➘ #16 Source Article
Symbolic Logic Formalization
(Let 
We define the salient evidential data as a vector 
- Here,
represents persistent ambiguity or low specificity (for instance, Isaiah 7:14 or Matthew 24).
denotes weak pre-event timing integrity or textual signs of redaction (e.g., Isaiah 45:1 or Daniel 11).
captures manipulability or staging potential (such as Zechariah 9:9, where a triumphal entry is arranged to fit the prophecy).
reflects clustered accuracy up to the author’s historical horizon but subsequent failure (as in Daniel 11).
stands for non-uniqueness across religions and genres, reflecting human universals of mythmaking.
Annotation: This vector specifies the main observed features of prophecy, broken into categories that capture ambiguity, timing, staging, horizon-limited accuracy, and universality.
Under likelihoodist reasoning, the evidential weight is captured by the likelihood ratio (also called the Bayes factor):
Annotation: This means the total evidential weight is the product of the likelihood ratios for each observed feature, comparing how probable each feature would be if prophecies came from God versus from human retrofit.
Assigning conservative probabilities tilted in favor of 
Annotation: Ambiguity is far less expected under divine prophecy than under human retrofit, so this heavily disfavors
Annotation: Weak timing integrity fits the human retrofit hypothesis far better.
Annotation: The ease of staging fulfillment is expected under human behavior but not under divine design.
Annotation: Horizon-limited clustered accuracy is highly indicative of human postdiction.
Annotation: Cross-cultural universality in prophecy types is much more expected under human authorship.
Multiplying these together, we obtain:
Annotation: This result means that the evidence is about 0.6% as likely under divine authorship as it is under human retrofit. In other words, the reverse Bayes factor is approximately 180, strongly favoring
Case-level examples reinforce this imbalance:
- In Daniel 11, accuracy persists only until the author’s own time, then collapses. Formally,
.
- In Isaiah 45:1, Cyrus is named retrospectively, so
.
Annotation: Both examples show that what would be unlikely under divine foresight is entirely expected under human postdiction.
Finally, under Bayes’ theorem, posterior odds are updated as follows:
Annotation: Even with neutral priors, multiplying by the large Bayes factor in favor of
◉ Clarification
When we write 
So 
This quantity asks: “How much more (or less) likely is the evidence 
By contrast, 
This version asks the flipped question: “How much more (or less) likely is the evidence if human retrofit is true, compared to if divine authorship is true?”
In our worked calculation:
. This means the evidence is only about 0.6% as likely under divine authorship as it is under human retrofit. In plain terms, divine authorship does a very poor job of predicting what we actually see in biblical prophecy.
. This means the evidence is about 180 times more likely under human retrofit than under divine authorship. In plain terms, human authorship overwhelmingly fits the evidence better.
To visualize this, think of it as a scoreboard:
- The divine-authorship side scores about 0.6 points.
- The human-retrofit side scores 180 points.
The lopsided result shows that the features of biblical prophecy (ambiguity, staging, postdiction, horizon-limited accuracy, etc.) overwhelmingly match what we would expect from human communities adapting texts, not from a God providing unambiguous foresight.
Note: The numerical probabilities used in this analysis are not meant to be precise measurements, as if one could calculate prophecy-likelihood down to decimal points. Rather, they serve as conservative estimates deliberately tilted in favor of 
◉ A plain English walkthrough of the symbolic logic above.
We are comparing two hypotheses:
- H₁: Biblical prophecies were authored by God with the intent of providing supernatural evidence.
- H₂: Biblical prophecies were written and later interpreted by humans, often through postdiction or retrofit.
To test which fits the evidence better, we look at the main features of prophecies and ask: “Are these features more expected if God authored them, or if humans did?”
The features we focus on are:
- Ambiguity: Many prophecies are vague or non-specific (e.g., Isaiah 7:14, Matthew 24).
- Timing problems: Some show signs of being written after the fact, or have timing mistakes (e.g., Daniel 11, Isaiah 45:1).
- Manipulability: Some prophecies can be staged or fulfilled by human planning (e.g., Jesus arranging the donkey ride in Zechariah 9:9).
- Clustered accuracy with horizon limits: Prophecies sometimes appear accurate up to the writer’s own time, but fail beyond it (e.g., Daniel 11 suddenly going wrong after Antiochus IV).
- Non-uniqueness: Similar prophecy styles exist in other religions and mythologies, suggesting human universality.
We then estimate: If prophecy really came from God (H₁), how likely are these features? And if prophecy came from humans (H₂), how likely are they?
- For ambiguity, under H₁ it should be rare (≈15%), but under H₂ it is common (≈60%). So the evidence is only about one-quarter as likely under H₁ as under H₂.
- For timing problems, under H₁ we’d expect accuracy (≈20%), but under H₂ we’d expect flaws (≈60%). So again, the evidence is about one-third as likely under H₁.
- For manipulability, under H₁ we’d expect no staged fulfillment (≈10%), but under H₂ we’d expect it often (≈80%). The ratio is about one-eighth as likely under H₁.
- For clustered accuracy with horizon limits, under H₁ this shouldn’t happen (≈15%), but under H₂ it’s expected (≈70%). So only about one-fifth as likely under H₁.
- For non-uniqueness, under H₁ prophecy should stand out as unique (≈20%), but under H₂ common overlap with other traditions is expected (≈80%). Ratio is about one-quarter.
When we multiply these together, we find that overall:
- The observed pattern is only about 0.6% as likely under divine authorship as it is under human retrofit.
- Put the other way around, it is about 180 times more likely under human authorship than under divine authorship.
Worked examples confirm this:
- In Daniel 11, accuracy continues only up to the writer’s lifetime, then fails — exactly what we’d expect from human knowledge, not divine foresight.
- In Isaiah 45:1, Cyrus is named retrospectively after his rise — fitting postdiction, not genuine prediction.
Conclusion: Even if we start neutral, the massive tilt of the evidence toward human retrofit means the posterior judgment lands overwhelmingly in favor of H₂. The shape of biblical prophecy matches what we would expect from human literary practices, not from a God attempting to provide clear, evidential prophecies.
◉ Flowing Narrative Summary
When we examine biblical prophecy through a careful comparative lens, two competing explanations emerge. The first, which believers often assume, is that prophecy was authored by God as a way of providing supernatural evidence of His existence and plan. The second, more naturalistic explanation, is that prophecy is a human product, shaped by literary creativity, retroactive fitting, and cultural patterns of mythmaking.
To assess these explanations, we can ask: what features do the prophecies actually display, and which explanation makes those features more probable? What we find is that prophecies are typically vague, riddled with ambiguity, and often require interpretive stretching before they can be linked to later events. They also reveal serious timing issues, such as signs of having been written after the events they describe or predicting accurately only up to the writer’s own historical horizon before collapsing into error. Many prophecies are staged or manipulated so that the narrative conforms to a pre-existing script—Jesus’ arranged donkey ride into Jerusalem being a textbook example. Moreover, prophecy as a genre is not unique to the Bible; nearly every religious tradition has its share of symbolic forecasts and retrospective fulfillments.
If prophecy were divinely authored, we would expect something quite different: precise predictions that are unambiguous, immune to manipulation, and consistently accurate well beyond the writer’s own time. Instead, what we actually observe fits far more naturally under the hypothesis of human authorship. The statistical weight of the evidence, when modeled carefully, indicates that the observed features of prophecy are roughly 180 times more likely if prophecy is human rather than divine in origin.
Case studies reinforce this imbalance. Daniel 11 remains accurate only until the events of the author’s own day, then collapses into failed forecasting—an outcome fully consistent with human guesswork but hardly with divine foresight. Similarly, Isaiah 45:1 names Cyrus retrospectively, not as a supernatural vision of the future but as a retroactive insertion once the identity of the Persian king was already known.
When taken together, these patterns paint a consistent picture: prophecy looks less like the voice of an omniscient God and more like the handiwork of human authors attempting to make sense of their world, often retrofitting their texts to align with later developments. The evidential arrow points firmly in one direction: the Bible’s prophecies are best explained not as divine revelations, but as literary constructions born of human ingenuity and cultural convention.
Free of Faith 
Leave a comment