Credibility of ancient texts

 

Of course there are truths in the Bible. No one doubts for example that there existed a city of Jerusalem or a country of Egypt. That is like saying there are truths in the Iliad, although few today believe in Apollo and Athena and so on. All texts, even modern ones, but even more, ancient texts, need to be read skeptically (not cynically), and texts that refer to outlandish things like talking asses and fire and brimstone from the sky and divine voices from burning bushes need a particular skepticism: one that says the more "out" a statement is, the more evidence is needed. Whenever an ancient text speaks of such thing, it should be doubted. This is simple enough, but the thing is the presence of such statements bring into question not just the miraculous parts but all of it. This is a fundamental rule of reading ancient texts: it it contains narratives contrary to normal events, then not just just the parts that make us doubt become questionable, the whole thing becomes questionable. This doesn't mean the ancient authors were deliberately lying, but only that they were recording stories that had come down to them in their culture, just as the stories of Robin Hood or King Arthur came down to the Middle Ages in European cultures and later got written down as fact.

What you’re outlining is very close to the standard historical method, though I’d nudge one part of your conclusion a bit, because historians tend to be more selective than your “the whole thing becomes questionable” rule suggests.

Your basic instinct—skepticism that scales with how extraordinary a claim is—is exactly right. If a text reports something like a census or a king building a wall, that sits comfortably within ordinary human experience and requires relatively modest corroboration. If it reports a talking animal or a suspension of natural law, then by ordinary evidentiary standards the bar rises sharply. That’s not cynicism; it’s just proportional judgment.

Where I’d push back slightly is on the idea that the presence of miraculous or legendary material makes everything in the text equally doubtful. In practice, historians don’t treat sources that way. They tend to “disaggregate” them—pull them apart into layers and claims, and evaluate each on its own merits. A single document can contain:

• mundane, reliable background (place names, political structures),

• plausible but uncertain narrative,

• and highly doubtful or symbolic material,

all at once.

To use your analogy, scholars don’t discard the Iliad wholesale because of the gods. They still take it seriously as evidence for Mycenaean culture, oral tradition, and even a possible historical conflict behind the story of Troy. Similarly, the Bible contains references to places like Jerusalem and Egypt that are independently attested and historically uncontroversial, even if other elements are treated as theological or legendary.

Your point about transmission is also well taken. Many ancient texts sit at the end of long oral traditions, where stories evolve, accrete meaning, and get shaped by the needs and beliefs of a community. In that sense, comparing biblical narratives to Robin Hood or King Arthur is methodologically sound: both likely have some historical kernel, layered over with centuries of storytelling.

But there’s another dimension worth adding. Ancient authors were often not trying to write “history” in the modern sense at all. They were conveying meaning—moral, theological, or cultural. A talking donkey or a burning bush may not have been intended as a testable report in our sense, but as a way of expressing something about divine agency, authority, or revelation. That doesn’t make the events historically reliable, but it does change how we interpret why they appear.

So I’d refine your rule this way: extraordinary claims don’t invalidate an entire text, but they do force us to read it more carefully—sorting genres, intentions, and levels of plausibility rather than accepting or rejecting it as a single unit.

You’re already doing the hard part, which is resisting both naïve acceptance and blanket dismissal. The interesting work lies in that middle ground—deciding, piece by piece, what kind of “truth” each part is actually trying to convey.

I would not say they invalidate the entire text and I don't think I said that. I said that the presence of the extraordinary makes the mundane parts questionable.

Ah—yes, that’s a more precise (and more defensible) position, and I take your point.

Framed that way, you’re describing something quite real: extraordinary elements in a source don’t disqualify it, but they do lower our prior confidence in the author’s reliability as a reporter of events. In other words, they shift the burden of proof even for the ordinary claims.

Historians would largely agree with that—but again with a qualification. They don’t treat that lowered confidence as uniform across everything in the text. Instead, they ask why the extraordinary material is there, and whether it reflects on the author’s general reliability or only on certain kinds of claims.

For example, if a text treats miraculous events as literal history in the same narrative voice as mundane events, that does indeed raise a question: is this author distinguishing between the probable and the symbolic? If not, then yes, one becomes more cautious even about the ordinary details.

But there’s a counterweight. Many ordinary claims in ancient texts can be independently checked—place names, administrative titles, trade routes, political events. If those repeatedly line up with archaeology or other sources, confidence in those specific kinds of claims rises again, even if the same text also contains the marvelous.

So what you’re really pointing to is something like this:

• Extraordinary claims → reduce initial trust in the author’s judgment about reality

• Independent corroboration → rebuild trust in particular domains of the text

That’s why a work like the Bible can be treated simultaneously as:

• unreliable for miracles (by historical standards),

• mixed for narrative history,

• and quite solid for certain geographical or cultural details.

And similarly, the Iliad is not trusted about divine intervention but is still mined for what it accidentally preserves about the world behind the story of Troy.

Your formulation—the extraordinary makes the mundane questionable—is essentially a caution about source reliability. The historian’s tweak is: questionable, yes—but not equally so, and not permanently so if other evidence steps in.

It’s a bit like judging a witness. If someone tells you one wildly implausible story, you don’t ignore everything else they say—but you do start asking, more carefully than before, “How do we know this part is true?”

The presence of independent confirmation in a more strictly mundane source, or maybe from archaeology or simple geography, reflects credibility, but in that case we are not treating the source as our evidence, but the independent confirmation as the evidence. The text itself is not being relied on.

You’re zeroing in on a real tension in how we use sources, but I’d draw the line a little differently than that.

It’s true that independent confirmation carries the real evidentiary weight. If an inscription, excavation, or second source lines up with a claim, then that external evidence is doing the heavy lifting. In that narrow sense, you’re right: we’re not believing the claim because the text says so.

But it doesn’t follow that the text “is not being relied on” at all. What’s actually happening is more subtle and, frankly, more interesting.

The text is still doing at least three kinds of work:

First, it makes the claim in the first place. Without the text, there is nothing to confirm or disconfirm. Even if archaeology later supports it, the text remains the original witness.

Second, once some of its claims are confirmed, the text acquires what historians sometimes call domain-specific credibility. Not blanket trust, but a track record. If a document repeatedly gets mundane details right—names, places, administrative structures—then historians become more willing to provisionally trust it on similar kinds of claims where independent confirmation is missing.

Third, it helps interpret the independent evidence. Archaeology rarely speaks in sentences. A ruin, an inscription fragment, a layer of ash—these don’t come labeled. Texts often provide the narrative framework that makes sense of otherwise ambiguous data.

So rather than a clean separation—“the text isn’t evidence, only the corroboration is”—it’s more of an interaction:

• The text proposes

• Independent evidence disposes (confirms, modifies, or rejects)

• The results feed back into how much weight we give the text next time

Take something like the Bible: when certain kings, cities, or events are corroborated, historians don’t suddenly treat the whole text as reliable. But neither do they ignore it. They begin to sort it—this part seems grounded in administrative reality, that part looks theological or legendary.

The key point is that sources aren’t used in isolation. A text is rarely the evidence, and archaeology is rarely self-sufficient evidence. Each becomes more informative in the presence of the other.

So I’d rephrase your point just slightly: when independent confirmation exists, it anchors the claim—but the text remains part of the evidentiary web, not something discarded once better evidence appears.

You’re absolutely right to resist granting authority to a text simply because it exists. The historian’s move is not to abandon the text, but to keep it on a short leash—tightened or loosened depending on how it performs under cross-examination.

We diasagree here. To me when a text reports fable or myth as fact, it loses credibility, end of story. There is no way to cherry pick the parts we want and the parts we don't. It may be that there are accurate things in it, but if so these must be determined independently and the text cannot be used as part of the argument

I see the line you’re drawing, and it’s a clean one: once a source shows itself willing to present the fabulous as fact, you treat it as an unreliable witness across the board and exclude it from evidentiary use. That’s internally consistent—and it has the virtue of protecting you from being misled.

Where I’d still push back (gently, but firmly) is that this standard is actually stricter than the one historians use, and if applied consistently, it would force us to discard a very large portion of the ancient record—not just religious texts.

Take the Iliad again. By your rule, it’s out entirely: gods intervene, speak, take sides in battle. Yet we still use it—carefully—as evidence for things like warfare practices, social structure, and oral tradition. Not because we trust it wholesale, but because parts of it behave like plausible human reporting embedded in a poetic framework.

Or consider early historical writers like Herodotus. He reports things that are clearly fanciful alongside things that are independently confirmed. If we discard him completely, we lose one of our primary windows into the ancient Mediterranean world. Instead, historians read him critically, not categorically.

Your concern about “cherry-picking” is the strongest part of your argument, and it’s a fair one. If selection is arbitrary—keeping what we like, discarding what we don’t—then it’s not serious method. But in good historical practice, the selection isn’t based on preference; it’s based on criteria:

• Does this claim fit what we know about human behavior and physical possibility?

• Is it internally consistent?

• Does it align with independent evidence where available?

• Is the genre (poetry, theology, court chronicle) one that signals a different standard of reporting?

That’s not cherry-picking in the casual sense; it’s closer to cross-examining a witness who is partly reliable and partly not.

The alternative—your approach—treats credibility as binary: once broken, it’s gone. The historian’s approach treats it as graded and local: a source can be unreliable in one domain (miracles) but still informative in another (geography, administration, social custom).

Now, there is a cost to your position, and it’s worth making explicit. You gain rigor and safety from error, but you lose a lot of potentially usable information, especially in periods where all sources are messy by modern standards. Ancient writers almost never separate the natural and supernatural the way we do.

So the disagreement really comes down to a philosophical choice about evidence:

• You prefer a high threshold that excludes compromised witnesses entirely.

• Historians prefer a lower threshold but compensate with constant, fine-grained skepticism.

Neither is irrational. Yours is cleaner; theirs is more pragmatic given the poverty of sources.

If I had to pinpoint the crux: you’re treating credibility like a switch, while historians treat it like a dimmer.

We can learn from Homer about things of Homer's time. I fear a lot of what you say derives from a desire to know things that better discipline tells us cannot be known.

That’s a fair caution—and honestly, one historians themselves worry about more than they sometimes admit. There is a constant temptation to squeeze more certainty out of the past than the evidence really supports.

Your example of the Iliad is actually a good place to sharpen the distinction. You’re on very solid ground in saying we can learn about Homer’s own world—the social assumptions, the poetic conventions, perhaps some lingering memories refracted through centuries of storytelling. That’s a disciplined, defensible use of the text.

Where historians sometimes go further—trying to extract details about a putative historical Troy or a specific war—your skepticism bites harder, and not unreasonably. At that point, the risk of overreach increases: reading history out of literature that was not designed to preserve it accurately.

I think the disagreement between your position and the mainstream one isn’t really about logic—it’s about risk tolerance.

You’re essentially saying:
If a source shows signs of unreliability (in your sense), then using it to reconstruct past events invites self-deception. Better to accept ignorance than to build on compromised foundations.

The historian’s counter-impulse is:
If we apply that standard too strictly, we may end up knowing almost nothing about large stretches of the past, because nearly all surviving sources are, by modern standards, compromised.

And you’re right to suspect that sometimes this second impulse shades into wishful thinking—what you nicely called “a desire to know things that better discipline tells us cannot be known.” That does happen. Whole reconstructions have been built on very thin reeds.

But there’s also a middle practice, at its best, that tries to avoid both extremes: not trusting the text, not discarding it, but using it almost against itself. Not “Homer says X, therefore X,” but “given that Homer says X, what must have been assumed, known, or imaginable for that statement to make sense?”

That kind of inference is less about recovering events and more about recovering mental worlds.

So I think your position serves as a necessary corrective—a kind of intellectual brake. It guards against overconfidence and narrative invention. Where I’d still hesitate to go all the way with you is in abandoning texts as evidence entirely once they’re “contaminated,” because in many cases they are the only windows we have—even if the glass is warped.

Put differently: you’re insisting we not mistake a distorted window for a clear one. Historians, at their best, would agree—but still try to look through it, cautiously, because there’s no other view available.

The publish or perish motive can lead to a great deal of ingenuity and jumping on things that may just be coincidence.

That’s not just a suspicion—it’s a well-recognized structural problem in academia.

When careers depend on producing something “new,” there’s a built-in incentive to find patterns, connections, or interpretations that haven’t been noticed before. Sometimes that leads to genuine insight. But just as often, it encourages what you’re pointing at: very clever arguments built on thin or ambiguous evidence, where coincidence starts to get dressed up as causation.

History (and related fields like biblical studies or classics) is especially vulnerable because the evidence base is so fragmentary. When you only have a handful of texts, a shard of pottery, and a few inscriptions, it doesn’t take much imagination to weave a story—and once written in scholarly language, it can look far more solid than it really is.

You see this in things like:

• elaborate reconstructions of events from a single passage,

• arguments built on linguistic coincidences,

• or confident narratives resting on what are essentially gaps in the record.

And yes, the “publish or perish” pressure can reward that ingenuity, even when it outpaces the evidence.

That said, there are counterforces. Good scholarship gets tested over time. Other specialists push back, replicate (or fail to replicate) arguments, and compare against new findings. A lot of those clever but fragile theories quietly die after a decade or two. The ones that survive usually do so because they’ve accumulated independent support.

So you’re right to be wary. Your instinct—to suspect that some conclusions are driven as much by professional incentives as by evidence—is a healthy one. It aligns with your broader caution about sources: don’t grant more certainty than the foundation can bear.

If anything, your standard acts like a kind of filter. It may exclude some things that turn out to be true, but it also protects you from a great many things that only look true because someone was motivated—and talented enough—to make them seem so.

I might suggest that the consequences of my view are not as bleak as you paint, that we may not be able to know certain refined details but we still get a good picture.

That’s a strong—and honestly quite defensible—way to frame it.

If what you’re after is a reliable outline rather than a richly detailed narrative, your approach holds up very well. You’re essentially trading resolution for confidence: fewer details, but the ones you keep are on firmer ground. And for many purposes, that’s exactly the right trade.

In fact, a lot of what historians feel most confident about already looks like what you’re describing. Broad structures tend to be much more secure than fine-grained stories. We can be quite confident about things like:

• the existence and relative importance of places like Jerusalem or Egypt,

• large-scale patterns—trade networks, agricultural systems, imperial administration,

• the general contours of conflicts and migrations,

even when the specific narratives attached to them are uncertain or embellished.

Where your approach bites hardest is on the “refined details,” as you put it—the names, motives, speeches, precise sequences of events. And it’s exactly there that overinterpretation tends to creep in, especially under the pressures we were just talking about.

So I think your position can be restated in a very productive way:

You’re not denying that we can know the past—you’re insisting that what we can know with confidence is mostly structural and coarse-grained, not narrative and fine-grained.

That’s not bleak at all. It just produces a different kind of history—less like a novel, more like a map. The map may lack the color and drama, but it’s far less likely to lead you astray.

Where I’d still keep a small reservation (and it’s a modest one) is that sometimes a few carefully handled texts—even imperfect ones—can give us glimpses of those finer details, especially when multiple sources converge. But I take your warning as a useful discipline: those moments should be treated as rare and provisional, not the default.

Put bluntly, your standard won’t give you a vivid past—but it will give you a trustworthy one. And there’s a lot to be said for that.