How AI Will Revolutionize the Study of Early Religion

[Secret Alias]

Hey everyone, I wanted to share some thoughts on a topic that has been weighing on me as someone deeply involved in studying early Biblical traditions. The core issue for me is the heavy reliance on the concept of “TRUTH”—an absolute truth, an all-encompassing understanding of what is "RIGHT." This perspective has dominated Biblical studies for centuries, especially with sanctioned texts of the "Bible" being viewed as infallible.

But here's where AI becomes a game-changer. Instead of being locked into these big “T” truths, AI opens the door to smaller “t” truths that deserve consideration. In traditional scholarship, the accepted method is to treat texts like the “gospels,” “Letters of Paul,” or “Acts” as coherent pieces that we piece together to create a seamless history of Christianity. Scholars often just default to this method and, if you dare challenge it—like the Marcionites did—by suggesting these primary sources were corrupted, you're labeled as fringe.

And I get it. The bias isn’t always conscious. But it’s there. One of the most glaring examples of this bias is in how terms like “Jew,” “Jewish,” and “Judaism” are treated. As a Jew, this hits home. Judaism is often painted as the pure root of tradition, but anyone who's delved deep into the texts knows that Judaism was just one sect, stemming from an older cult based on Mount Gerizim. The Pentateuch hints at this if you’re willing to look past the biases of later interpretations.

AI is powerful here because it isn't biased by centuries of theological tradition. It can take in texts, analyze them, and give a more level-headed perspective. AI doesn’t subconsciously favor one perspective over another simply because it’s been deemed “traditional.”

Building on that, I think AI will fundamentally change the way we approach ancient traditions by moving the discussion from "what is the truth" (which is often impossible to know) to "what is possible." This shift allows for a much more open-ended and dynamic exploration of history and traditions. Instead of trying to fit everything into a preconceived framework of an absolute truth—something that’s essentially unknowable after two thousand years—we can start to explore a wider range of possibilities and interpretations. It’s a more scientific approach, grounded in probability and open to multiple interpretations rather than trying to determine a singular "correct" version of events.

AI can sift through the texts without the emotional or cultural baggage that comes from centuries of religious or scholarly traditions. This could lead to healthier discussions and better scientific exploration. What do you all think—is this shift in focus the way forward for Biblical studies?

[Secret Alias]

For instance, take my assertion that Tertullian never actually had the Marcionite gospel in front of him but was instead adapting an earlier text by Irenaeus. Irenaeus argued against Marcion using "the portions of Luke that Marcion retained." What’s fascinating is that Tertullian never explicitly says, “I have a copy of the Marcionite canon.” Instead, what we see in Adversus Marcionem is what looks like a Latin reworking of Irenaeus’s argument.

In fact, Tertullian’s text even mentions an earlier version of Adversus Marcionem written by an "apostate." Some scholars have suggested that this apostate could be Irenaeus himself, especially given Tertullian’s ties to the Montanist sect, which had some tensions with the Roman Church. The attack Tertullian lays out against Marcion reads like a section-by-section refutation using Luke, with little emphasis on Marcion "cutting out" portions of the text. It fits the structure you’d expect if someone were working from Luke, not a Marcionite gospel.

Now, German scholarship has long been fixated on reconstructing an "original" Marcionite gospel from Tertullian’s Adversus Marcionem. This has been attempted time and again because there’s a long academic trail of scholars taking for granted that Marcion's gospel was a version of Luke. Of course, Irenaeus claimed that Marcion falsified Luke, but his argument was based on the assumption that Marcion only retained parts of Luke. Tertullian simply took that material from Irenaeus and presented it as his own work.

Here’s the kicker: Irenaeus is known to have had a collection of Pauline letters that began with Galatians. This wasn’t unusual in the East, where Roman missionaries brought such collections. So it’s entirely plausible, even preferable, to consider that Tertullian was copying Irenaeus’s Adversus Marcionem, where Marcion was disproven using Luke. Marcion’s gospel wasn’t a corrupted version of Luke, but scholars have largely ignored this possibility because “so many great minds” have accepted the alternative narrative.

This is exactly where AI can help. It can cut through these long-standing academic biases and help explore overlooked possibilities that have been brushed aside simply because they don’t fit the accepted scholarly tradition. In short, AI often has more imagination than traditional scholars of religion which is strange because it speaks more of the "deadness" of traditional scholarship than of a machine actually possessing creativity.

[Peter Kirby]

I hope you're right. On the other hand:

More mutually exclusive possibilities technically just means more wrong stuff, unless everything so far was wrong.

There is great value in your explanation, not because it is possible and non-traditional, but because it's true or at least contains elements of the truth.

That's rare enough in late-stage scholarship that you should be proud of yourself for having the intuition to pursue it.

[StephenGoranson]

If AI helps, fine. I'm skeptical, or, at least, I don't see it as a black and white divide.

Some historians do a good job. When, for example, Peter Brown wrote about Late Antiquity, and about opinions about embodiment and icons, and about Saint Augustine, and about the life of the mind, he is illuminating. People have biases, but different people have different ones. Sometimes dialogue works well, and some folks can discern good bits. Some scholars are better on some subjects than others.

AI has some sort of intelligence, but a sort with limits. So far, it seems largely to regurgitate what it picked up from texts, texts written by people. Also, its prose is wooden and boring.

[Secret Alias]

I agree that AI has its limitations, and it's not a magic solution to historical research. It's important to recognize that AI is, as you said, largely regurgitating existing texts written by humans, and yes, the prose can feel wooden. But rather than focusing on truth or black and white distinctions, I think it's more helpful to view AI as a tool that opens up new possibilities.

Historians like Peter Brown are brilliant at illuminating aspects of the past, but even they are working within certain frameworks and biases. AI, while limited, can help us explore patterns, compare sources, or suggest connections we might not have considered. It's not about relying on AI for definitive answers, but about using it as one of many tools in a broader dialogue that allows us to entertain different perspectives, enrich our discussions, and maybe even uncover new insights.

In the end, it’s about possibility rather than truth, and AI is just another way to widen that scope.

[Secret Alias]

That's where we differ on the Letter to Theodore. I don’t see its discovery as any more "corrupt" than any other relic from early Christianity. I don't think orthodox Christianity’s version of "truth" should automatically negate Marcionism. Instead, I say, "let it breathe," "bring it back to life," and "learn from it." In the same way, even if the Letter to Theodore might have been ancient, we should "keep it alive."

I see myself aligned with the scholarly camp of Hillel—show mercy on everything. Don’t dismiss things just because they’re hard to fully grasp. It’s better to show mercy than to rush to judgment.

[StephenGoranson]

Well, just another tool sounds different than a revolution.

[Peter Kirby]

I see myself aligned with the scholarly camp of Hillel—show mercy on everything. Don’t dismiss things just because they’re hard to fully grasp. It’s better to show mercy than to rush to judgment.

This makes more sense than advocating for the enumeration of infinite possibilities so long as they are non-traditional and therefore more interesting.

For example, I argue that early Christianity was clearly an extension of Alexandrian Judaism as outlined by Philo. And regarding Second Commonwealth Judaism, I propose that the Sadducees were, in fact, more like neo-Samaritans. This is my main argument, but I don't see these points reflected in the mainstream narrative of the first century.

And it's the connections formed between these ideas and the source material that makes these proposals compelling. Such connections find and ground their value, ultimately, in the truth.

[StephenGoranson]

SA, above, in part:
"That's where we differ on the Letter to Theodore. I don’t see its discovery as any more "corrupt" than any other relic from early Christianity."

Holy mackerel.
In this thread I didn't bring up that Letter.
Do I really need to remind you of its first claim in 1958--among other reasons--makes it shakier and possibly more corrupt than some other evidence?

[Secret Alias]

It’s not impossible that it’s authentic. In the same way, the canonical gospels are likely forgeries, but that doesn’t mean we shouldn’t study them. The same goes for the rest of the New Testament and much of early Patristic literature. If we dismiss texts simply because they might be forgeries, we’d be left with very little to work with. Instead, all of these texts should be studied with an asterisk—acknowledging their potential for forgery, but not discarding them outright. Our judgments are often made on a collection of likely forgeries, and very few, if any, completely "authentic" texts. Yet, these works still provide valuable insights into early Christianity and its development.