I had lunch the other day with some friends of mine who were spouting off, in a Rush Limbaugh like manner, about 'Christmas' being replaced by 'Holiday'. As the conversation moved on one guy was saying how he met with a born again Christian friend. He told me it was tough because 'I'm not a Christian and had a hard time coming up with safe jokes.'

Perhaps it's easier to be an agnostic/atheist 'Christian' if you're in the majority. After all 'Christianity' is the background noise you become accustomed to. And if by accepting certain moral precepts and observing some nominally religious holidays where good old fashioneed pagan rituals like eating and drinking to excess swamp out the tedium of being dragged to Church Christmas morning is all that is required to being a 'Christian' then I'm a 'Christian.'

"Bertram lives in England, where brown ("Asian") and Muslim intersect pretty well (though many Asians are not Muslim, almost all Muslims are Asian)."

There are many black African Muslims in Britain, IIRC Somalis are one of the fastest-growing immigrant groups.

My goodness, you really do need to brush up those reading skills.

In my post I expressed no opinion about whether Muslims should be afforded special protections. Nor did I affirm in that post that Muslims anywhere constitute an ethnic group. What I did suggest was that it is quite possible for religious affiliation sometimes to function as a marker setting the boundaries of ethnic groups and that therefore Muslims might turn out to be an ethnic group at some times and in some societies. That may or may not be stupid but it certainly doesn't license you to attribute to me a swathe of opinions I nowhere express.

yes, but see here. the conditional probabilities of a black african being muslim are far lower than an asian (~20% vs ~50%). and that doesn't take into account that many carribeans on visual inspection might be hard to distinguish from black africans, and very few of them are muslim (richard reid notwithstanding).

as an aside, in the usa, the vast majority of brown people are from india (with relatively few from pakistan and bangladesh, unlike the UK). the vast majority of these are non-muslim, the result being that the odds are if you see a south asian in the USA they are from a hindu or sikh background. ergo, people usually assume i am hindu or sikh until proven otherwise (though who assume i'm muslim are also the types to assume hindus or sikhs are muslims! so they don't assume i'm muslim because they know anything about the demographics of american islam, they assume i'm muslim because all brown people are muslim in their mind, and i'm part of all brown ppl).

As it happens I agree with Bertram. If religion creates barriers to integration ( i.e apostasy is punishable by death or by social exclusion from the InGroup, or by eternal damnation) and if out-marriage is apostasy then the religion creates an InGroup which intermarries with itself. Of course, not being of the left I wont go on to blame Whitey or Western society ( or even the blander " we as a society") for that: but credit instead the in-group with the barriers.

Even after the religious belief is gone the cultural barriers to out marriage remain - which is why you do have catholic atheists, and protestant atheists in Northern Ireland, and why such rhethoric makes sense.

There was a discussion on Crooked Timber about whether Class was a ethnicity - the answer is that is can be ( or at least an InGroup).

Being lefties they asked was the "bourgeoisie" an ethnic group: the answer to that is no, as the capitalist classes - the self employed - come from all classes, and therefore any family will have the bourgeoise and proles in it's extended mix: however the white collar/ blue collar barrier is a thick black impregnable line.

Traditionally, an ethnic group is defined as a group united by common language and customs. A group united by religious beliefs does not normally meet these criteria.

On the other hand, religion sometimes has the same social consequences as ethnicity, notably with regard to marriage and residence choices.

And then there are cases, like Jews, and perhaps Sikhs, where it is difficult to distinguish religion and ethnicity. As a matter of interest, both Jews and Sikhs have been accepted as ethnic groups by Courts in Britain for the purposes of race relations legislation.

Ultimately, definitions are a matter of convenience rather than objective right and wrong. Personally, I think it is more useful to keep a distinction between ethnicity (based on language and custom) and religion (based on beliefs). It may sometimes happen that a religious group is also an ethnic group, but this can only be decided case-by-case.

Chris Bertram's comments in discussion of his own post on Crooked Timber include the following:

"But given the importance of community and ritual behaviour in Islam as compared to doctrine in contrast with Christianity (but in this similar to Judaism), I’d have thought that Muslims in the West were highly likely to function as ethnic groups."

Compare and contrast this with his claim above that "Nor did I affirm in that post that Muslims anywhere constitute an ethnic group."

Maybe Chris is distinguishing between his post and his subsequent comments, or maybe he is distinguishing between 'affirming' that something is the case and saying that it is 'highly likely'. Both would be legitimate distinctions, but rather subtle ones, and a poor basis for accusing someone of poor reading skills for failing to notice them.

which is why you do have catholic atheists, and protestant atheists in Northern Ireland

northern ireland is the ideal case of this. that is the problem, most situations don't necessarily follow it. intermarriage barriers created by religion vary. in the USA, ~40% of "jews" outmarry (30% of americans of all religions outmarry).* the presumed barriers between ingroup and outgroup seem not to be barriers to discourse, shall we discount ingroup-outgroup status? as i point out, muslims are multiracial and multilingual in origin. they eat different foods, and don't necessarily socialize across "ethnic groups" (for the purposes of argument, assume prima facie that somali muslims and pakistani muslims do view themselves as fundamentally distinct). the contrast with jews and sikhs is illustrative, the vast majority of british jews are ashkenazi and the vast majority of sikhs are punjabi. the coupling between ethnic origin and religious affiliation is very tight, and in fact, religious profession can be irrelevant (disraeli is a jew despite his anglican convictions). this suggests that the religious label is simply a correlated proxy. the problem mapping this sort of system with muslims is that islam is a transnational religion. in nations where all muslims are of ethnic group x it can serve as a proxy, but in the USA, or to a lesser extenet even britain (due to black muslims amongst south asians as well as arabs) it is less tight a coupling.

the rub of it is of course that this is relevant because the gov. gives recognition and rights to groups, and so taxonomies become important. the group 'jew' or 'sikh' is coherent enough that it is workable, because legal systems presume a deterministic and precise boundary to these sort of classifications (ergo, blood quanta being important in racial hygiene laws in nazi germany). but with something like a 'muslim' ethnic group it exists nowhere except in the minds of the outgroups, 'that amorphous muslim mass.' hispanics to some extent are similar in the USA, cubans and mexican americans might have little affinity besides their spanish language and roman catholic religion. but here you have at least two points of commonality, as well as the geographic perception of coming from the 'south.' in contrast, muslims can be all colors, speak many different languages, and originate from 3 continents. not that an incipient deracinated muslim-only identity is not emerging (i.e., pakistani american children abandoning punjabi language and dress for a more arabicized set of markers), but, this is not an omnipresent or overwhelming phenomenon. frankly, the leftist tendency** to clump all non-whites, or 'group X,' together, is using the same logical process as racists (this is justified via utilitarian means because racial justice demands group clustering, but ultimately i wish more would ask if the means justifies the ends, or subverts the ends), except they 'mean well.' ultimately, i wish that they would get past their equivalence between different sets of values and accept individuals and dissenters from neo-medieval interpretations of islam as a positive development and attmept to enable them instead of pretend that they are abberations who are simply interference in the 'big picture' of accomodation of non-western cultures within the western matrix. as it is, they trivialize and dismiss people like ayaan hirsi ali because she does not toe the line that they expect from people of group X (and accuse her of 'islomophobia'). in my book, the quality of life of people like ayaan hirsi ali is worth 1,000 X black-moving-objects. but hey, perhaps i'm a bigot.

* america is differrent than northern ireland, which goes to show that these definitions are locally contextual, so grand theories of ethnicity are pretty unreliable.

** obviously not all people on the political left are like this, and perhaps not even a majority. yet the fact that anyone who considers themselve a progressive would dismiss someone like ayaan hirsi ali, whose trangressions are against reactionary medievalism are heroic in both personal narrative and public activism, makes me want to vomit. surely many leftists who accuse her of making common cause with right-wing extremists are right on some counts (from what i have seen), but the reason she does these things is that you take the lifeline you are offered if you are drowning.

You probably could make a case that the (catholic) Irish are an ethnicity defined almost entirely by religion, being the only sizeable Roman Catholic population left in the British Isles after the reformation, and themselves historically denying the protestant Irish the right to be considered truly Irish. Conor Cruise O'Brien has written interestingly on this last point.

being the only sizeable Roman Catholic population left in the British Isles after the reformation

i think the world 'sizeable' is imprecise. i mean, there was an indigenous catholic population of english origin even before the emigration of irish to england. alexander pope was from a catholic family for example. and i believe many of the highland scots remained catholic.

I'd defend "sizeable": there were of course catholic English, especially in the North, but they would have been a very small part of the population by, say, 1700. There were catholic Highlanders, but more Episcopalians (i.e. protestant dissenters from the Presbyterian Kirk); the Highlands had, by the Reformation, regressed to being substantially irreligous/pagan, which is probably part of the reason that the Scottish reformation was pretty unbloody. There was some successful counter-reformation conversion particularly in the Outer Isles, but the bulk of the Highlands is now presbyterian, having been converted from irreligous/pagan as late as the 19th century. Or so I understand: I am just an amateur in these matters.

I've tried to come up with definitions of "racial group" and "ethnic group" that match how the U.S. Census Bureau uses those terms in practice (e.g., they make a big deal about how "Hispanic" is an ethnic group, not a race).

You are probably familiar with my definition of a race ("a partly inbred extended family"), but my definition of ethnic group might be relevant here: "A group united by traits that are frequently passed down within biological families, although they don't have to be."

Thus, "Muslim" could be an ethnic group in some situations. It would not, however, be an ethnic group in Saudi Arabia, where everybody is Muslim.

I would not expect Muslims (as Muslims) to be recognized as a racial/ethnic group in the United States anytime soon. Racial and ethnic identity in this country is relevant mainly in terms of determining whether one qualifies for preferential treatment under various affirmative action programs. It is unnecessary to categorize Muslims as their own race/ethnicity for affirmative action purposes simply because the overwhelming majority of Muslims *already* fall within various nonwhite classifications and hence already qualify for many, though not all, affirmative action programs. Granted, white converts like John Walker Lindh and Cat Stevens sort of fall through the cracks, as do Bosnians and Albanians, but they're not quite numerous enough to require special attention.

Rob B. Since the remarks made in the comments here consisted in asserting that I was stupid, was discussing something I didn't know about and was making a particular empirical claim, I think I was well within my rights to react as I did.

My initial post at CT was not intended to make an empirical claim of any kind. Rather, it was intended to rebut the claim that the question of whether religious groups might also be ethnic ones could be settled just on the basis of the meaning of the terms. I was saying that this was not a conceptual question, but an empirical one.

Now for the question raised by my "highly likely". Is it the case that, as a matter of fact, there are some circumstances where Muslims function as an ethnic group? (Only those who accept that this is an empirical question need follow this far.) Here I would think that, as anxiety about Muslims grows in Western Europe, it becomes more likely that being a Muslim will function as a marker of separate identity, will govern access to important societal goods etc, and that those who are (or who are not) Muslims will think of that as distinction between "us" and "them".

In The Netherlands, the Muslims are mainly split among Turks, Moroccans and Indonesians. It seems to be a perfect case study. These still are three distinct ethnic groups. Is there any evidence that these groups are more likely to intermarry among themselves than the Dutch population in general?

If so, they may become a new 'ethnic' group. What you call them is up in the air. If these three groups represent 90%+ of Dutch Muslims, then calling them a 'Muslim' ethnic group may make sense. If there is intermarriage between Turks and Moroccans but not between either of those groups and Indonesians, then calling the former combined ethnos 'Muslim' wouldn't make sense because it excludes the Indonesian Muslims. You'd have to come up with a new name.

Are European Americans now their own ethnic group since there isn't any real barriers between the once distinct ethnic groups marrying?

I think it's clear that Bertramm uses the term "ethnic group" in a way that can only cause confusion. The term is normally an anthrological and statistical term. Bertram expropriates it to promote a political agenda.

When an observer experiences the buzzing confusion that is reality he or she will impose an abstraction or grouping on the phenomena. In statistics one tool that helps with this is factor analysis. The methodological mill produces constructs that glom together.

Antropologists in the field do the same thing - they look for patterns that clump together and they give a name to this construct - ethnic group.

A simple model of an ethnic group would be three orthogonal dimensions - race, language, culture.

In times past the clustering was certainly better than it is today. The cloud of data points was more compact. That meant that knowledge of an individual's position on one or more of the axies gave you a good idea of their position on the other axis.

These multidimentional correlations do not persevere automatically. The cloud of data points cohere or dissipate as a function of circumstance and social evolution.

For example the connection between race and religion has been a function of the religion's means of expansion. Jews don't proselytize and until recently have not intermarried much. Therefore to be a Jew means something genetically as well as in terms of doctrine. For Christians this connection is much less strong because Christianity spread and continues to spread by missionary action. Islam is an extreme case. This is the only major religion that has spread principally by conquest and coerion. A consequence is that the link between belief and the race of the believer hardly exists.

Hence to use the uni-dimensional category of religion as a proxy for the mullti-dimensional concept of ethnicity is least appropriate for Islam. It maybe makes a little sense for Jews but none whatever for Muslims.

Of course Bertram means something else again. His sophistry expropriates the antropoligical notion of ethnic group to advance a political point of view. Sailer hit on this point above. Bertram means ethnic group in the leftist sense of of oppressed group.

A few years ago in England there was a movement to widen the term "BlacK" to include all oppressed people. This silly idea meant that in England Jews would be classified as Blacks.

The concept of a Catholic atheist isn't so stupid. I think there is a catholic cultural style which is seperate from whether you believe in God. I don't believe in God but I got 4 kids.

Joe,

I don't agree with your ascertion, as I find, as an avowed Atheist, I have more in common with fellow Atheists everywhere, than with non-Atheists anywhere.

I think you may be conflating Catholic with a particular cultural context, which is outwith its religious context.

I once asked a coworker about his religion, he said (I'm pretty sure this is an exact quote) "well, I'm a Catholic only I don't believe in this God shit". It may have been "but" rather than "only".

He was German though, and they do things a litle differently over there.

Being a Catholic or a Unitarian or a Muslim, etc. means a person adheres to a variety of beliefs and cultural practices. It might seem odd that belief in god and the particular tenants of the religion is only part of belonging to the religion, but from my experience, many Church going people aren't necessarily religious. They may tacitly accept the dogma, but it isn't very central to their life. I accept the existence of quarks, but I'm not exactly moved to action by that belief.

If you're an essentialist, that is, belief is God is the essential part of being a Christian or Muslim, then it would be absurd to say 'Christian Atheist.' But if being a Christian means adhering more or less to a particular set of traits, I think it's possible to be a Christian Atheist.

I remember hearing a quote from a bishop in the Church of England that went something like this: I don't believe in the immaculate conception or the divinity of Jesus. I don't even believe in God, but I'm still a Christian.

I don't think a person's professed beliefs are as important as their actions. Someone could say, 'yes, I believe that Jesus Christ is my Lord and Savior, blah blah blah' and then go home and beat his wife and kids.

That being said, Razib is right that only if Muslims come from one ethnic group can then Muslims be thought of as one ethnic group within the country they live. Chris is right to point out that there is a potential for Muslims within the host society to become an ethnic group. I don't think that's likely to happen.

Bertram means ethnic group in the leftist sense of of oppressed group.

My goodness, I learn something new about myself every day!

For Christians this connection is much less strong because Christianity spread and continues to spread by missionary action. Islam is an extreme case. This is the only major religion that has spread principally by conquest and coerion.

that's not true. see the saga of olaf trygvasson and see how eagerly the pagan masses embraced christianity. don't make up history pat, it irriates me. without state-elite backing missionaries tended to minimal impact.

chris,

let me make this short. yes, i stipulate that religion may be a necessary and sufficient condition for ethnic demarcation. but its utility is strongly framed by a priori assumptions and preferred policy prescriptions. for you, using islam as a marker is common sense perhaps, you seem positively predisposed toward gov. treatments of group x according to its relation to the general society, so you need a heuristic to separate groups. but conservatives (or the ppl at butterflies & wheels) are not being disingenuious when they assert that islam isn't an ethnic marker because they own public policy prescriptions make the necessacity of markers such as this weak, so they do not weight the operational argument as more worth or valid than the categorical/ontological one.

i still think that there are empirical arguments we can have. but basically your point of initiation strong shapes one's perspective on what is, and isn't, reasonable.

Chicago is an interesting laboratory for this question. There has for some time been a non-trivial rate of Mexican-Polish Catholic intermarriage and it shows up in two different neighbors. Stanislaus Gonzalves and Rosita Lupinski are typical actual names which come to mind.

"...without state-elite backing missionaries tended to minimal impact."
Razib's right. Many Europeans converted to Christianity at the tip of a sword or the order of state, though I do remember reading that Ireland was the exception to this. The conversion really was voluntary in their case. Elsewhere, they didn't waste time in those days, life tending to be so brutish and short.

Who was the great missionary to the Goths? Did he do it by the sword? Granted, when the chief converts by persuasion, his underlings may do so under a sense of compulsion. Even so, this is quite different in Islam where everybody was under compulsion from the top down. Indeed, apart from a few of Muhammad's early followers, and some of the latest converts in the U.S. (black Muslims, Cat Stevens) are there any examples of voluntary conversion to Islam? I can't think of any off hand. The original spread of Christianity within the Roman empire by contrast was completely voluntary, and often in the face of ferocious persecution. Jesus never advocated the sword, but Muhammad most unequivocally did. etc, etc, etc.

The guy who converted the goths to Arian Christianity was Ulfilas or Wulfila (same guy, different names). I believe the only Germanic people that went directly from paganism to Catholocism without stopping at Arianism were the Franks. Except for the odd case of the Icelanders, who missed out on Arianism and had their official religion decided by this guy after some political deliberation.