Now I’m not a historian, but I do know a bit about the scientific method. From what I’ve read about historiography, it seems to follow a basic scientific methodology:
Physical evidence seems to be the primary way that history is reconstructed (hence Biblical Minimalism) and written sources are used as secondary way that history is reconstructed. In other words, we shouldn’t go to written sources first since physical sources can tell a lot more about the stories of history.
This is how it’s done in every other field of science, and it’s logical and consistent that historians do the same.
The problem with Christianity is that we have no “primary” evidence, so we have to rely on writings. What I’ve learned from my few ethnography/sociology classes at NYU, people’s writings are terribly biased when processed through the filters of their own epistemology. This is the problem that sociologists have to contend with when they’re studying their subjects. I bring up sociology because history is a mix of both the hard science of archaeology and the soft science of sociology – basically sociology in the past.
In science, even in hard sciences, all knowledge is tentative. The Theory of Relativity basically says “this is how it’s worked so far”, and it will always be open to modifications to that theory if evidence pushes things in that direction. Soft sciences are even more tentative in their conclusions, since human behavior is even harder to predict than gravity.
Back to the problem with early Christanity and its history: in this case we only have writings. For the most part, early Christian writings are undated, and the earliest narratives about Jesus are anonymous. Another problem is that these writings are opinions about religious beliefs. They are writing what they want us to believe, not what actually happened. But this is not a problem restricted to Chritian writing, it’s a problem for all writing, both modern and ancient. And this is why writings are secondary material. They need corroborating data, or external controls. Secondary sources like writings can tell us more about the sociological context of the writers than what they’re actually writing about.
Related to these writers writing what they want us to believe, we know that in the 2nd century, Christians were editing these writings even further in the battle between “orthodoxy” and “heterodoxy”. So not only were the original writers writing what they wanted us to believe, but these writings arrive to us today edited by later Christians wanting us to believe what they believed. Thus you end up not with just Mark… but Mark, Matthew, Ebionite Matthew, Luke, Marcion’s Luke, Gnostic John, Orthodox John, and myriads of other permutations. All variations of Mark… all of the data are hopelessly corrupted.
But this isn’t a “show stopper” for history, since in many other fields of history they don’t treat undated and anonymous writings as a primary source. They aren’t forced to.
Thus, for example, Virgil’s Aeneid counts as evidence of both the founding of Rome, and of the social climate of the Age of Augustus. It is very bad evidence of the former, and fairly good evidence of the latter.
So at the very least, this is why I’m agnostic about the historical Jesus.