Faith and Power: Religion and Politics in the Middle East, By Bernard Lewis

A 'clash of civilisations' might yet be averted, says the historian who coined the phrase

There are few 94-year-olds whose work remains as controversial as that of Bernard Lewis. The Stoke Newington-born academic, long resident in the US and currently rejoicing in the title of Emeritus Cleveland E Dodge Professor of Near Eastern Studies at Princeton University, is either "the world's foremost Islamic scholar", if you believe The Wall Street Journal, or "someone who knows nothing about the Arab world" if you'd rather go with the late Palestinian-American scholar Edward Said.

He, not Samuel Huntington, is the man who coined the term "clash of civilisations", and stands accused of providing intellectual cover for the Bush administration's disastrous invasion of Iraq. Lewis was so bewitched by his 1950 visit to Turkey during which he saw how the heart of the old Ottoman empire had combined Islam with modernity, say his critics, that he views the imposition of an Atatürk-type secularism as the only viable long-term future for any Muslim state. "This is a person who does not like the people he is purporting to have expertise about," says one rival academic. Dick Cheney, no less, headed the guest list at his 90th birthday party.

If all that makes Lewis's latest book sound off-putting, it shouldn't. The reader might struggle to reconcile the man in the comments above with the author whose name adorns the cover. For most of this volume, he comes across as highly philo-Islamic and pacific. But it is as well to be aware of the greater context that you won't find spelt out in this collection of essays and lectures (most of which have not been printed in English, or at all, before).

As a primer for Lewis's stated subject, though, it would have been useful reading for Dubya. Even the repetition that characterises it would have been a virtue for a president who thought the inhabitants of Greece were the Grecians. And the notes that Lewis strikes again and again are both wise and valuable.

The separation of church and state that the populations of Christian and post-Christian countries take for granted is unknown in Islam. Israel has often been less the cause of fundamentalists' ire than rulers who have attempted to "de-Islamise" their own states. (Lewis notes that when one of the killers of the Egyptian president Anwar Sadat said "I have killed Pharoah", he was not accusing Sadat "of being soft on Israel; he was accusing him of being a pagan tyrant".) Pressing "moderate autocracies" towards hasty democratisation can lead to the establishment of "more ferocious and determined" fundamentalist dictatorships.

Lewis also stresses that early Islamic states did contain checks and balances that could be used to develop a particular and relevant form of democracy today. The Caliphs' rule was, in theory at least, contractual, consensual and limited. The authority of 20th- century autocrats such as the Ba'athists, both in Syria and Iraq, he argues, came from an "imported fascist ideology and apparatus of tyranny that suppressed and supplanted their own, older Islamic traditions of lawful, limited and responsible government".

So he does offer a path from within Islam itself that would avert the need for any clash of civilisations. This is enormously to be welcomed, as is his unequivocal statement that "at no point do the basic texts of Islam enjoin terrorism and murder".

Scattered among the reasonable argument, however, are odd sentences that stop one dead. "Either we bring them freedom," he concludes one chapter about Muslims' future, "or they destroy us." At another point he writes of Iraq and Iran, "Democratic ideas have deep roots in these countries, and given the chance, they may soon prevail." He offers no evidence for this claim – nor could he.

It is lines such as these that led Ian Buruma to ask once whether Lewis possessed two minds, one of the diligent and cautious historian, the other belonging to a glib and strident advocate of neo-liberal force. Fortunately, it is the first that is most in evidence here. Although the second makes the occasional appearance, most readers will award the garlands to, and be convinced by the arguments of, the former.