Since the 19th century, Western scholarship has taken for granted that in the first 100 years after Muhammad’s revelations, Islam was practiced much the same way it is today. Western scholars explained the birth and early expansion of what is now one of the world’s largest religions through the development of its army and political institutions, the need for social change among Arabian nomads, or simple economics. But “they seldom talked about the religious motivation,” says Islamic scholar Fred M. Donner.

A professor of Near Eastern history at the Oriental Institute and head of Chicago’s Center for Middle Eastern Studies, Donner instead believes Islam’s origins shared features with the genesis of Christianity.

The idea that Christianity didn’t spring fully formed from Judaism with Jesus’s preaching is well accepted; scholars and laypeople alike understand that there was an early germinal stage before the canon was worked out at the Council of Nicea and subsequent Church council meetings.

Donner says that Islam too went through an early “ecumenical phase” when Muhammad’s followers were a loosely defined community—Donner, following the Quran, calls them “the Believers”—that may have included Jews and Christians. These followers were committed more to monotheism than they were to Muhammad. “It was more of a monotheistic revival movement,” Donner says. In 2010 he posited this theory in Muhammad and the Believers (Belknap). Islam, he writes, began as a religious movement, “not as a social, economic, or ‘national’ one. The early Believers were concerned with social and political issues but only insofar as they related to concepts of piety and proper behavior needed to ensure salvation.”

Donner’s conclusions diverge from the traditional view, which “sees Islam as being codified from the very first day,” he says. According to that story, the prophet Muhammad settled in the Arabian town of Medina after being expelled from nearby Mecca, and soon afterward he began to spread Islam. After Muhammad’s death in 632 AD, his teachings disseminated through the Middle East via military and bureaucratic expansion, eventually moving beyond Arabia.

But according to Donner’s interpretation, it took perhaps 100 years after Muhammad’s death for the religion to consolidate and establish a Muslim identity apart from other monotheistic religions. “At a certain point, the people leading the movement redefined it to be a separate confession,” he says. “They defined Jews and Christians out of it and defined themselves as Muslims with the Quran as their scripture, the prophet Muhammad as their prophet.”

As evidence that early Islam included Jews, Donner cites the treaty Muhammad drew up with the Medinese, which mentions Jews within the community he headed. To Donner this indicates that for Jews at the time, believing Muhammad was a prophet didn’t conflict with the Torah. In fact, Donner interprets the Quran’s use of the word muslim, literally “one who submits,” to connote a monotheist rather than a follower of Muhammad.

This definition seemed to persist: seventh-century Christian sermons indicate that some Christians of the time still considered Islam a new and errant form of their religion. And in the early eighth century John of Damascus, a Christian monk who served in the court of the Muslim King of Syria, wrote the Heresy of the Ishmaelites, a treatise indicating he saw Islam as an outgrowth of his own religion.

More than a year after the release of Muhammad and the Believers, Donner calls reception of his book “fairly positive.” Some peers shake their heads at his conclusions, but others perk up, and Donner says he’s gratified by an enthusiastic, “if mutedly so,” reaction from scholars working in Muslim countries. His view of Islam’s origins is revisionist, but it’s less extreme than some theories over the past 30 years. Donner doesn’t, for example, propose that Muhammad didn’t exist, as theologian Sven Kalisch of the University of Münster did in 2008.

Muhammad was a real person, says Donner, who has studied the documents surrounding Islam’s history for more than 25 years, from seventh-century papyri, inscriptions, and coins to later chronicles and books of collected traditions. In 1981 he wrote a history of the early Muslim conquests, and in Narratives of Islamic Origins (Darwin Press, 1998) he studied the development of traditional Muslim sources for Islam’s beginnings.

Partly what compels Donner’s revision of the historical record is the scarcity of sources written during and right after Muhammad’s lifetime. The date of the Quran itself is “a subject for debate,” he says, although he believes it was written by the end of the seventh century, within 30 or 40 years of Muhammad’s death. “There’s a serious source problem for anything dealing with early Islam,” Donner says. “This has been known for a long time. The problem is, the sources describing Islam’s origins are mostly written later,” in some cases hundreds of years later. “Actual documentary evidence is sparse, but the bits we have suggest that the traditional narrative isn’t exactly right.”

Some evidence is in early Arabic papyri written by Muslims or those in contact with them. Recently Donner has turned fuller attention to the papyri, which date to the period of Muslim rule in Egypt, soon after Islam began. They are “unmediated” sources, he says. “It’s what somebody wrote at that time for a particular purpose: a bill of sale, or a manumission document of a slave, or purchase of a house, or a shopping list, or some narrative that was being told at the time,” he says. “Precious evidence.”

Donner tempers his own claims about Islam’s origins, pointing out that sources aren’t entirely clear, so his interpretation remains in some ways hypothetical. “It’ll be interesting to see how things evolve,” he says. “No historical interpretation lasts forever. All you can do is contribute to the conversation and move it along.”