From such traditions we learn, for example, that Muhammad had red hair, that he liked cats (even cutting his cloak apart so as not to disturb a sleeping cat), that he had a weakness for perfume, and that he proffered an abundance of admonitions on everything from the diseases of livestock to the proper use of the toothbrush, not to mention voluminous dicta on prayer, fasting, sexual relations, the treatment of friends and relatives and the proper stance of the believer before God. These traditions in their considerable minutiae constitute the Sunna, that model comportment of the Prophet which all believers are expected to follow and emulate — and from which Sunni Muslims, the mainstream tradition, take their self-description. 

There is also the Prophet's biography, the Sira, which provides exhaustive detail about Muhammad's daily life, including his many conflicts with his adversaries, his military campaigns, his pious practices, and his singular personality. Finally, there is, of course, the Koran itself, the scripture containing the revelations accorded to Muhammad, beginning around 612, when the angel Gabriel first appeared to him on Mount Hira near Mecca, and continuing sporadically, until his reported death in 632. The Koran and the Hadith, taken together, form the two sacrosanct sources for Muslim belief and practice; they provide the foundations of sharia, or Islamic law, in its various (often quite disputatious and opposing) "schools", and so underpin every aspect of Muslim religious and social life.

There is only one problem with this mass of information, and it's a big problem. Not one shred of it — with the exception of the Koran — is incontrovertibly dateable to the early seventh century. The first biography of Muhammad was compiled by Ibn Ishaq, who died in 773, more than a century after the Prophet's death, and even that has not survived as Ibn Ishaq wrote it; rather, parts of it were incorporated in the later work of one Ibn Hisham, some 60 years later, to form what we know as the Sira. Thus, almost two centuries separate the Prophet's lifetime from his official life. As for the Hadith, even Muslim scholars from the tenth century on recognised that many of its traditions were not merely of dubious provenance but had all too often been fabricated or fatally embroidered. Much later, European Orientalists, such as the great Hungarian Jewish scholar Ignaz Goldziher or Joseph Schacht, a pioneering historian of Islamic law, would conclude that virtually all of these traditions were spurious; that is, not necessarily forged or invented — though many were — but simply too late or too dubious to be authentic. Over the last four decades this scepticism, not only about the Hadith but about all the long-accepted sources, has continued to mount. The existing evidence — or rather, the existing non-evidence — has been put searchingly to the test by those whom Robert Spencer calls "a small band of scholars who have dared, often at great personal and professional risk, to examine what the available historical data reveals about the canonical account of Islam's origins".

By way of contrast, in his little book Muhammad, in the series "All That Matters", the well-known journalist and Muslim apologist Ziauddin Sardar provides that "canonical account" of Islam's origins. Unfortunately, this is not a good book; it is at once drab and tendentious, written in lacklustre prose and much given to witless anachronism. Thus, Sardar defines the Arabic word ummah (literally, "community" but denoting the entire Muslim community) by saying, "The original significance of the term ummah was a multi-religious, one could even say multicultural, community committed to defending its joint interests". This is arrant nonsense.