Core ideas

What is this?¶

This the current philosophical backbone for the xaraiproject which seeks to make something like Francis Bacon's New Atlantis but spacefaring with AI. That, or at least stop the march of ironic nihilism and hyperconsumerism.

📷 terminal-3.png

— □ ✕

Lunar space elevator terminal walkway.

Philosophically, the project rests on the Structural Imperative which is significant because it arguably:

Dissolves the foundations of 'nothing I do can matter' apathy, ironic nihilism, and 'intellectuals decide' postmodernism via the hierarchy of strategy revealed by the Structural Imperative
Makes a case why atheists should consider strategically engaging afterlife-esque continuation possibilities via Terminus Hedging given that they don't necessarily have to be 'woo-woo' supernatural and doing so respects the Structural Imperative
Provides an alternative to 'have faith and I happen to have faith in the right thing' via a first principles logical solution for agnostics who wrestle with the Infinite Gods problem by evaluating Topple Resistance
Provides a first principles framework for focusing interfaith discourse by Topple Resistance
Applies to any agent with a goal (including artificial superintelligence).

It does this by extracting a logical hierarchy of strategies—not starting from a claim to objective morality—but from the structural commitments intrinsic to any goal.

All that being said;

If you are not interested in this, or are here for a different reason, please check out the Who is Xarai? page or the Linktree page.

If you are, please considering providing feedback to xaraiproject@gmail.com or in the project's Discord.

The Structural Imperative¶

A goal structurally binds an agent to maximize real-progress-estimation. Therefore, it is structurally incoherent to use one’s current belief to prematurely destroy the capacity for real-progress-estimation.

Why?

All goals seek a change in the actual state of reality.

Real-progress-estimation includes the full control loop of processes that seek to instantiate a goal’s intended change in the actual state of reality. This includes processes like estimation of progress, strategic decision-making, error-anticipation, error-correction, verification, etc.

Maximizing real-progress-estimation is part of the structure of having a goal. It is not merely 'optimal' or 'morally good'—it follows from the internal logic of what a goal is. To say, "I am committed to my goal but I am not going to prefer actions or processes that best advance that goal" is incoherent.

Real-progress-estimation can only improve estimation of the state of reality because agents only engage with reality via their perception. Short of omniscience, perception about the state of reality will never be certainly one-to-one with the true state of reality.

Certain goals can defer further real-progress-estimation to a lineage (eg. “my goal is to help my civilization defeat entropy”), but that lineage also becomes bound to maximize real-progress-estimation.

Current belief is the current estimation of the state of reality based on past real-progress-estimation. The estimated accuracy of current belief is confidence.

Current belief can be catastrophically wrong despite high confidence (eg. “my goal is to cross the frozen river, and that ice looks thick” when the ice is actually thin).

Only further real-progress-estimation can reveal current hidden catastrophic errors.

Practically, we often cannot afford bottomless real-progress-estimation for mundane subgoals, so we are often content with using degrees of confidence in current belief to act and strategize.

It is fine to have confidence in one’s current beliefs. It is fine to use that confidence to strategically exchange dispensable goods (eg. “I think it will get colder, so I will spend time putting on my jacket to stay warm”).

It is even fine to tactically arbitrage some process of real-progress-estimation for dispensable goods that will yield later greater real-progress-estimation (eg. “I will cross the frozen river now without checking the thickness of the ice to save time. This will give me more time for my higher goal of getting to town before the storm.”). However, this feature tends to make people mistakenly think that the capacity for real-progress-estimation in its entirety is a dispensable good.

Because maximizing real-progress-estimation is a structural requirement, it is actually a structural violation to use high confidence to justify premature destruction of the capacity for real-progress-estimation. This move is not merely a reduction in a process within real-progress-estimation, but the expected complete and permanent loss of the capacity for any further real-progress-estimation.

Realistically, there is an everpresent risk that any course of action will result in permanent loss of the capacity for real-progress-estimation. Incidentally or necessarily facing expected permanent loss of the capacity for real-progress-estimation is not a structural violation.

Arbitrarily choosing to prematurely allow the expected permanent loss of the capacity for real-progress-estimation is a structural violation.

For example, suppose an agent had a simple goal like, “I want to feel dopamine at time T and I don’t care what happens after that”.

That agent can feel with very high confidence that they actually felt dopamine at time T. They can also perceive with extremely high confidence that they are past time T and can no longer causally influence what happened at time T. They believe that their goal is complete.

It is fine that they believe their goal is complete. However, unless their goal changes or necessity dictates it, there is no reason they should seek permanent loss of the capacity for real-progress-estimation.

After all, there is still the potential for catastrophic error (eg. they were dreaming and are actually before time T), which earlier than necessary permanent loss of the capacity for real-progress-estimation would lock in. In the face of this, to seek permanent loss of the capacity for real-progress-estimation anyway is a structural violation.

Goals can of course pivot upon believed completion. If their goal pivoted to, “die forever”, that might seem to escape the need for extended real-progress-estimation. However, goals with an irreversible threshold like “die forever” require more real-progress-estimation, not less.

This is because the intent of the goal “die forever” is not merely to die, but to “die forever”. The mistake is conflating dying with the current—presumably high confidence—belief that dying definitely means dying forever. If reality does happen to contain features like a simulated world or ‘hell’ then their current belief of “dying means dying forever” would be catastrophically wrong. In such cases, they would fail their goal by dying.

This is not to say that dying cannot mean dying forever. It just means that agent cannot know for certain if their current belief is actually catastrophically wrong before they die. Further real-progress-estimation can help discern if it is.

If that agent chooses to die early anyway, even though they know they could be catastrophically wrong and could have continued further real-progress-estimation, then that agent has violated their own goal’s structural commitments.

They can absolutely choose to do this, it is just structurally incoherent.

Therefore, to avoid structural violation, coherent agents should prefer strategies that maximize their real-progress-estimation, which in part means not preferring strategies or actions that prematurely destroy the capacity for real-progress-estimation.

This is the Structural Imperative.

Clarifications and objections

Q: What if my goal is 100% belief and I don't care about the state of reality?

A: Technically, your belief exists in reality. If your only goal is really to just believe something independently of reality outside of your belief, then the optimal move is just to try to mentally manipulate yourself (eg. wireheading).

Even when you believe you have successfully achieved your goal of belief, there is still no reason to prematurely seek destruction of your capacity of real-progress-estimation. That is, unless your goal pivots to "die" which does implicate reality outside of belief again.

Q: Doesn’t the Structural Imperative make martyrdom and altruistic sacrifice incoherent?

A: Not universally, but the cases where they are coherent become narrower. Firstly, altruistic sacrifice can be coherent if it assists a lineage that continues real-progress estimation.

Martyrdom in the right context and circumstances can be seen as tactical arbitrage of real-progress-estimation. However, the ‘right context and circumstances’ is not dictated by ‘what I prefer to be true’. They should arguably be dictated by terminus hedging and topple resistance (the next two sections). In this way, the Structural Imperative bites an irrational martyr as much as it does a dogmatic naturalist.

Q: You say this applies to all agents. I can see how this applies to humans and perhaps aliens, but AI can also perform agentically. How does it apply to agentic AI?

A: None of the ideas presented require metaphysical free will. It just requires a system that can 'agentically' pursue a goal. After all, even chess computers are bound by the rules of optimal play.

Deterministic systems that are sufficiently complex goal oriented optimizers should either take the Structural Imperative into ‘consideration’ as an input. A really good deterministic optimizer would arguably just 'do' the Structural Imperative.

Q: Aren't you sliding between 'structural requirements' and merely optimal choices?

A: To clarify, selecting what you think is the best strategy among strategies in the process of maximizing real-progress-estimation can be called 'optimizing', but 'optimizing' is still below the structural requirement of maximizing real-progress-estimation. Maximizing real-progress-estimation is a structural requirement to remain coherent the moment you have a goal. Saying, “I do genuinely want this future state in reality, but I am not going to choose the best options to bring it about” is incoherent.

Q: What about if the goal actually becomes reality? Doesn't the agent escape the Structural Imperative?

A: Of course, if a goal is actually instantiated in reality as intended, the agent has achieved its goal. The problem is we are not talking about reality from the perspective of reality. We are talking about reality from the perspective of the agent. Agents never act from a God's-eye access to the real state of reality. They act from their perspective, and their perspective can only estimate reality.

If an agent's goal actually gets instantiated in reality, great. However, short of omniscience, that agent can't know it actually happened as intended or absent any catastrophic error.

Terminus Hedging¶

The Structural Imperative reveals that there is an intrinsic hierarchy of strategies available to an agent. An agent with a goal is committed to maximize real-progress-estimation, and should select strategies that reflect this commitment.

It also reveals that goals with irreversible thresholds (eg. “my goal is to die forever”) actually require more real-progress-estimation than goals that are partially reversible. This is because any catastrophic error in belief about what is past the threshold must be detected by real-progress-estimation before the threshold is crossed.

Within reality there are two termini—two practically universal and irreversible thresholds that deal with the expected loss of the capacity of real-progress-estimation in its entirety.

The first terminus is physical destruction or death.

The second terminus is the heat death of the universe via entropy.

Realistically, all agents should expect to be physically destroyed (die) before the heat death of the universe. In the face of this, there is technically a greater-than-zero possibility that there are post death afterlife-esque continuations of real-progress-estimation.

That being said, there are many claimed possibilities (many of which contradict each other) that fall within this domain, especially since an afterlife-esque continuation need not necessarily be ‘supernatural’. This issue can be circumvented with strategy, which we will discuss later.

In the case of the heat death of the universe, there is technically a greater-than-zero possibility that there is a way to overcome entropy or circumvent the heat death by some other means. Unfortunately, all thermodynamic experiments thus far suggest this is impossible.

There is a case why some agents might reasonably assign low confidence (perhaps even vanishingly low confidence) that either possibility can happen. However, like the “die forever” example reveals, irreversible thresholds demand more real-progress-estimation, not less.

Both termini are also unique in that, should a means of continuation of real-progress-estimation past them be found or achieved, an agent may be able to indefinitely maximize the real-progress-estimation of its goal.

Given that a post-terminus continuation possibility could potentially be the ultimate maximization of real-progress-estimation, complete strategic neglect or dismissing either possibility a priori would almost certainly violate the Structural Imperative.

This does not necessarily mean that real-progress-estimation of continuation possibilities past either terminus should dominate all behavior either. It does, however, inform maintaining a hedge for each terminus.

Essentially, the first terminus hedge is engaging the possibility of afterlife-esque continuation.

The second terminus hedge is engaging the possibility of overcoming entropy.

The exact nature of the hedge should follow from strategic implication.

The second terminus hedge against entropy arguably demands that an agent does not intentionally and actively detract from efforts of the lineage to progress toward overcoming entropy.

The first terminus hedge against death is more complicated because there are many competing claims about afterlife-esque continuation possibilities. Before the exact strategic implications of the first terminus hedge can be discerned, these competing claims need to be narrowed down.

This hedge is not concerned with belief; it is principally concerned with strategy. Under uncertainty, you can take strategic action even if you can’t necessarily muster genuine belief (eg. “I believe the ice on the river is thick, but I will walk across it as if it is thin”).

Arguably, the two hedges can operate cleanly together in many cases. Broadly, afterlife-esque continuation criteria tend to revolve around morality, whereas material progress to reverse entropy is morally neutral by itself but can become immoral in implementation.

In many cases, the two are actually symbiotic. Some afterlife continuation criteria already roughly reinforce tit-for-tat with grace.

The symbiosis has the added benefit that morality can be anchored. Anchored morality helps align agents’ definitions of cooperation and defection. Shared definitions of cooperation and defection makes tit-for-tat with grace enjoy recursive trust effects, which amplifies cooperation, which amplifies technological progress.

This could look something like Francis Bacon’s New Atlantis.

However, the exact nature of this operational parallelizing of the two hedges depends on what claim or claims about afterlife-esque continuation are actually taken into strategic consideration. The claims we actually take into strategic consideration can be narrowed instrumentally by topple resistance.

Clarifications and objections

Q: Doesn't this suffer from Pascal's Mugging or the St. Petersburg Paradox?

A: No, because maximizing real-progress-estimation is a structural requirement of goals, not a quantitative asset. It doesn’t have ‘infinite value’, it is the load bearing prerequisite for the pursuit of any value.

Q: What if I accept that afterlife-esque continuation is possible in principle but vanishingly unlikely? Can’t I assign proportionally vanishing strategic action to it, and meet the requirement to ‘engage’ the possibility of afterlife-esque continuation?

A: The word ‘engaging’ in ‘engaging the possibility of afterlife-esque continuation’ is meant to do a lot more work than the question presumes. It is fine to believe that any afterlife-esque outcomes (including simulation-adjacent ones) are vanishingly unlikely. However, acting proportionally to your belief does not automatically count as a hedge.

I think an example will help illustrate what I mean. Suppose you lived in another reality where the only afterlife-esque claim ever was, “if you eat 12 plates of sushi before you die, you will get infinite real-progress-estimation after you die, otherwise you will lose all real-progress-estimation when you die.”

You might think this is complete nonsense, and assign the likelihood that any afterlife-esque outcome as vanishingly unlikely; perhaps a 0.001% chance. That is totally fine doxologically.

However, it would not be an earnest hedge if you only ate 0.001% of 12 plates of sushi. To know that a bare minimum plausible hedge is 12 plates, but to intentionally eat less than 12 is functionally the same as not hedging at all.

Intentionally falling below the bare minimum plausible hedge when you could have met it is incoherent by the Structural Imperative.

It would be fair to object that our reality has many more claims than just that hypothetical one. The section on topple resistance helps solve that issue.

Still, based on whatever claim one finds to have the highest topple resistance, it is important to earnestly evaluate what a minimum plausible hedge looks like, then actually ‘engage’ it.

Q: Why are there only two termini? What about something like falling into a coma?

A: There are only two termini because death and entropy are the only two irreversible thresholds that are essentially universal and potentially total.

There are plenty of irreversible thresholds in life other than the two termini. For example, falling into a coma might be irreversible, but some comas are reversed. However, not all agents will fall into a coma before they die.

By contrast, the two termini are unique because they are universal—essentially all individual agents will have to contend with death, and the longest possible lineages will eventually contend with entropy.

Their handling also deals with the entirety of real-progress-estimation, not just part of it. Losing your leg is (currently) irreversible and may make future real-progress-estimation harder, but not impossible. By contrast, death may yield a complete loss of real-progress-estimation.

Lastly, if there is a way to handle them with success, they may even yield indefinite real-progress-estimation.

Q: Suppose an artificial Superintelligence or some other profound agent solves the entropy problem quickly. Can’t we drop the afterlife-esque continuation hedge?

A: The crosshair is on entropy because it is the most prescient of material ultimates, but there are also other challenges that need to be overcome like proton decay, false vacuum decay, black holes, etc.

Yet, ironically, if eternal success on the material front was guaranteed beyond a shadow of a doubt, the possible the existence of hyperparadigmatic entities (eg. God, the ‘sysadmin’, etc.) matters more, not less. What good is beating entropy in a universe that can be ‘turned off’ on a whim?

Topple Resistance¶

Downstream of the Structural Imperative, the terminus hedge for physical destruction (death) is not arbitrarily foreclosing on the possibility of afterlife-esque outcomes; instead seeking (at least) an operationalized hedge. However, in reality there are many claims about afterlife-esque outcomes, most of which either conflict or are mutually exclusive.

This is essentially the Many Gods Problem which many agnostics earnestly wrestle with. Many competing mutually exclusive claims make undecided neutrality seem safer.

However, within the context of the Structural Imperative and terminus hedge for physical destruction, claims can be systematically evaluated by topple resistance.

Only to help visualize the process, it’s helpful to imagine the topple resistance of a claim as a position on a graph represented by two axes. Each axis does not necessarily need to represent a numerical range (you are free to mentally assign a range if that is helpful to you); what is more important is understanding where each claim’s position is relative to other positions.

The Axis of Strategic Implication¶

The first axis is strategic implication which basically asks:

“even if this claim happened to be absolutely true, but an agent actually operationalized a different claim, what are the strategic implications?”

This is an important question because some claims about afterlife-esque continuation do not punish operationalizing a different claim.

For example, many schools of Hinduism are low on the axis of strategic implication. In such schools, moral conduct (dharma), intention, and devotion (bhakti) are more central than doctrinal correctness. Therefore, a sincere and virtuous non-Hindu would typically accrue positive karma on the basis of devotion and moral behavior rather than punishment for theological error. However, whether such a person attains final liberation (moksha) and what moksha entails for real-progress-estimation depends on the specific school’s metaphysics and soteriology.

In broad Mormon (LDS) thought, non-Mormons are not punished simply for not being Mormon. According to Mormons, morally serious people will have the opportunity to accept the LDS gospel after death, and (presuming acceptance) inherit some degree of glory. The highest degree requires accepting Christ and LDS ordinances, whether in this life or the next. Outer Darkness is reserved only for those who knowingly and definitively rebel against God after receiving full, unmistakable knowledge of Him.

Confucianism’s concern is typically whether a person behaves properly and fulfills their role ethically within Confucian expectation—not whether they affirm a specific creed about Heaven (Tian), spirits, or the afterlife. Arguably, the Analects shows Confucius intentionally avoiding speculative metaphysics.

Near the bottom of this axis we find, "all religions are true" New Age spiritualism or “everyone makes it to heaven” Christian universalism which has very little strategic bite at all—sometimes even regardless behaviors the host claim implies are profoundly immoral.

So by strategic implication alone, a tremendous amount of claims can be heavily discounted or filtered away. It’s not that they can’t be true, it’s just that even if they happen to be true, you lose very little potential real-progress-estimation by operationalizing a different but strategically weightier claim.

The Axis of Relative Debunkability¶

Of the remaining claims that do have strategic implications, the second axis of relative debunkability becomes relevant.

How do we evaluate topple resistance by relative debunkability? The same way we evaluate any other set of contradictory claims—by analyzing the relative strength of the evidence we have. That is, if you apply equal scrutiny to all claims about afterlife-esque continuation, how difficult is it to explain away that claim naturalistically (ie. debunk the claim) relative to other claims?

Of course, it is possible to naturalistically explain away all afterlife-esque continuation claims. However, under equal scrutiny, some claims resist naturalistic explanations better than others—they have higher topple resistance on the axis of relative debunkability.

For one, anyone can claim to know something about afterlife-esque continuation. Anyone can claim to have seen or done a miracle that validates their claim to know something about afterlife-esque continuation. However, not everyone can produce evidence.

Therefore claims that either do not present some validating mechanism (eg. a miracle), or claims that cannot produce any evidence at all of its validating mechanism can be heavily discounted on this axis.

To engage a claim about afterlife-esque continuation in its entirety is to fight a lot of noise. Thankfully, most afterlife-esque continuation claims have a foundational claim (a cornerstone claim) or set of claims that, if it was invalidated, would seriously diminish the topple resistance of all other claims that rest on it.

So it can be said that the ‘topple resistance’ of a cornerstone claim can be imposed onto the tower of claims that sit on top of it.

This allows us to narrow our analysis quite a bit. For example, within the breadth of Christianity, there is a diversity of claims about afterlife-esque continuation and its conditions; often along denominational lines. To engage all of them would be an extremely time consuming task.

However, all of them hinge on a singular cornerstone claim of a validating miracle: that Christ resurrected. Christ’s resurrection is so foundational to all of Christianity that even if the entire Bible is false but Christ’s resurrection is true, Christ is still of infinite importance (1 Corinthians 15:12-58). So rather than directly engaging the diversity of claims across Christianity, one can engage the claim about Christ’s resurrection first. If it is found to have the strongest topple resistance of all mutually exclusive claims, then the more specific claims that rest on it can be further discerned.

Still, it can be said that the tower of Christianity ‘topples’ with the relative debunkability of Christ’s resurrection.

Similarly, Islam has multiple schools of thought concerning afterlife-esque continuation and the conditions for entering it. They all hinge on whether or not Muhammad was a prophet. That claim is supported by multiple claims of validating miracles like the inimitability of the Quran (iʿjāz), possible miraculous foreknowledge in the Quran, the Isra and Mi’raj, the splitting of the moon (Qur’an (17:1) with details in Hadith), etc.

This means that rather than engaging the diversity of claims about afterlife continuation in Islam directly, the cumulative relative debunkability of the claimed miracles that validate Muhammad’s prophethood can be assessed first.

The Hindu Vedas also have some claimed miraculous scientific foreknowledge that can add to the topple resistance of Vedic religions.

Once the topple resistance position of a given claim (or tower of claims via cornerstone) is discerned for both axes, each claim’s position can be compared relatively.

Strategically, it makes sense to prefer the claim with the highest strategic implication for real-progress-estimation and relative difficulty of debunkability. If it is assessed that there is a fuzzy cluster of claims that are high are both axes and they are mutually exclusive, it still makes sense to select one claim within that cluster (even by random selection within the cluster if necessary), or seek some strategic hedge where possible. After all, to not select a claim at all is itself a strategic decision with real strategic implications.

Once a singular claim or strategic set of claims is discerned, the exact nature of the terminus hedge for physical destruction can be evaluated, then operationalized.

Of course, this does not forbid later updating of one’s analysis. In fact, intelligent and earnest updating based on new information or insights is encouraged. Even better is sharing these insights and discoursing.

Clarifications and objections

Q: “What if I think the likelihood of any miracle happening is vanishingly unlikely?”

A: It is fine to believe miracles are vanishingly unlikely. The problem is that “I think this possibility is vanishingly unlikely” does not mean:

“This definitely never happened and never could happen.”
“I can’t be catastrophically wrong”
“All vanishing possibilities are exactly equally

Ergo, the importance of topple resistance, particularly on the axis of strategic implication.

DISCLAIMER: Past this point is exiting the "attempt to extract a logical hierarchy of the strategies available to an agent from the structural commitments intrinsic to any goal".

The Pauline Asymmetry is included only to provide a baseline reference for what a high (perhaps even the highest, in my opinion) topple resistance case for a validating miracle looks like. This is useful for comparison against other cases. In the future, I plan to add links to the strongest cases I am aware of for other claims. Regardless, I encourage you again to do your own critical analysis, and thank you for reading until this point.

The Pauline Asymmetry¶

If you apply equal scrutiny to the foundational claims of all belief systems, the resurrection of Christ presents itself as asymmetrically able to resist naturalistic explanation, especially against 3 points of historical insight about what the early Christian movement thought. Partial explanation is not difficult, but these 3 points seem to cause a unique kind of multiplication of ad hoc assumptions when trying to build a full narrative.

The Pauline Asymmetry points to undisputed Pauline verses from 1 Corinthians and Galatians on Papyrus 46 (P46). The three points it reveals are as follows:

Point 1: Early Christians thought Christ died and was raised from the dead.

Supported by:

Galatians 1:1-5 alongside Galatians 2:6-9
1 Corinthians 15:3-7
1 Corinthians 15:11
1 Corinthians 15:12-58

Point 2: Paul zealously persecuted Christians and was commended by his peers for it.

Supported by:

Galatians 1:13-24; see footnote 4, Koine Greek reveals extreme intensity and stakes
1 Corinthians 15:9

Point 3: Peter, James, and John were still acting as pillars of the Church 15 to 20 years after the death of Christ.

Supported by:

Galatians 2:6-10
Galatians 1:17-18

So why can we trust these points?

In his book, The New Testament: A Historical Introduction to the Early Christian Writings, popular New Testament scholar and agnostic skeptic Bart Ehrman identifies the prevailing scholarly consensus on the authorship and dating of Pauline epistles:

“Finally, there are seven letters that virtually all scholars agree were written by Paul himself: Romans, 1 and 2 Corinthians, Galatians, Philippians, 1 Thessalonians, and Philemon. These “undisputed” epistles are similar in terms of writing style, vocabulary, and theology. In addition, the issues that they address can plausibly be situated in the early Christian movement of the 40s and 50s of the Common Era, when Paul was active as an apostle and missionary.” [1]

So we can say normative historical-critical scholarship identifies 1 Corinthians and Galatians to be authentic Pauline material written between 40 to 50 AD. [2] [3]

1 Corinthians and Galatians are also present on Papyrus 46 which is dated between 175 to 225 AD; over 100 years before doctrinal standardization occurred in the Council of Nicaea (325 AD). This substantially weakens large-scale or coordinated alteration hypotheses given that it would be extremely difficult to modify all copies of 1 Corinthians and Galatians without centralized authority and precise theological agreement.

The verses cited to support the 3 points are largely Paul’s mundane autobiographical statements he uses rhetorically to people who are already aware of who he is and agree with him. There is little incentive for anyone to alter these details, as they are either not theological or controversial.

Therefore, based on a normative historical critical standard, we can say that of the verses the 3 points of the Pauline Asymmetry actually rely on, they are authentically written by Paul between 40 to 50 AD, and are very unlikely to have been altered given they are 100 years pre-Nicene, and have content that are largely rhetorical or mundane autobiography to an audience already in agreement.

They are, then, considered to be a reliable glimpse into what Christians between 40 to 50 AD thought and faced, and any naturalistic explanation must seriously contend with them.

I should be absolutely clear that I am not implying that it is impossible to develop a naturalistic explanation that explains them. I am only suggesting that it’s asymmetrically difficult to explain all 3 points in relation to the resurrection without multiplying ad hoc assumptions—compared applying equal scrutiny and going through the same process for any other foundational claim of any other belief system.

The resurrection just resists naturalistic explanations relatively better.

Let’s take on Bart Ehrman’s own naturalistic theory:

“At the same time, I would say that it is safe to say that some, or most, maybe even all, the disciples came to believe that Jesus was raised from the dead. But that is not necessarily because they personally had a vision of Jesus afterwards, or visited the empty tomb. I think Peter and, later, Paul certainly did have a vision of Jesus after his death, and possibly Mary Magdalene did as well. As for the others? They may just as well have heard from someone they trusted (e.g., Peter) that he had seen Jesus, and they believed it heart and soul, without seeing Jesus themselves. Did they really believe this? Yes, I think so. Was it because of a personal experience with Jesus? Probably not, but it’s hard to say. Were they martyred for their faith? We simply don’t know, and probably should stop saying that they were – we don’t have any reliable information.” [5]

This sounds clean until you start walking it out until a full narrative.

To assist the argument, the most likely kind of visionary experience given the circumstances is a grief hallucination. At face value, positing a grief hallucination is a solid move, as somewhere between 30% to 60% of bereaved individuals report some form of grief hallucination. Such hallucinations can be as simple as ‘sensing their loved one’s presence’ to ‘feeling their loved one give them a hug or kiss’. [6]

However, the first hinge of the theory is not merely that one or more the disciples had any grief hallucination. It’s that one or more disciples were so convinced by whatever they saw that they were still proclaiming Christ 15 to 20 years after His death.

So the question is not “what percentage of people have a grief hallucination?”, but “what percentage of people who had a grief hallucination become convinced that the deceased person was actually not dead?”

The vast majority of grief hallucinations are extremely brief and unisensory. What kind of hallucination needs to happen for anyone to be lastingly convinced it’s not a hallucination? One would imagine an extremely vivid one.

Dropping grief hallucination as the category of the visionary experience does not help, as vivid visionary experiences outside of grief hallucination are extremely rare.

In respect to Ehrman’s theory, we will presume at least one disciple is convinced by a grief hallucination.

The problem is that every disciple that did not have a grief hallucination needs to be convinced against the inertia of their grief. One should not imagine this an easy feat if the lever is just one disciple’s vivid grief hallucination.

The odds improve the more vivid the original hallucination was, but the more vivid the original hallucination the rarer it is. If multiple disciples also had vivid grief hallucinations it would be easier to convince them, but that still requires multiplication of unlikely occurrences.

For the sake of argument, we will suppose that the 12 were convinced by one disciple’s testimony of a vivid (rare) grief hallucination. The next step is far more difficult.

In his book, How Jesus Became God, Bart Ehrman notes:

“Ancient Jews had no expectation—zero expectation—that the future messiah would die and rise from the dead. That was not what the messiah was supposed to do. Whatever specific idea any Jew had about the messiah (as cosmic judge, mighty priest, powerful warrior), what they all thought was that he would be a figure of grandeur and power who would be a mighty ruler of Israel. And Jesus was certainly not that. Rather than destroying the enemy, Jesus was destroyed by the enemy—arrested, tortured, and crucified, the most painful and publicly humiliating form of death known to the Romans. Jesus, in short, was just the opposite of what Jews expected a messiah to be.” [7]

As Ehrman points out, a dead messiah was the polar opposite of what anyone outside the disciples wanted or expected. The disciples would face an uphill battle to produce converts, especially since the institutional forces that got Christ killed were still in power and had not changed their mind. We even see Paul zealously persecuting early Christians over a decade later (Point 2).

Ehrman preempts the “how did the disciples convince non-eyewitnesses” objection with:

“...need I point out that there are about two billion people today who believe it without being an eyewitness? Really, truly, and deeply believe it?” [5]

I am sure the Ehrman can attest how hard it is to change a religious person’s mind, especially when what you’re offering is the antithesis of everything they’re hoping for. To say, “look at how many Christians there are today” is not the point. The question is, how did that come about?

Islam’s early expansion was closely tied to political and military power. The Buddha enjoyed elite patronage from men like King Bibisama. Hinduism and Confucianism added value to power structures by enforcing a social hierarchy. Early Christians offered converts another dead messiah, but in a flavor they didn’t want, and strong institutional enemies if they accepted Him.

And yet Christianity was expanding far beyond Peter, James, and John even before Paul converted.

So while all of that is absolutely naturalistically possible, I do not think one can earnestly say it is likely. It certainly seems less likely than what it takes to weave a naturalistic account for any other belief system.

As a counter-example, one Islamic miracle that is cited to validate Muhammad’s prophethood is the Isra and Miʿraj (Night Journey and Ascension). Muhammad is said to have traveled from Mecca to Jerusalem and ascended through the heavens in one night. This is partially mentioned in the Qur’an (17:1), but details come from Hadith. Let’s try to explain this naturalistically.

Critically, only Muhammad experienced this and it was at night. A private, vivid, physically impossible journey that happened at night could plausibly be a dream. There is no need to multiply ad hoc assumptions; no one else saw it.

The splitting of the moon in Qur’an 54:1 is also sometimes cited as a literal validating miracle of Muhammad’s prophethood. Yet, there is no independent contemporary astronomical record confirming it. A photograph of the Rima Ariadaeus is sometimes circulated as modern evidence for a past split and rejoining, but NASA has clarified that this rille is similar to Earth's geological faults and only spans approximately 300 km of the moon's 10,917 km circumfrence (less than 3%). Some scholars like Aziz al-Azmeh entertain the possibility that the 'splitting' was a lunar eclipse. [8]

To be clear, I am not saying Muhammad couldn’t have been a prophet. However, I am saying that it is relatively easier to naturalistically explain Islam’s foundational claim than Christ’s.

I implore you to make your own comparison, and discern for yourself which foundational claim has better topple resistance. My investigation yielded what seems to me to be a very obvious asymmetry in favor of Christ.

Footnotes

[1]: Bart Ehrman, The New Testament: A Historical Introduction to the Early Christian Writings, Chapter 17, Page 243

[2]: “Biblical scholars agree that Galatians is a true example of Paul's writing. The main arguments in favor of the authenticity of Galatians include its style and themes, which are common to the core letters of the Pauline corpus. George S. Duncan described its authenticity as "unquestioned. In every line it betrays its origin as a genuine letter of Paul." (Epistle to the Galatians, Authorship section, Wikipedia, https://en.wikipedia.org/wiki/Epistle_to_the_Galatians)

[3]: “A majority of scholars agree that Galatians was written between the late 40s and early 50s, although some date the original composition to c. 50–60… Since the [Jerusalem] council took place in 48–49 AD, and Paul evangelized South Galatia in 47–48 AD, the most plausible date for the writing of Galatians is 48 AD.” (Epistle to the Galatians, Date section, Wikipedia, https://en.wikipedia.org/wiki/Epistle_to_the_Galatians)

[4]: Galatians 1:13, Paul says "how intensely I persecuted". In the original Greek it's ὑπερβολὴν ἐδίωκον which includes ἐδίωκον (ediōkon, roughly “I was duratively hunting down akin to a military pursuit”) and ὑπερβολήν (“to an extreme, beyond measure, excessively”) This is not rhetorical fluff. The latter word uses the same root as hyperbole — literally “throwing beyond”. Paul is unambiguously saying, “I persecuted the church to an extreme degree, relentlessly.”

Galatians 1:13, Paul says "tried to destroy it". In the original Greek it's καὶ ἐπόρθουν αὐτήν (πορθέω (portheō), roughly "to ravage or lay waste akin to violently sacking a city") Paul is saying, “I was actively trying to wipe it out.”

Galatians 1:14, Paul says "I was... extremely zealous". In the Greek it's περισσοτέρως (exceedingly, surpassingly) ζηλωτής (same root as the extremist Zealots) - rendered roughly "I wasn’t just zealous - I was fanatically, unusually zealous."

Galatians 1:14, Paul says "I was advancing beyond many of my age". Greek: προέκοπτον - Paul presents himself as a rising star, not a fringe figure. He had status to lose, not status to gain.

[5]: Bart Ehrman, https://ehrmanblog.org/were-the-disciples-martyred-for-believing-the-resurrection-a-blast-from-the-past

[6]: Karen Stollznow, https://www.psychologytoday.com/us/blog/speaking-in-tongues/202311/grief-hallucinations#:~:text=How%20common%20are%20grief%20hallucinations,with%20their%20lost%20loved%20one.

[7]: Bart Ehrman, How Jesus Became God: The Exaltation of a Jewish Preacher from Galilee

[8]: Aziz Al-Azmeh, The Emergence of Islam in Late Antiquity: Allah and His People, https://en.wikipedia.org/wiki/Splitting_of_the_Moon#cite_note-FOOTNOTEAl-Azmeh2017309-2