? Is it because fewer people take Christianity all that seriously any more, and we have "Churchianity" instead? People trust SCOTUS to defend the First Amendment despite radically changing voter demographics? Do people think the differences between Protestants and Catholics are small compared to those between Christians and Humanists? Has Christianity evolved in ways that are not reflected in changes in the Bible?

Once a Muslim has US citizenship, we're stuck with the problem of figuring out to live together. This brings me back to my list of possible meanings of BAUUC's support for CLIC. The two possibilities that interest me here are #2 and #6.

Possibility 2 is the idea of deliberately promoting good versions of Islam at the expense of the bad ones. Possibility 6 is "Love the sinner, hate the sin", which suggests that CLIC is a sort of halfway house for Muslims who are hopefully transitioning to some other, more felicitous religion, like Christianity or Progressivism.

But how do you promote a good version over a bad version of Islam if you can't identify bad versions or can't acknowledge that they exist? And how do you "hate the sin" if you can't identify or can't acknowledge what the sin is?

I understand the motive for a politician like G. W. Bush to tell a diplomatic lie and claim that Islam is a "religion of peace". I don't want to declare war on 23.18% of the world's population, either. But this is another false dilemma: We either need to go to war against everyone who is even slightly associated with a problem, or else we need to deny fundamental facts about the nature of the problem (i.e. that religion is part of the problem). It makes sense to say that there are degrees of association with a problem. Lying about it may be diplomatically convenient, but in a democracy like ours, there are too many people who "don't get the joke", or who want to use the lie as an excuse for slandering their political rivals. We need to be serious about what the problem is.

I realize that my amateur statistical puttering about doesn't prove causation. But we know that Islam correlates with trouble. Even if the best we can do is little more than speculation, we need some working hypotheses about the nature of the trouble with Islam so that we can figure out how to deal with it.

There are other possible explanations for Islam's bad reputation besides the religion itself. Elsewhere, I suggested that cultural context and political environment were important in what we call "religion". Richard Fernandez once commented that "Terrorism is an externality of rotting societies." In the case of Saudi Arabia and Iraq, part of the dysfunction may be "the resource curse" associated with oil. Middle Eastern and South Asian cultures tend to be low-trust and tribal.

Religion and political power are hard to separate because one of the functions of religion is to determine who has the authority to adjudicate and make changes to social norms (i.e. morality). That's why Mormons care about apostolic succession. Neither religion nor government can afford to be indifferent towards moral concerns. Religion is a path to political power and legitimacy, and no government can afford to be indifferent towards its own legitimacy or to the problems caused by perverse social norms. But if religion is too subservient to political power, it is discredited.

One could also suggest that genetics is a factor. For example, the problem in Rotherham is with Pakistanis, not Indians. Pakistan used to be a part of India before splitting off, largely along religious lines, but in so far as conversion from Hindu religion to Islam was voluntary, Islam was more attractive to low caste Indians than high caste. Maybe the separation was partially along former caste lines, and caste membership was hereditary? According to two studies of Y-DNA haplogroup R (here and here), apparently not. Pakistanis seem to have more Y-DNA in common with high-caste than low-caste Indians. Inbreeding among British Pakistanis is a problem, and that's cultural rather than religious, but it doesn't have much to do with the Rotherham scandal.

We can see some of the interplay between religion, politics, and violence by looking at the 30 Years War and the English Civil War, as well as internal religious strife in France (e.g. Huguenots). Plenty of killing was done in the name of Christianity, and at least some of it was sincere. When people stopped killing in the name of Christianity, it wasn't because of changes in genetics. History seems pretty clear that religion does in fact exert power "to mold the minds of men and to motivate their behavior". The burden of proof is on anyone who would deny this. The statistics also strongly suggest that Islam does so in ways that are less desirable than modern Christianity.

Part of the problem is clearly governments that use Islam in order to encourage soldiers to fight wars, including proxy wars. Part of this is in order to distract people from the governments' own failings. But why don't we see more of this being done in the name of Jesus or Buddha?

My problem here is to figure out what's different about Islam and Christianity, and how they've changed over time. What can I attribute to the cloth, and what can I attribute to the tailor? What causes Islam to be correlated with terrorism, corruption, and coercion? What is the "sin" that we ought to hate?

This is a difficult question in part because there are obvious similarities between Islam, which I want to criticize, and Christianity, which I reluctantly praise. Both religions have traditionally claimed to be the One True Path to salvation, where you only get one lifetime to get it right, and failure is infinitely horrible. The importance of this justifies horrifically desperate measures to try to convert infidels. Except that this is not so true of Christianity any more. Steven Weinberg writes,

Even though American atheists might have trouble winning elections, Americans are fairly tolerant of us unbelievers. My many good friends in Texas who are professed Christians do not even try to convert me. This might be taken as evidence that they don't really mind if I spend eternity in Hell, but I prefer to think (and Baptists and Presbyterians have admitted it to me) that they are not all that certain about Hell and Heaven.

Weinberg interprets this as "a retreat that can be attributed to no new revelation, but only to a loss of certitude." My interpretation is that Christians still believe in God, but not Hell. Purgatory, maybe, but not Hell. The Universalists have won ("Universalist" in the sense of believing in universal salvation). In other words, Christians no longer typically believe that Christianity, let alone any particular version of it, is the One True Path. Modern Christians who do believe in One True Path typically also believe, like Roger Williams, that religious coercion is a sin, and that forced conversions are invalid.

The exclusive "One True Path" claim is basically incompatible with multiculturalism and peaceful globalization. To combine the two requires a bit of polite hypocrisy. Modern Christianity gets around this in several ways. The formal Christian Bible still claims One True Path, but the informal "universalist" modern doctrine has pretty much abandoned this. The Christian Bible also lends itself to being interpreted as opposing religious coercion (e.g. there is little precedent for it in the life of Jesus). Christianity explicitly holds itself somewhat aloof from the State ("render unto Caesar"), which is the principal means of legitimate coercion in modern societies. So while Christianity often requires a bit of polite hypocrisy in order to get along with other religions, the amount of hypocrisy required is pretty minimal. The amount of hypocrisy required in order for Islam to play well with others is much greater.

Another issue is apostasy. Any religion that tries to promote credence goods, (eternal salvation, moral superiority, etc.) by creating an illusion of consensus among the knowledgeable is going to have issues with apostates. Apostates spoil the illusion of consensus among the knowledgeable. Many Islamic governments (e.g. Afghanistan) threaten apostates with capital punishment. This is clearly a deal-breaker if you're trying to get along in a global, multicultural civilization. Timothy Shah claims that the Islamic basis for capital punishment for apostasy isn't in the Koran, but in certain hadiths. (Robert Spencer finds support for killing apostates in the Koran.) Also, he points to capital punishment for apostasy in Christian history. If you're going to try to promote "good" Islam over "bad" Islam, you definitely need to make sure the version you support emphatically rejects the hadiths in question. But if you think of "Christianity" and "Islam" as religions rather than families of religions, then you have to approach the problem differently. Some proportion of Muslims are going to embrace the problematic hadiths, and interpret them literally, rather than looking for excuses to sweep them under the rug. As I argued earlier, the difference between Christianity and Islam is that the different respective "cloth" provides different "tailors" with different probabilities of winning these sorts of arguments. For a Christian, it's possible to argue for killing apostates, but it's a stretch. On the other hand, for any version of Islam that is even slightly respectful of the hadiths in question, the stretch is more likely to be in arguing against killing apostates.

Christianity also seems quite a bit different from Islam in the degree to which it lends itself to separation of church and state. As far as I know, there is nothing in Islam comparable to Jesus' "render unto Caesar" statement. I am tempted to claim that the lack of such a clause in Islam is also a deal-breaker. But it is possible to read too much into this. As I said above, neither religion nor government can afford to be indifferent towards moral concerns, and no government can afford to be indifferent towards its own legitimacy. The attempt to forbid the state from establishing a state religion tends to result in word games about what is a "religion" and empowers demagogues to invent new religious and quasi-religious movements in order to sieze power.

Nevertheless, the "render unto Caesar" clause in Christianity supports a degree of distinction between public and private morality, between what people ought to do voluntarily and what they are justified in forcing others to do involuntarily. This distinction appears to me to be much weaker in Islam than it does in Christianity.

Another issue is the treatment of infidels. Nietszche accused Christianity of fostering "slave morality". The way I understand it, early versions of Christianity were oriented towards avoiding running afoul of the Roman army. Later versions were oriented towards being in the Roman army, but there was always a tendency towards sympathizing with the slave. Islam, on the other hand, seems to have started out as a retrospective justification for conquest, brigandry, and slavery. The Religion of Peace website has a disturbing page on Muhammad's teachings regarding the rape of female prisoners. Some of these teachings are from hadiths that might be relatively easy for Muslims to discount, but some of them are from the Koran.

Islam has evolved and branched out to some extent. People who were raised as Muslims or forcibly converted have often tried to take the lemons that were forced on them and make lemonade (as have many Christians, et al). To return to the sword and plowshare metaphor I borrowed earlier, Islam comes from the mill in a convenient shape for making a sword, and it takes quite a lot of hammering in order to turn it into a plowshare. In addition to "coming from the mill" in a much more sword-like state, Islam is also harder to work than Christianity. It is easier for Christians to cherry-pick from the conflicting views presented in the Bible than it is to cherry-pick from the less internally conflicted Koran. In addition, Islam is like Protestantism in making a fetish out of its holy book. Catholicism also has issues, but it is less focused on the Bible.

I'm left with the impression that, if we treat Islam as a single religion, and want to reform it, we need to call something like the Council of Nicea, that has the authority to scrap large parts of the Koran and Hadiths, write a new book like the Book of Mormon, and to rebuke anyone who makes a fetish over the result.

Another issue is church polity. Islam is said to have been "born Protestant". That is to say, there is no Pope or other central authority who can speak with much authority across national boundaries. Various national governments support and control local versions of Islam, which makes Islam seem a bit like Orthodox Christianity, but their control seems weak. The Saudi government can't stop the Wahhabis whom it supports from occasionally trying to attack the Saudi family for being insufficiently pure.

David Patel reports that this generalization is more true of Sunni than Shiite Islam, and that the difference has consequences in places like Iraq. Patel distinguishes between solving a public good problem and solving a coordination problem. The more hierarchical structure of Shiite Islam gave it an advantage in both areas, but especially with solving coordination problems.

The good news about Islam being "Protestant" is that it can "fork" easily, and thus has the potential for rapid evolution. The bad news is that, outside of local governments dictating religious doctrine to the imams, this evolution is out of control. There is no Islamic Pope who can ban polygamy and regulate consanguinity the way the Catholic Church did. There is open entry into the career of "public sanctimony entrepreneur". There is no way to suppress insane holier-than-thou competition ("supererogation"), and in so far as religious ideas are a public good, they are likely to be produced only as a side effect of also being private goods. In Mancur Olson's terms, Muslims as a whole (like Protestants) are a "latent" group, unlikely to be able to solve coordination problems easily. A related observation was made by Richard Fernandez in his Three Conjectures essay, that without a Pope or a dominant national government, it is impossible to negotiate a peace treaty with Islam as a whole.

Regarding (1) foreign Muslims who want to immigrate, I've already come down on the side of the immigration skeptics. That leaves me with the questions of what attitudes to take towards (2) American Muslims in general and (3) CLIC.

I spent a lot of pixels wringing my hands about what is the right unit of analysis in understanding Islam. My tentative answer is that the correct unit of analysis is the book. Anyone who venerates the Koran is essentially the same religion, i.e. belongs in the same statistical group. See my Markov chain model. If you give your child a book, you have limited influence over how your child will interpret it. Maybe you get a vote, and the political environment gets a vote, but the book also gets a vote. As long as they venerate the Koran, I reject the view that the different versions of Islam are fundamentally distinct.

There are two reasons for taking a negative view of people who venerate the Koran. One reason is that, for various reasons (population genetics, political environment, culture), venerating the Koran puts you in with a bad crowd. You can't do anything about your genes, but you can stop affiliating with evil political leaders and dysfunctional culture.

But I also think that the book is bad. This opinion is based partly on the statistics I presented above on how people do, in fact, frequently interpret the Koran. But it is also partly based on reports by people who have studied the book itself. This is admittedly second-hand, but it seems to me that the things said by the Koran's politically correct defenders (e.g. Robert Wright, The Evolution of God) are damning enough; like the Bible, only more so.

Voltaire famously described Christianity as being full of "villainy and nonsense", much of which is traceable directly to the Bible. The Bible is a mixture of various genres, including some genuine wisdom literature, but consisting in large part of contradictory propaganda written in support of various power grabs and other political maneuvers by priests and kings, including wars of aggression. Some of it is polytheistic, some "monolatrous", and some monotheistic. The poor reader who tries to interpret it for himself as a guide to moral behavior is told that "To every thing there is a season, and a time for every purpose under heaven." The wisdom of the Christian wisdom tradition seems to me to consist of the work of professional clergy who have spent two millenia figuring out which parts to quote under what circumstances, and which parts to sweep under the rug.

The Koran is similarly full of contradictory propaganda to support opportunistic political maneuvers, largely divided between (A) preaching at Mecca while out of power and (B) engaging in military conquest at Medina while in power. The "Satanic verses" involved failed attempts to schmooze polytheists, which contradicted other parts of the Koran and were quickly repudiated. If there is a moral lesson here, it seems to be incorrigible violent opportunism. Again, anyone seeking moral guidance is advised to find professional clergy, and a good clergyman's job consists largely of sweeping things under the rug.

That's my answer regarding American Muslims in general: "Love the sinner, hate the sin", where the sin is taking the Koran seriously. It is analogous to saying, "Love the alcoholic, hate the alcohol," except that with Islam, I'm much more concerned about peer effects than I am with alcohol.

Proposition 5: There is a rational basis for thinking that the Koran is generally a bad influence on people.

The next question is whether I should take a different attitude towards CLIC than I do towards American Muslims in general. My answer is no. As far as I can tell, CLIC is not a halfway house for recovering Muslims, it is genuinely Islamic (i.e. they appear to take the Koran seriously). It also seems to me that UUs who are exposed to CLIC are as likely to convert to Islam as CLIC members who are exposed to UU are likely to convert to UU. If UUs think that every religion, properly understood, is really Progressivism, then CLIC members are equally justified in thinking that every religion, properly understood, is really Islam. (Stephen Prothero would beg to differ.)

Part of my skepticism is that there aren't enough degrees of separation between CLIC and more radical versions of Islam. On one hand, Imam Basyouni has gotten death threats from ISIS. On the other hand, CLIC is on friendly terms with CAIR, and CAIR is tied to Hamas and the Muslim Brotherhood. There's too much history of justifying lying as a legitimate ruse of war (taqiyya and kitman). For all I know, someone could be playing what James A. Donald described as a game of good cop/bad cop:

Hey, I am a moderate Muslim. Give me everything I demand, or you might have to deal with my pal the immoderate Muslim.

Alternate version:

We will be your friends until we are strong enough to be your enemies.

I also don't believe that BAUUC members in general are even trying to proselytize to CLIC members or to influence how their religion evolves. If they are, they seem to be going about it in a perverse way. UUs generally (e.g. overwhelmingly on Facebook) treat Islam as having a privileged status, of being above criticism, and accusing anyone who does criticize it of bigotry. UUs certainly don't treat Christianity this way. Giving a group this sort of privileged status, on condition that they don't change their religious views, doesn't look to me like an attempt to get people to change their views.

Brendan O'Neill writes,

This censorious flattery of Islam is, in my view, a key contributor to the violence we have seen in recent years. Because when you constantly tell people that any mockery of their religion is tantamount to a crime, is vile and racist and unacceptable, you actively invite them, encourage them in fact, to become intolerant. You license their intolerance.

Loving the sinner and hating the sin means trying to separate the two. If I really believe that Islam is bad for people, I should be trying to convert Muslims to some other, less harmful, religion, just as I would encourage an alcoholic to stop drinking. As Tommy Robinson of PEGIDA put it,

I'm opposed to Islam as a fascist ideology. We are not an anti-Muslim group. We feel Muslims are victims of Islam.

Let's go back to possibility #3, "We have enough information about the individual people at CLIC to be able to treat them as individuals." That's the situation I'd like to be in. As Martin Luther King, Jr. put it, I want to judge a man by his character, not by the color of his skin, or in this case, by his nominal religion. The question then becomes, do I have enough information to be able to do this?

My answer is that, in order to evaluate what I think is important in what we call a man's "character", I need to know a great deal about where his loyalties lie. I generally won't have this kind of information about someone unless I have known him for a long time, and have seen how he behaves under stress, in many different circumstances. It may well be that his character is full of loyalty to his ethnic or religious community rather than to my "proposition nation". For my purposes, this is likely to make him a bad neighbor, and if his culture persists over several generations, is likely to make his grandchildren bad neighbors to my grandchildren.

Garett Jones calls this loyalty to kith and kin over abstract principles "amoral familism". He also observes that this has long term political consequences.

We wanted workers, we got voters instead.

"Bad character" could simply mean incompatibility combined with refusal to assimilate. This incompatibility could be as innocent as the British driving on the left. Rudyard Kipling described this sort of friction-without-blame in his poem, The Stranger:

The Stranger within my gates,
  He may be evil or good,
But I cannot tell what powers control—
  What reasons sway his mood;
Nor when the Gods of his far-off land
  Shall repossess his blood.

The men of my own stock,
  Bitter bad they may be,
But, at least, they hear the things I hear,
  And see the things I see;
And whatever I think of them and their likes
  They think of the likes of me.

August, 2017: In The Righteous Mind, Jonathan Haidt mentioned Emile Durkheim's idea of man as being "homo duplex": a man is simultaneously an individual and a member of various groups. It's like having to think of an electron as acting like a particle in some circumstances, and like a wave in others. I wanted to end this essay by talking about "homo duplex" and the transition between treating someone as a member of an untrusted statistical population and treating him as (1) a trustworthy individual or (2) a member of a generally trustworthy statistical population (ie. "Who are my people?"). But it's been something like a year since I started writing this. Apart from being busy, I've been having writer's block, and I think a lot of the writer's block has to do with Sam Keen's question, "Who are my people?"

I think the other issue is a red herring. I never have enough information about someone to be able to predict his or especially his grandchildren's behavior reliably. All I can do is lump people into categories that are loosely sorted by index of suspicion.

The real problem is that I don't know who my people are. Whom do I trust with the right to vote over matters relating to my life, liberty, and property? Anglo-saxons and Dutch? Classical liberals (moderate libertarians)? Conservative Christians? Some small subset of reactionaries? I keep coming back to Spandrell's point that we need a new religion, whose outline is only very dimly visible. That's where I am.

You kids! Stay out of my yard! I don't trust any of you.

Addendum, 9-16-2017: Dear God, teach me to utter kind and tender words, because tomorrow I may have to eat them (the Houellebecq option).

Universalism and Particularism
(Immigration and Religion, pt. 3)
Controlled Flight Into Terrain