Lewis of course doesn't consider religious beliefs impervious to social, economic, or political influence. However, he does consider them pretty darn stubborn; he traces the attitudes of contemporary Islamic extremists all the way back to doctrines forged in Mohammed's time. Lewis writes, for example,

If the fighters in the war for Islam, the holy war "in the path of God," are fighting for God, it follows that their opponents are fighting against God. And since God is in principle the sovereign, the supreme head of the Islamic state—and the Prophet and, after the Prophet, the caliphs are his vicegerents—then God as sovereign commands the army. The army is God's army and the enemy is God's enemy.

Now, you can find passages in the Bible, too, which imply that "the army is God's army and the enemy is God's enemy," that God's people "are fighting for God [and] their opponents are fighting against God." You can also find plenty of Jews and Christians since the Bible was written—including no small or inconsequential number who are alive at this moment—who have persisted in that attitude. So, what is Lewis' point? In what sense is Islamic terrorism distinctively Islamic in its causality?

Lewis certainly isn't unaware of the Judeo-Christian parallels to Islamic doctrine. He acknowledges that a) Islam isn't the only monotheism with such chauvinistic scriptural themes; b) some Christians inspired by scripture have (e.g., during the Crusades) behaved about as badly as some Muslims inspired by scripture; c) various Muslims at various times in history have been quite progressive and humane; d) more generally, religious belief and behavior are shaped by contemporary social forces; and so on. That's the trouble: Lewis acknowledges everything, including lots of things that sit awkwardly alongside his concluding emphasis. His Atlantic essay contains such an ample list of to-be-sure paragraphs that you start wondering why they're not his central thesis.

In fact, if you ask why Lewis has become the darling of conservatives, the answer is "the order in which his paragraphs are arranged." He starts out asking why so many Muslims seem so angry and then launches a series of to-be-sure paragraphs. Yes, it's partly America's support of Israel; but, he stresses, that alone can't explain all the rage, so we must look further. Yes, it's partly America's support of Arab regimes; but that alone can't explain all the rage, so we must look further. And so on. … Then, having shown that none of the usual suspects alone can explain Muslim rage, he brings on stage his own favored explanation: some combination of Islamic doctrine and the triumphant encroachment of Western culture. And that explanation is the one that all readers remember as the winner!

But of course, this "winning" explanation can't alone explain Muslim rage any more than any of the to-be-sure explanations could. After all, there have at various times and places been large Muslim populations that, though poorer than their relatively modern, relatively secular, non-Muslim neighbors, were nonetheless not inclined toward terrorism. Indeed, Lewis' many to-be-sure paragraphs amount to a concession that his favored explanation can't alone account for Muslim rage. So, logically speaking, Lewis could just as easily have trotted out his favored explanation first, stressed that it can't alone explain all the rage, and proceeded on down the list of usual-suspect explanations. Then whichever explanation he resoundingly ended on would be the one readers remembered as the winner, even though what he'd actually said was that Islamic terrorism has lots of causes, and no single explanation alone suffices. In other words: The "conservative" take-home lesson of Lewis' famous essay is essentially a result not of its analytical content (i.e., that no explanation alone suffices), but of its rhetorical framing.

Having said that, I should add that Lewis' vast knowledge and literary fluency make his writing both useful and enjoyable. So read him. But when deciding what the take-home lesson is, carefully distinguish between what he says and how he says it.