That's true, perhaps, but it's not what I asked for. How we determine what's the "real" Islam?
I think we must analyze a) the roots of the religion b) the real preactice of the religion.
Both look very violent. The recent form of Islamic violence (jihadist terrorism) is just a continuation of its violent tradition which started with Muhammad's conquest of Mecca.
Since collective human behaviour tends to subvert attempts to a priori deductively define it, I think you need to go with the functional approach, with a decent level of induction.
Disuccing the history of, Islam warts and all, I have no problem with, as long as it's on the understanding that history has the privilege of being non-reductionist, which again means that it's not a fit basis for making essentialist calls on what any phenomenon "really" means. What it was then doesn't decide what it is now. It may detail certain predispositions, a basic conceptual framework open to interpretation, addition, excision of various parts.
"Roots" don't work as people tend to think they do. Westerners, and others, are predisposed to think that an intellectual geneaology works by having a a Great Ancestor who triggers everything. Genealogical thinking in a society is one of the more powerful model which unwittingly predisposes peoples thinking.
The bullet is fired in the past, and then everyone else is obliged to follow the trajectory. It allows the assumption that somehow we can work out what an intellectual tradition really means by looking at the Great Guy at the inception, the Original Form.
In reality there's more to suggest that the process works the other way around; every generation is obliged to reconnect, and at this point there is more or less severe tinkering with the tradition. Which is why there is a constant necessity in intellectual history to reassess the Great Ancestors, and it invariably turns out their present day followers have made something very different of the tradition they started.
So the problem with postulating a deductive analytical definition of something tends to be that people then go and do things differently. In that case one can of course claim to simply have a negative result in this case it might allow someone to conclude that the people calling themselves Muslims are in fact not "real Muslims" in accordance with the definition (this is actually what the al-Q. is doing at present).
It could work as a first step of an analytical process, but in itself it's insufficient. Someone will come along and say "Yes, yes! But since these people still admit to a collective identity as Muslims, no matter what your definition says, we want to know what's really going on with them here. All your definition does is tell us to go look at some other people."
Short of it: Analytical definitions of Islam made by non-Muslims tend to be a form of the "glass-bead game", academic or otherwise. The trick is to work out what's going on with Islam and Muslims in general, not what one thinks they should be up to, given things like past history, tradition etc. If that was how humans worked, the world would be easy-peacy to understand due to the deadening regularity and conformity of human behaviour.
But it's not.
- we can, in fact, try to see if there are general tendencies in certain ideologies/religions/"national culture" which tend to influence a certain degree of a given population toward specific behaviour, but even this is very sketchy scientifically. Christianity is as possibly "harmful" to an already anomic individual as Islam or Conservatism etc. is and each can be used to create incentives for directed action from different bodies inside a society, so it is not really a valid category if used on a broad scale, so I really don´t see the point.
