---

I recently listened to an episode of the Gathered in One podcast in which the hosts presented a large collection of alleged similarities between Old World (Mesopotamian) and Mesoamerican art, iconography, religious symbolism, and material culture. The episode makes a broad case for pre-Columbian transoceanic contact and ultimately connects that possibility to questions surrounding the Book of Mormon.

There is a great deal in the presentation that is worth investigating. Some of the proposed parallels are genuinely interesting. Some may deserve considerably more scholarly attention than they have received. The episode also raises potentially testable questions about artifact provenance, chronology, and cultural transmission.

My concern, however, isn’t primarily with the conclusions. It is with the method of getting there. I’m not arguing here that Old World–New World contact didn’t occur. Nor am I arguing that every proposed parallel is coincidental. My argument is narrower: The existence of a large number of claimed similarities doesn’t, by itself, establish the historical mechanism that produced those similarities. That distinction is important.

The basic argument used in the podcast

The episode repeatedly follows a pattern something like this:

1. A motif appears in the Old World.
2. A similar motif appears in Mesoamerica.
3. Another similarity is identified.
4. Another is added.
5. The similarities are said to become increasingly difficult to explain as coincidence.
6. The cumulative similarities are then interpreted as evidence of cultural diffusion.
7. Cultural diffusion is subsequently used to support a model of transoceanic migration or colonization.

For example, the presenters discuss similarities involving bird-human figures, animal-human hybrids, “handbag” motifs, fish imagery, eagle feet, jaguars, trees, crocodiles, underworld imagery, and various religious compositions. The argument is explicitly cumulative: individual examples might be dismissed, but when they are “stacked” together, the presenters argue, coincidence becomes increasingly implausible.

There is an intuitive appeal to this. If two cultures share one unusual feature, perhaps it is a coincidence. If they share ten unusual features, surely something is going on. But there is a problem: We can’t determine how compelling a collection of similarities is simply by counting them. We first have to establish what those similarities mean.

Similarity is an observation, not an explanation

This is perhaps the most important methodological distinction. Suppose we establish that two artifacts are similar in iconography.

That gives us: A resembles B.

It doesn’t yet give us: A was derived from B.

And that still doesn’t give us: The people who produced A traveled from the region where B was produced.

And that still doesn’t give us: Those people were the particular population proposed by a specific historical account.

These are different propositions. The inferential chain then looks like this: Similarity → transmission → contact → migration → specific historical population

Each arrow requires evidence. The problem with the episode is that evidence supporting the first proposition is frequently treated as though it establishes the entire chain. But scientific reasoning requires us to ask what alternative explanations could produce the same observation.

A similarity could result from:

• independent invention;
• convergent symbolic development;
• common human cognitive or religious patterns;
• similarities inherited from a much older common tradition;
• indirect transmission through an intermediate culture;
• direct cultural transmission;
• actual migration;
• or, in some cases, simple visual resemblance without historical connection.

The presence of similarity doesn’t tell us which of these mechanisms produced it. That is the question that needs to be investigated.

The problem with “stacking” similarities

The episode’s strongest rhetorical device is also one of its methodological weaknesses. The presenters repeatedly add one similarity to another until they reach the point where they argue that coincidence is no longer a reasonable explanation. At one point, the argument describes several motifs being combined into a single cumulative case and concludes that there are simply too many similarities to dismiss.

But cumulative evidence only becomes stronger if the individual pieces of evidence are sufficiently independent. This is an important concept in both statistics and historical inference.

Imagine discovering that two religious compositions contain:

• a tree;
• birds;
• a feline;
• water;
• an underworld;
• a divine figure;
• and a horizontal boundary.

It might be tempting to count that as seven independent icons. But they may not be seven independent observations. They could all be components of one underlying iconographic system. Counting every shared component separately can exaggerate the strength of the evidence. This doesn’t mean the similarities are meaningless. It means we need to determine whether the similarities constitute one piece of evidence or seven independent pieces of evidence. That distinction is very important.

The multiple-comparisons problem

There is another problem that becomes particularly important when dealing with iconography. Ancient cultures produced enormous quantities of art. If we compare thousands of artifacts from one culture with thousands from another, we have an enormous number of possible comparisons. Eventually, we will find similarities. The relevant question therefore isn’t: “Can we find similarities between these cultures?” Of course we can. The more important question is:

“Are the similarities substantially more numerous, more specific, and more structurally organized than we would expect in cultures that didn’t have contact?”

That requires a baseline. Without a baseline, statements such as “What are the chances?” or “There is no way this is a coincidence” are intuitively persuasive but scientifically incomplete. The episode repeatedly asks the audience to consider how unlikely the similarities would be as coincidences, but it never actually establishes that probability. Without a baseline, we can’t evaluate whether it is a coincidence or not.

If we want to argue that a particular iconographic pattern is unlikely to have arisen independently, we need to know how frequently comparable patterns occur in cultures where contact is known not to have occurred. That gives us a control. Without one, we are evaluating the unusualness of the pattern largely by intuition.

The Texas sharpshooter problem

This introduces a related problem often called the Texas sharpshooter fallacy. Imagine someone fires hundreds of bullets randomly at a barn. Afterward, they draw a target around the area where the bullets happened to cluster and announce that they have demonstrated remarkable accuracy. The problem is that the target was selected after seeing the data. A similar problem can arise in comparative iconography.

Suppose we begin with the hypothesis that Mesoamerican and Near Eastern cultures were connected. We then search Egyptian, Assyrian, Babylonian, Greek, and Mesoamerican art for similarities. We find a bird-human figure. Then a tree. Then a feline. Then a serpent. Then a particular type of headdress. Then a water motif. Eventually, we have a large collection of parallels. But how many potential parallels were examined and rejected?

That number is just as important as the number that was accepted. If 20 similarities were found after examining 10,000 possible comparisons, the evidentiary situation is very different from finding 20 similarities in a dataset containing only 30 possible comparisons. This is why comparative research needs explicit comparison criteria. Ideally, those criteria should be specified before the comparison is made.

How to test the hypothesis

This doesn’t mean that iconographic comparison is useless. Quite the opposite. It means that we should make it more testable. For example, rather than saying, “These two figures look remarkably similar.” We could define a set of measurable characteristics:

• anthropomorphic body;
• avian head;
• feline feet;
• associated container;
• particular hand position;
• specific headdress;
• relationship to a tree;
• relationship to a captive;
• astronomical symbol;
• position within a ritual composition.

Then we could ask: How frequently does this combination occur elsewhere?

We could compare:

• Mesoamerican cultures;
• Near Eastern cultures;
• unrelated Old World cultures;
• unrelated New World cultures.

Now the argument becomes testable. If the proposed Near Eastern/Mesoamerican configuration occurs dramatically more frequently than comparable configurations elsewhere, that is meaningful evidence. It doesn’t prove contact by itself, but it substantially strengthens the case.

Iconography and Cultural Transmission

There is also an important distinction between shared symbolism and cultural transmission. Suppose we establish that a Mesoamerican image really does share a complex structure with an Egyptian image. The next question is: How did the similarity get there? Perhaps Egyptians traveled to Mesoamerica. But perhaps a third culture transmitted the concept. Perhaps the similarity derives from an older tradition. Perhaps the symbolism arose independently. Perhaps the apparent similarity reflects our own interpretation of the images rather than the ancient artists’ intended meanings. This is why archaeological arguments for contact become particularly powerful when different categories of evidence converge. For example:

Iconography + artifact provenance + chronology + material culture + linguistic evidence + biological evidence

is considerably stronger than:

iconographic similarity + iconographic similarity + iconographic similarity.

The first gives us multiple independent lines of evidence. The second may simply be different manifestations of the same observation.

Provenance is much more powerful than appearance

Interestingly, the episode occasionally identifies the kind of evidence that could make the argument stronger.

For example, when discussing jade artifacts, the presenters suggest isotopic provenance analysis as a way of determining whether some material could have originated outside Mesoamerica. They explicitly acknowledge that this would be necessary to distinguish between locally sourced material and material actually transported from elsewhere. This is the direction I would like to see the argument move. If an artifact is made from a material whose geological source can be identified, we can potentially ask: Where did this material originate? That is much more difficult to explain away than: “This face looks Chinese.”

Likewise, an authenticated inscription with a secure pre-Columbian archaeological context would be far more significant than a symbol that merely resembles an Old World writing system. This isn’t because iconography is unimportant. It is because material provenance provides a different kind of evidence.

Starting with the conclusion

Another issue is the relationship between the evidence and the hypothesis. The episode begins with a fairly broad conclusion: that there was substantial pre-Columbian transoceanic interaction. It then gathers evidence that appears consistent with that conclusion. This creates a confirmation bias.

The problem isn’t that the researchers intend to confirm their existing beliefs. Confirmation bias can happen unintentionally. Once we have a hypothesis in mind, we naturally notice observations that fit it. A rigorous test therefore requires us to actively search for observations that could falsify it. For example: If Near Eastern influence is proposed, what specific features should we expect to find? And what features should we not expect to find? If those predictions fail, how would the hypothesis change? A hypothesis becomes scientifically useful when it makes predictions that could turn out to be wrong.

“Academics are dismissing it” isn’t evidence for the hypothesis

The episode also spends considerable time discussing the supposed reluctance of academics to accept these kinds of arguments. The podcast suggests that scholars may be constrained by disciplinary expectations or professional incentives and characterizes this tendency as something of a “party line.” There may certainly be historical examples of academic communities resisting unconventional ideas. That happens in science. But this doesn’t establish the truth of the alternative hypothesis. There is a subtle but important distinction: “Academics can be wrong” is obviously true. But “Academics reject this evidence because they are protecting an established paradigm.” is a separate empirical claim. And: “Therefore, the rejected evidence is probably correct” doesn’t logically follow.

The quality of an argument should ultimately be determined by the evidence, not by whether the people evaluating it belong to an academic institution. Ironically, the best way to challenge an academic consensus isn’t to attack the people who hold it. It is to produce evidence that the consensus can’t adequately explain.

What would actually convince me?

We shouldn’t begin with: “The conventional explanation must be correct.” Nor should we begin with: “The similarities are so extraordinary that contact must have occurred.”

Instead, we should ask: What observations would distinguish these competing hypotheses? For example:

Hypothesis 1: Independent development

The similarities should generally be explainable through common cultural, environmental, or cognitive processes.

Hypothesis 2: Indirect diffusion

We should find evidence of intermediate populations or transmission routes.

Hypothesis 3: Direct transoceanic contact

We should expect some combination of:

• securely dated artifacts;
• foreign material provenance;
• diagnostic technological traits;
• distinctive artistic conventions;
• linguistic borrowing;
• biological evidence;
• settlement patterns;
• nautical evidence;
• and archaeological contexts demonstrating interaction.

Hypothesis 4: A specific Near Eastern migration

This requires even more. The evidence shouldn’t merely demonstrate “contact.” It should demonstrate characteristics that specifically distinguish Near Eastern migration from other possible forms of contact. And if the argument ultimately seeks to connect that migration to the Book of Mormon, an additional layer of evidence is required again. This is the key point: The more specific the conclusion becomes, the more specific the evidence must become.

The conclusions may be correct

Nothing I have written establishes that the conclusions of the episode are false. It establishes something else: The argument presented by the podcast doesn’t establish its conclusion. That is important. Future research may demonstrate that some of these similarities really do represent ancient cultural transmission.

Some artifacts currently regarded as anomalous may eventually prove to be genuine evidence of transoceanic interaction. The conventional archaeological model may eventually require substantial modification. But if that happens, the case should be established through better evidence and better methodology, not simply through a larger collection of visual similarities.

This is where I think proponents of unconventional historical hypotheses can actually strengthen their arguments. They don’t need to abandon their hypothesis. They need to make the hypothesis more falsifiable, more quantitative, more comparative, and more scientifically grounded.

There is potentially an interesting research program hiding underneath this type of argument. Instead of asking, “Can we find similarities between Mesoamerica and the Old World?” we ask:

“Do Mesoamerican iconographic systems contain statistically unusual combinations of features that are diagnostic of specific Old World traditions, and can those similarities be independently corroborated through material, chronological, linguistic, or provenance evidence?”

That’s a much stronger question. And more, it is a question that can produce an answer either way. If the similarities disappear when properly controlled, the hypothesis becomes weaker. If they survive rigorous controls, and especially if they converge with independent archaeological evidence, the hypothesis becomes stronger. That is how an unconventional hypothesis should be tested.

I think there is a tendency in these debates to frame the disagreement as:

Believer: “Look at these similarities.”

Skeptic: “They’re all coincidences.”

That is too simplistic. There is a third position: “Yes, the similarities are interesting. Now let’s determine exactly how much evidentiary weight they deserve.” I don’t think we should automatically dismiss anomalous similarities. But neither should we automatically convert similarities into historical conclusions. The appropriate response is to investigate them systematically. And that means asking uncomfortable questions of both sides.

If conventional archaeology dismisses an unusual artifact without adequate investigation, that deserves criticism. But if an alternative hypothesis treats every resemblance as evidence for contact without adequately testing competing explanations, that deserves criticism as well. The goal shouldn’t be to protect a consensus. Nor should it be to overthrow one. The goal should be to determine what the evidence actually allows us to conclude and build from that.