What I have been saying, and I continue to maintain, is the fact that:

1. the percentage of reprehensible acts committed per capita by Muslims is no greater than that of any other major monotheistic religion, as well as people of no religion... (Carol)

A very successful argument could probably be made that "Christians" have committed more reprehensible acts per capita over the history of that religion than any other (Carol)

My other premise is that it is simply not possible for anyone to make a legitimate and verifiable argument about which religion or religions, if any, are more prone to any kind of tendencies whatever, and that any kind of generalization one tries to make in answer to that question will be opinion and nothing more. (Carol)

Fact, premise, opinion?? Is your first statement that comes in form of an empirically true statement just an opinion or are your premises facts or your facts premises?

I can only repeat what I have said above: I have no empirical knowledge on any relationship between religion and number of crimes but I would consider it very likely that religion, like nearly all other cultural or biological contexts, also influences the number of per capita crimes. Incidence of crimes is rarely statistically independent of age, gender, social standing of the offender. So why should it be independent in the case of religion?
And add: Something like religion that influences so deeply ethics, morals and beliefs of humans is bound to have observable influences on behaviour. To state categorically that in this case the statistical null hypothesis will be true whatever relation is studied, does not make any sense.

I think what you mean, Carol, is very noble (you will correct me, if I don't get what you mean). Even if there should be any observable difference in behaviour (criminal, or other, perhaps even recommendable) being correlated with the variable 'religion', those mentioning such a fact would most probably have sinister motives. Any data of that kind could/would be used for political manoeuvres of a despicable kind. The interpretation some might make based upon such data may be biased and just one of many possible interpretations, the interpretation of the choice being influenced by prejudices or political opportunity. Therefore you prefer shunning even the thought that there could be data showing a difference.

That's the wrong and ill-fated approach in my eyes. Data are data, even if they can be abused and have been abused. The attack must be on the sinister interpretation and not on the data. It is a fact, for instance, that in the USA IQ-data differ depending upon ethnic origin. It is of no use to attack that fact, but it is worth looking closely at (and criticising) the possible interpretations ((1) shows genetic differences in IQ (2) demonstrates culturally biased tests (3) can be explained by economic disadvantages of some groups). The problem is that if one attacks a fact when you means one possible, or even popular, interpretation one looks quite silly if the facts are corroborated beyond doubt. Even worse, the sinister interpretation seems to gain undeserved support.

One example: The criminal statistics in Germany show higher rates of some crimes in some ethnic groups (and since religion is not equally distributed among ethnic groups, of course, religion would be a statistically valid predictor as well) living in Germany. The Neonazis like to quote those statistics for their "foreigners out" agenda. One approach, we call it the 'social worker approach', is to claim that this is not true, and to mix this way statement of fact with critique of a Neonazi agenda. They mean well, but they do damage to their good cause this way. The Neonazis then make the demonstratable empirical truth of the fact their issue and in some people ("well, at least they dare to tell the truth") their interpretation gains a support it should never have.

The right approach is not to question the facts (if they are true; sometimes they just aren't, but that's a different story) but to go the more difficult way and to question the interpretation (or the motives for chosing one particular interpretation). For instance, one could point out that a variable that explains much more of the variance in the data is economic background, and that a subgroup of Germans matched with the ethnic minority on economical variables shows the same amount of criminal behaviour (if that is true; it's not always true). Or one could point out that the average age of foreigners lies below the average age of Germans (that's true) and since the variable 'age' can explain a lot of criminal behaviour (that's true; males from puberty to roughly 30 dominate the criminal statistics) an age-matched group of Germans also shows a higher incidence of criminal behaviour.

However, the I-don't-want-to-hear-any-of-this approach of closing the eyes can backfire.

Wolfgang