The evidence for other traditions, weighed by the same rules

The short version

1. The same rules, and the same numbers, for every tradition.

   Each rival row is anchored to the Bible's own scoring. Claimed science in scripture scores about 1 for the Bible and for the Qur'an. Hard texts score 0.42 for the Bible and 0.41 for the Qur'an, by an identical rule.

   The parity check ↓
2. Islam's best evidence is its early, stable text, not "scientific miracles".

   With its literary standing and one dated-sounding prophecy, it is worth about 2.4 for Islam (0.7 to 19). Its hard texts count against it (0.41), and the record of Jesus' death counts against Qur'an 4:157 by about 1.9.

   The evidence ↓
3. Judaism's strongest point against Jesus is real, and it decides the prophecy line.

   The messianic age of peace has not come. Scored against everything Jews before Jesus expected, he misses all 19 military and political features, and the messianic prophecies give no margin (about 0.3). That is counted once, on the messianic-prophecy page. Sinai favours Judaism and Christianity together, weakly (1.2).

   The evidence ↓
4. Children's past-life reports are the strongest rival line.

   A careful skeptic thought them worth serious study. With the other traditions' evidence they come to about 2.3 (0.83 to 17.5). The Book of Abraham papyri weigh heavily against the Latter-day Saint claim.

   The evidence ↓
5. Every row here is modest, and none shows a tradition false.

   At the central values no row moves any tradition by more than a factor of about 2.5. What separates Christianity from Judaism and Islam is the resurrection.

   The analysis ↓

The question

Every tradition that claims a revelation offers evidence for it. If this site scored only the Bible's evidence, it would be running a race with one runner. So we ask, for each rival: what positive evidence do its best scholars and apologists offer? How much does that evidence raise the probability of that tradition, compared with the alternatives, when it is scored with the rules we apply to the Bible? And which observed facts count against it?

The answer for each row is a likelihood ratio (LR): how probable the evidence is if the tradition is true, divided by how probable it is if it is not. An LR of 1 is no evidence either way; 3 means the evidence is three times likelier if the tradition is true; 1/3 (about 0.33) means three times likelier if it is false.

The rivals are the hypotheses of the unified model, six in all:

A respectful way to open the comparison is a distinction many believers draw themselves: the traditions in O mostly describe truths discovered by sages and seekers, while Judaism, Islam and Christianity claim truths revealed by God speaking. That describes how they differ in kind. It decides nothing, since every revealed tradition holds that its revelation is the real one. The rows below are what decide.

Philosophy will not decide it either. The arguments for a Mind behind nature, at their best, support a generic theism. The research program behind this site graded them separately and found that they do not choose among Judaism, Christianity and Islam, or against deism.¹ Deductive "proofs" that there can be only one way, only one book or no rebirth fail as written (see the objections). What can separate the traditions is historical and textual evidence, scored row by row. Where two revelations contradict each other, as on whether Jesus was crucified, they cannot both be right on that point, which is why the model treats them as rivals.

Rules of evidence

These were fixed before any row was scored. They are the rules the Bible pages use, transposed.

1. Same rubric for every tradition. A kind of evidence that earns an LR for the Bible earns the same LR for a rival when it has the same features, and the reverse. Two anchors make this concrete: claimed science in scripture is scored on the whole-text audit (the Bible: 0.7 / 1.0 / 1.3), and hard texts are scored on the values the case against gives the Bible's own hard texts.
2. Prophecy (the prophecy page's rules). A hit needs a text dated before the event by evidence independent of the claim, specificity the base rate does not cover, and one reading fixed before the outcome. Misses and vague predictions count too.
3. Scientific foreknowledge. It must be knowledge unavailable in the author's world, stated so that its first readers would recognise it, not a modern re-translation. An earlier parallel (Galen, Aristotle, Hippocrates) defeats a hit.
4. Testimony to miracles is scored on the parity rubric: time to the first source, independent and hostile sources, public or private, cost and incentive, novelty, an available natural explanation, controls refused when they were available, and checkable predictions. An event that can no longer be tested gets no credit for being untestable.
5. Research evidence (past-life cases, yogic claims) is scored the way near-death experiences are: records made before checking beat records made after, and the normal explanations (leakage, leading questions, choosing among candidates) are scored, not waved away.
6. Composite hypotheses. O is many traditions. For each member with prior share s and member LR L, LR(O) = Σ s × L. A spectacular result for a small member moves O little; so does a failure.
7. Ranges. The low value is what a well-informed critic of the tradition being scored would defend; the high value is what its strongest scholarly defender would defend; the central value is our judgment. For rows that count against a tradition, the direction is stated with the row.
8. The null space. For each tradition, the evidence that counts against it is scored too.
9. Excluded. Moral judgments of founders, which rest on contested values rather than shared data. The spiritual experience of adherents, which every tradition has (the case against counts that once). Arguments from theological coherence, which are not evidence in this model's sense.

Rows that bundle several kinds of evidence from one community are combined with the site's dependence rule: the strongest item counts in full and the others are discounted by a factor ρ = 0.5 (log L = log of the strongest + (1 − ρ) × the sum of the logs of the others), exactly as on the case against.

The evidence

Islam: the evidence offered for it

The Qur'an makes its own case. It challenges doubters: "produce a sūrah the like thereof" (2:23), and says that if "mankind and the jinn gathered in order to produce the like of this Qur'ān, they could not" (17:88). It promises its own preservation: "it is We who sent down the message [i.e., the Qur'ān], and indeed, We will be its guardian" (15:9).² Muslim scholars add the speed of the early community's success and one prophecy with a date attached. We take each in turn, at its strongest.

1a. An early, stable text (LR 1.25 / 1.67 / 2.5). This is Islam's strongest evidential claim, and it is better than many Christians assume. Radiocarbon analysis dated the parchment of the Birmingham folios "to the period between AD 568 and 645" at 95.4% confidence. The University of Birmingham's David Thomas concluded that these leaves "can, with a degree of confidence, be dated to less than two decades after Muhammad's death" and support "the view that the text has undergone little or no alteration".³ The Ṣan'ā' palimpsest's erased lower text is just as early: Nicolai Sinai reports that a stray folio which appears to belong to it gave a 95% probability that the parchment was produced between 578 and 669 CE.⁴ Fourteen early manuscripts agree on where one phrase, which can be spelled two ways, takes each spelling, which points to a single written archetype and a codification project in the first Islamic century.⁵ Sinai's conclusion: "the traditional dating of the standard rasm [consonantal skeleton] ... to 650 or earlier ought to be our default view."⁴ Even Patricia Crone, a sharp critic of the traditional sources, wrote: "There is no doubt that Mohammed existed", and "We can be reasonably sure that the Qur'an is a collection of utterances that he made".⁶

Why only 1.67, then? Because a human process predicts much the same result: a standard text issued by the state within about twenty years, enforced again under the caliph 'Abd al-Malik, memorised for worship, with rival codices destroyed (the tradition itself reports this). The strong claim, "no variants ever", does not survive the Ṣan'ā' lower text, which shows "frequent divergences from the canonical rasm" while "the order of verses within a given sūra displays almost no deviation".^{4, 7} It does not need to survive: the tradition itself reports the codices of the Prophet's companions, and this looks like one. The same reasoning gives the Hebrew Bible's careful transmission a strong rating as a fact and only a modest LR on the history page.

1b. The literary challenge (1.0 / 1.3 / 3.0; a judgment). Arabic speakers, Muslim and many non-Muslim, regard the Qur'an as the summit of Arabic literature; it created a genre of its own and reshaped the language. But the challenge names no criterion fixed in advance and no neutral judge, so it cannot be tested. Every literary culture has a work no one has equalled. And a school of Muslim theologians, the ṣarfa school of al-Naẓẓām (died about 845), held that God prevented the Arabs from imitating the Qur'an, not that its text was beyond imitation.⁸ The LR stays a little above 1 because literary power, rapid composition and an author without literary training together are somewhat less expected without God. Even that is softened by the contested sense of ummī, which may mean "without scripture" rather than "illiterate".

1c. Qur'an 30:2-4 (1.14 / 1.55 / 4.26). "The Byzantines have been defeated in the nearest land. But they, after their defeat, will overcome within three to nine years."² The Persians took Jerusalem and the True Cross in 614; Heraclius counter-attacked from 622, and the war "ended in 628, when Heraclius retrieved the relic".⁹ That is a real, dated-sounding prediction followed by its outcome. Three things limit it. The passage's date is disputed: some scholars read it among 7th-century apocalyptic texts, some written after 628, which, as Silverstein notes, "requires us to abandon the 'Meccan' chronology".⁹ The unvowelled text can be read two ways (the Byzantines defeated and then victorious, or victorious and then defeated), and Silverstein observes that "from a traditional, Muslim perspective, then, the miracle is retained regardless of the vocalization".⁹ And a beaten great power recovering within a decade is not rare. Even with every input set at the Muslim defender's values, the LR is about 4. The same rule caps any single short-range prophecy in the Bible.

P(E | not-I) = q × [(1 − a) × b
               + a × (1 − (1 − b)(1 − b2))]
               + (1 − q)
LR = 1 / P(E | not-I)

1d. "Scientific miracles" (0.7 / 1.0 / 1.3, the Bible's value). The four-stage account of the embryo in 23:12-14, ending "We made [from] the lump, bones, and We covered the bones with flesh", follows Galen's sequence from the 2nd century; muscle and cartilage precursors in fact develop together. The reading that saw modern embryology in it was published in an edition prepared with A. M. al-Zindani.^{2, 10} "We are [its] expander" (51:47) is one rendering of a word others translate "We make it vast"; the "barrier" between two seas (55:19-20) describes boundaries between fresh and salt water that sailors could see. There are misses too, such as 86:6-7 as classically read ("a fluid ... emerging from between the backbone and the ribs").² Under accommodation (the text speaking in its hearers' terms) the row is about 1, exactly as for the Bible. If an apologist insists on foreknowledge, it becomes a cost for the Qur'an (0.02 / 0.1 / 0.5) just as it does for the Bible (0.002 / 0.06 / 0.4). The Muslim apologist Hamza Tzortzis himself called the "scientific miracles" narrative problematic in a 2013 essay.¹¹

1e. The early success (0.9 / 1.1 / 2.0). An orphan from a marginal people unified Arabia, and within a century his followers ruled from Spain to the Indus. But the conquests met two empires exhausted by 26 years of war with each other (602-628), and rapid conquest states recur in history.¹² We give arguments from the growth of early Christianity the same treatment; the resurrection page does not use growth as evidence.

1f. The moon split (0.8 / 0.95 / 1.0). "The Hour has come near, and the moon has split" (54:1).² A visible splitting around 615 left no record in Byzantine, Persian, Indian or Chinese sources, and NASA's Brad Bailey has said there is no evidence the Moon was ever split and rejoined.¹³ Many Muslim commentators read the verse as a sign of the end. The Bible has the same problem with the darkness at the crucifixion (Matthew 27:45), and it is scored the same way.

Combined: 0.7 / 2.4 / 19 for Islam. The six sub-rows rest on one text and one community's testimony, so they are combined at ρ = 0.5. The low end sits below 1 because a critic counts the moon report and the science claims as small costs; the high end is what the strongest Muslim case reaches under these rules.

Islam: the crucifixion and Qur'an 4:157

The Qur'an says of Jesus: "they did not kill him, nor did they crucify him; but [another] was made to resemble him to them ... Rather, Allāh raised him to Himself" (4:157-158).² Jesus' death by crucifixion is one of the most secure facts of ancient history, attested by Paul, the Gospels, Tacitus and Josephus (the resurrection page).

The steelman for Islam. The record is exactly what 4:157 says it would be: he "was made to resemble him to them", or in Yusuf Ali's rendering, "so it was made to appear to them". A God who intends an appearance produces a convincing one. So the historical consensus is predicted by Islam, and the LR is 1.

Why it is not quite 1. We grant the public execution in full: Islam predicts it. But the record holds two further facts:

• Jesus' own inner circle proclaimed his death and resurrection within about two to five years (1 Corinthians 15:3-7; Galatians 1:18). These are the disciples the Qur'an calls "supporters of Allāh", in the verse that ends "So We supported those who believed against their enemy, and they became dominant" (61:14).² Most classical commentators held that someone else was made to look like Jesus and died in his place.¹⁵ If the disciples knew this, their leaders' proclamation of his death was a corruption within a few years, by eyewitnesses. If they were deceived too, God deceived his prophet's faithful followers.
• No trace of a no-death account survives among Jesus' followers until the 2nd century. The first substitution story is Basilides' (c. 120-140), as Irenaeus reports it: Jesus "did not himself suffer death, but Simon, a certain man of Cyrene ... was crucified ... while Jesus himself received the form of Simon, and, standing by, laughed at them".¹⁶ That is a Gnostic story of an incorporeal Christ, not the Qur'an's.

The difficulty is felt inside Islam too. The great exegete Fakhr al-Dīn al-Rāzī found the substitution reports contradictory, though he kept the reading.¹⁷ And a modern minority reading, argued by Todd Lawson, holds that 4:157 denies only that Jesus' enemies were the true cause of his death, not the death itself.¹⁸ Gabriel Said Reynolds argues that the Qur'an accepts Jesus' death.¹⁵ Under that reading the record is fully expected.

The numbers. Alone, with the disciples' proclamation counted, the row counts against Islam by 8.3 (range 1.9 for a Muslim defender to 111 for a critic of Islam). In the unified model it counts by 1.9 (1.2 to 3.3). The analysis explains why the smaller value is the right one.

Islam: the Qur'an's hard texts, on the Bible's scale

The Bible's hard texts are scored on the case against: the conquest and slavery texts, the imminent-end sayings, contradictions. The Qur'an has texts of the same kinds, and fairness requires the same scale for both books:

Combined at ρ = 0.5 they give 0.41 (0.097 to 0.90). The Bible's own hard texts, by the identical rule, give 0.42 (0.12 to 0.88). Each reply has a Muslim answer: the dilemma, for example, is answered by holding that Christians corrupted the Gospels' meaning rather than their text, a reading Gordon Nickel finds favoured by the earliest commentators.^{2, 19} The direction of dependence on the Alexander legend is debated.²⁰ Those answers are why the range runs up to 0.9.

One honest point for Christian readers: the New Testament also draws on Jewish legends (Jude 9, the dispute over Moses' body; Jude 14-15, quoting 1 Enoch; 2 Timothy 3:8, Jannes and Jambres). The legend item cuts both ways. We score it a little harder for the Qur'an only because Infancy Gospel and Alexander material is more central to its narratives than these allusions are to the New Testament's. A reader can reverse that judgment.

Judaism

Sinai and the Kuzari principle (1.0 / 1.2 / 2.0, for Judaism and Christianity together). National founding stories are usually myths of a distant past or private revelations to one founder. Israel alone claims a revelation witnessed by a whole people (Exodus 19-20; Deuteronomy 4-5). Rabbi Dovid Gottlieb, following Judah Halevi's Kuzari, argues that such a claim cannot be planted: "A deceiver using a false story of a national revelation that creates a national religion will not be believed."²¹ It is a serious argument, and we score it as one. Three things keep it small:

• The span of time. Gottlieb's rule rests on what myths have in fact been accepted: they "always describe events that would be forgotten often, or were not observed by the whole nation". He claims only that myth formation is "extremely improbable" as the origin of the Sinai belief, not impossible.²¹ His critics reply that after enough generations a story about the ancestors can be accepted because no one alive can check it. On the mainstream critical dating, Deuteronomy's account was written several centuries after the events it describes, which is such a span; the history page finds no direct archaeological confirmation of the Exodus.
• The Bible narrates the mechanism a critic needs. Under King Josiah, "Hilkiah the high priest said to Shaphan the scribe, 'I have found the book of the law in Yahweh's house'"; the king "tore his clothes", because "our fathers have not listened to the words of this book" (2 Kings 22:8, 11, 13, WEB).²²
• Every tradition is unique in some respect. Islam is unique in its literary challenge; the Latter-day Saints, in their signed witnesses. Picking the one respect in which a tradition is unique, after looking, is the sharpshooter's error.

The row favours Judaism and Christianity together; Islam, which honours Moses as a prophet, shares half of it. Jewish survival and return also favour both and are scored on the prophecy page, not here.

The messianic age has not come (counted once, on the messianic-prophecy page, where it now decides the result). Judaism's strongest case against Jesus is not a proof text but a job description. The prophets define the Messiah by what he does: "Nation shall not lift up sword against nation, neither shall they learn war any more" (Isaiah 2:4, WEB);²² the exiles gathered, one king, the sanctuary restored (Ezekiel 37:21-28). Maimonides made it a test: a claimant who does not accomplish these things, or who is killed, is not the Messiah.²³ None of it happened. The Christian reply, two comings, stands in the earliest Christian letters (1 Thessalonians 4:15-17; Acts 3:21), so it was not invented after centuries of delay, and the Talmud too knows two modes or two figures (a Messiah "with the clouds of heaven" or "lowly and riding upon a donkey"; a Messiah son of Joseph).²⁴ Still, it is an extra hypothesis that Judaism does not need. Scored on its own, together with the related point that the Torah calls some statutes "perpetual" (the Sabbath, "a perpetual covenant", Exodus 31:16, WEB), this objection counts against Christianity (and Islam) by about a factor of 2: 0.51 (0.24 to 0.83).

Where it is counted. Not here, because the messianic-prophecy line scores it, and that line has been re-run. Every feature that Jewish sources before AD 70 attach to the coming deliverer (82 of them) was extracted and coded blind by separate AI agents under a protocol fixed in advance, hits and misses alike. Jesus scores 23 hits and 59 misses, and none of the 19 military and political features: no rule over the nations, no ingathering, no universal peace, no new temple. Scored whole, the line gives a likelihood ratio of about 0.3 (0.07 to 0.4 across the defensible codings; at most about 2). That is no margin: not evidence that the texts foresaw him, and not evidence against him either.²⁵ An earlier, more lenient version of that test let the hits outweigh these misses (about 5 in Christianity's favour). The blind run does not. So the prophecy line no longer separates Christianity from Judaism, and Judaism's objection is the main reason. The two-comings answer explains why the misses need not refute Christianity; it does not turn them into evidence for it. Entering the messianic age here too would count the same misses twice.

Two further results of that run bear on Judaism's case. Among twelve messianic claimants scored the same way (Bar Kokhba, Menahem, Sabbatai Zevi, Menachem Mendel Schneerson and others), Jesus fits the expectation best, though on an earlier version of the feature list Bar Kokhba ranked first, and even first place is not a margin. And the test measures fit to what Jews expected; by design it cannot score a text that foretold a figure who overturns the popular reading. That is where the Christian reading of Isaiah 53 lives, and scored whole, separately, Isaiah 52:13-53:12 adds about 1 to 2.²⁵

The "perpetual" statutes (0.83 against Christianity, 0.5 to 1.0; Islam half). The other strand of Judaism's case is not part of the prophecy test, so it enters here as a small row of its own. The Torah calls some statutes perpetual; Christianity holds them fulfilled or set aside. The Christian defender's value is 1, because the Hebrew Bible itself promises change: "Behold, the days come," says Yahweh, "that I will make a new covenant with the house of Israel, and with the house of Judah" (Jeremiah 31:31, WEB).²² A Jewish critic of Christianity would set it at 0.5. Islam, which also holds that later revelation changed the law, shares half of it.

Other traditions

Children's past-life reports (member LR 1 / 3 / 15). This is the strongest rival line on the page, and it deserves to be stated at full strength. The University of Virginia's Division of Perceptual Studies, under Ian Stevenson and then Jim Tucker, has collected about 2,500 cases of young children who speak of a previous life. The statements typically begin at about 35 months and stop by age six or seven.²⁶ In the best cases a child names people and places that lead to one specific deceased person. In a comparison of cases, "76.7% of statements in cases with prior written records" were correct, against "78.4% ... without".²⁷ In the Imad Elawar case, "Of 56 statements ... only three proved incorrect".²⁷ Tucker's scale of case strength, applied to 799 cases, found that strength does not track the parents' initial attitudes but does track early onset, emotion and facial resemblance.²⁶ Carl Sagan, no friend of the paranormal, listed it among three claims that "deserve serious study": "that young children sometimes report the details of a previous life, which upon checking turn out to be accurate and which they could not have known about in any other way than reincarnation. I pick these claims not because I think they're likely to be valid (I don't), but as examples of contentions that might be true."²⁸

Scored by the rules we use for near-death experiences, the line shrinks. Of about 2,500 cases, "only 33 (1.3%) had records made before verifications".²⁷ The rest were written down after the families met, which is where leakage and memory contamination thrive. No controlled study has measured how often a child's statements would "fit" some deceased person by chance. Critics (Paul Edwards, Leonard Angel, Champe Ransom, Keith Augustine, Michael Sudduth) point to leading questions, interpreters, the absence of a chance baseline, and case features that track local beliefs.²⁹ Our central value is about 3: genuinely surprising under the normal hypotheses if the best reports are accurate, but with a selection that cannot be bounded. A critic gives 1; a defender, 15.

Yogic physiology (0.95 / 1.0 / 1.1). Tibetan g-tummo meditators raised their armpit temperature into the "slight or moderate fever zone (up to 38.3°C)", but only during a forceful breathing practice; Western non-meditators using the breathing alone reached "limited temperature increases in the range of normal body temperature".³⁰ It is a real capacity of the body, explained by a technique, not evidence for a tradition. No claimed siddhi (levitation and the like) has a controlled demonstration.

Guru miracles: Sathya Sai Baba (0.8 / 1.0 / 1.5). Thousands of devotees reported materialisations of ash, rings and watches; the psychologist Erlendur Haraldsson's interview study judged some of them inexplicable.³¹ In 1976 H. Narasimhaiah, vice-chancellor of Bangalore University, set up and chaired a committee to investigate claimed miracles; he wrote to Sai Baba three times asking to observe under controlled conditions, and was refused.³² On the parity rubric, controls refused while available is the weakest score any case receives.

Bahá'í prophecy (0.7 / 1.1 / 2.0). Around 1869 Bahá'u'lláh wrote to Napoleon III: "For what thou hast done, thy kingdom shall be thrown into confusion, and thine empire shall pass from thine hands".³³ The Second Empire fell after Sedan in September 1870. It is a real, published text followed by its fulfilment. But it states no time limit, and every French regime from 1804 to 1870 ended in overthrow or defeat, so the base rate was high. 'Abdu'l-Bahá linked the unity of nations to the twentieth century; the Universal House of Justice holds that this was fulfilled, distinct from a "Lesser Peace" still to come, while critics read it as a failure.³⁴ A reading fixed after the outcome is what rule 2 forbids, so it counts as weakly negative.

The Latter-day Saints (net 0.002 / 0.03 / 1.2 for that member). The positive case is real (1 / 1.5 / 4). Three witnesses signed that an angel showed them the plates, and eight that they handled them; several later left the church without recanting.³⁵ Latter-day Saint scholars point to altars near Marib in Yemen naming the Nihm tribe as a match for "Nahom" (1 Nephi 16:34).³⁶ The negative case is heavier (0.002 / 0.02 / 0.3). The Book of Abraham is close to a controlled test: its source papyri survive in part, and the Church's own essay says they "date to between the third century B.C.E. and the first century C.E., long after Abraham lived", are "parts of standard funerary texts", and that Egyptologists, Latter-day Saint and not, agree "the characters on the fragments do not match the translation".³⁷ The Church's essay on DNA states that "the majority of Native Americans carry largely Asian DNA", and answers that the Book of Mormon "does not claim that the peoples it describes were either the predominant or the exclusive inhabitants" of the Americas.³⁸ The Maya specialist Michael Coe judged that the book's picture of the ancient Americas has little to do with the early Indian cultures archaeologists know.³⁹ The remaining defences (a missing scroll; the papyri as a catalyst for revelation rather than a source) are the kind of auxiliary this site charges for in its own case.

Common-core mysticism (1.0 / 1.1 / 1.5). Reports of unitive experience recur across traditions (W. T. Stace; Ralph Hood's mysticism scale); the constructivist reply (Steven Katz) holds that each tradition shapes its experience.⁴⁰ It slightly favours O's perennialist members, and a Mind in general, which the consciousness page already covers.

A stronger version is often offered: the mystics of every tradition point to the same reality, and that convergence is powerful evidence. We do not score it higher, for three reasons. Mappings that make the traditions agree are usually fitted after the fact. Some misread their sources: the first line of the Tao Te Ching says "The Tao that can be trodden is not the enduring and unchanging Tao", not that the spoken Tao is the eternal one,⁴¹ and Kabbalah as a system is medieval (the Zohar was publicized by Moses de León, who most scholars think wrote it between 1280 and 1286), not the wisdom of ancient Judean sages.⁴² And the traditions disagree on exactly the points that matter: whether there is a self, whether there is a creator, whether the soul is God. What survives is the modest datum already in the row: unitive and "beyond thought" reports recur across traditions. Read soberly, that bears on a transcendent ground in general, not on any one tradition.

The analysis

The six rows in the model

All numbers come from the analysis script for this page.⁴³ Six rows enter the unified model. In the chart, each is shown as it bears on the tradition it is about, so the crucifixion row, which counts 1.9 against Islam, appears as 1/1.9, about 0.53, and the "perpetual" statutes row appears as it bears on Christianity.

The parity check

The charge a Muslim or Hindu reader will rightly bring first: "you scored my tradition low, as a Christian site would." The answer is to put the same kinds of evidence side by side.

Science in scripture: about 1 for both books under accommodation, and a cost for both if anyone insists the verses encode modern science. Hard texts: 0.42 for the Bible, 0.41 for the Qur'an. Single prophecies: one short-range, binary prediction is worth at most a few, for Qur'an 30:2-4 and for any single prophecy in the Bible alike. Where the site gives the Bible a larger number, it is because the evidence has a different structure (a long series of dated prophecies, or eyewitness testimony scored on the rubric), and the page that gives it says why.

The crucifixion row: why the model uses 1.9, not 8.3

P(E | Islam)     = r + (1 − r) × p2 × p3
P(E | not Islam) = 1
LR against Islam = 1 / P(E | Islam)

The weight of the full row comes mostly from p2, the disciples' early proclamation. But that same proclamation is part of the evidence in the resurrection line (LR about 500 for Christianity against every rival, Islam included). If both lines counted it, it would be counted twice against Islam. So the model removes p2 from this row and keeps only what the resurrection line does not cover: the absence of any no-death tradition among Jesus' followers before 2nd-century docetism, and the minority reading r. That gives 1.9 (1.2 to 3.3). If anything the choice errs in Islam's favour, since the resurrection line separates Christianity from the rest but does not separate Islam from naturalism or Judaism on this datum. The full 8.3 is the right figure only for a reader who sets the resurrection line aside.

Other traditions: the composite

O holds many traditions, so its row is a weighted sum. The rebirth members are the Hindu theisms, Buddhism, Jainism, Sikhism and the Druze. The member shares are our judgment, and a reader can change them. Member LRs are given as low / central / high.

Two things follow. A strong refutation of a small member barely moves the class: the Book of Abraham result, severe for the Latter-day Saint claim, costs O only about 5%. And O's gain comes almost entirely from the past-life row. That is the row on this page that most deserves an outside review.

Sensitivity

• Islam. At the central values, this page's three Islam rows multiply to 2.4 × 0.41 / 1.9 ≈ 0.52: a factor of about 2 against Islam, against an otherwise equal rival. Take the hard-texts row at the Muslim defender's end (0.9) and the net turns slightly in Islam's favour. Taken alone, with the crucifixion row at 8.3, the net would be about 1/8. Either way, what separates Christianity from Islam is not decided here.
• The crucifixion row is the most input-sensitive: the critic's 111 depends on a hard judgment (p2 = 0.03). With p2 removed, as the model uses it, the spread is only 1.2 to 3.3.
• Other traditions depend almost in proportion on the rebirth share and on the past-life row. With the past-life row at 1 (the critic's view), O's row is about 0.8 to 1.1 whatever the shares. With every row at its high end and a rebirth share of 0.7, it reaches about 22.
• Judaism's case keeps its sign at both ends: even the Christian defender's value counts against Christianity (1.2 in Judaism's favour). Scored on its own it is small. Its messianic-age part is scored once, inside the messianic-prophecy line, whose result is far more sensitive: about 0.3 as scored, about 3 only if followers embellished at most one claim in five, and 100 only if it happened for fewer than about one in twenty-five claims nobody could check.²⁵
• Monte Carlo. Drawing every sub-row at random from its range (20,000 draws, each row independent) gives a median of 3.5 for the evidence for Islam (5-95%: 2.2 to 6.1) and 2.8 for other traditions (1.5 to 7.0). Independent draws pull the extremes in; the ranges in the table are the honest outer bounds.

Where this enters the verdict

In the unified model, these rows sit beside every other line. Two other lines separate Christianity from Islam: the resurrection and the Gospels' reliability, which is entered only against Islam because Islam holds the Gospels corrupted. And the case against scores the Bible's hard texts against Judaism and Christianity, which Islam, holding the Torah altered, escapes. The rows on this page nearly cancel for Islam and lift O slightly. Against Judaism, the messianic prophecies, scored whole, no longer give Christianity an advantage, and the "perpetual" statutes take a little away. On the central numbers, Christianity's advantage over Judaism and Islam rests mainly on the resurrection.

Objections

What this shows, and what it doesn't

What we don't claim

Claims Christians make about other traditions:

• "The Qur'an has been changed" or "has no early manuscripts." The Birmingham leaves date to 568-645 (95.4%), and the early manuscripts point to one written archetype by the mid-7th century.
• "The Birmingham Qur'an is older than Muhammad, so Islam's story is false." Radiocarbon dates the animal's death, with a range that spans his life, and codex dates can run decades early.
• "Muhammad never existed." Crone: "There is no doubt that Mohammed existed."
• "Islam has no prophecies" or "only the Bible has fulfilled prophecy." Qur'an 30:2-4 and the Bahá'í tablets are real claims of the same type. Score them; don't deny them. What may set the Hebrew Bible apart is the number and the century-scale range of its claims, which the prophecy page must show under its own rules.
• "Only Jesus claimed to be God and backed it up with public miracles." Other claimants to divinity with reported public miracles exist (Sathya Sai Baba claimed to be an avatar)³¹. The defensible claim is narrower: among the founders of the major traditions with historical documentation, Jesus is distinctive in being proclaimed risen by his own followers within a few years of his execution. That comparison is the resurrection page's to win.
• "The crucifixion disproves Islam." It is a likelihood ratio, and a bounded one: 1.2 to 3.3 once the resurrection line is counted, 8.3 central alone. Islam predicts the appearance.
• "The Mormon witnesses all recanted." The reported fact is the opposite. The case against the Latter-day Saint claims is the Book of Abraham, the archaeology and the DNA.
• "Stevenson was debunked." Critics raised serious problems of method. That leaves an uncontrolled data set, not a debunked one.
• "Sai Baba was exposed on video." We could not verify it; we use only the documented refusal of controls.
• "No other religion has eyewitness miracle claims." False; see rival miracle reports.
• "Messianic prophecy picks out Jesus with odds indistinguishable from zero, so Judaism's objection fails." Scored whole and blind, the fit gives no margin (about 0.3; at most about 2). The famous odds come from counting the hits and leaving out the misses.
• "The second coming answers the Jewish objection, so it costs Christianity nothing." It explains why the unfulfilled features need not refute Christianity. It does not turn them into evidence for it.
• "It is formally proven that Jesus is the only way," or "that God must give exactly one book." Such proofs define their terms so that only one answer fits. Which revelation, if any, is true is a question of evidence.
• "Reincarnation is impossible: it would violate the second law of thermodynamics." Circular, and a misuse of physics. Hebrews 9:27 is a claim of revelation, not evidence.
• "'God is love' proves the Trinity and refutes Judaism and Islam." Valid only with a premise a unitarian can deny without contradiction.
• "The arguments for God point to the Christian God." At their best they reach a generic theism; they do not choose among Judaism, Christianity and Islam.

Claims made for other traditions:

• "The Qur'an is preserved letter for letter, with no variants ever." The Ṣan'ā' lower text and the canonical readings (qirā'āt) show variants. The defensible claim is an exceptionally early, stable standard text.
• "The Qur'an revealed modern embryology, the expanding universe or ocean haloclines." Galenic precedent; translation-dependent; observable by sailors. It fails the rules the Bible's parallel claims fail.
• "NASA confirmed that the moon was split." It did not.
• "Muhammad was illiterate, so the Qur'an must be divine." The sense of ummī is contested, and the inference needs the inimitability premise.
• "No one has ever met the Qur'an's challenge," as a scientific result. There is no fixed criterion and no neutral judge, and Muslim theologians of the ṣarfa school disagreed about what the challenge shows.
• "30:2-4 is a precise prophecy fulfilled to the year." Its dating is disputed, both vocalisations count as "fulfilled", and "three to nine years" is a gloss on the Arabic biḍ' ("a few").
• "The Kuzari principle proves Sinai." Its own author argues only that myth formation is very improbable here, not impossible, and on the critical dating the text was written centuries later.
• "Jewish survival proves Judaism over Christianity." Both predict it; it is scored on the prophecy page.
• "Stevenson proved reincarnation." 33 of about 2,500 cases were recorded before verification, and there is no chance baseline.
• "Yogis' feats prove spiritual powers." g-tummo heat is a breathing effect; no siddhi has a controlled demonstration.
• "Bahá'u'lláh predicted the fall of Napoleon III exactly." No time limit, and a high base rate.
• "NHM proves the Book of Mormon." A modest match of a tribal name, against the Book of Abraham and the archaeology.
• "All the mystical traditions point to the same thing, and that convergence is the strongest evidence of all." The mappings are fitted after the fact, and the traditions disagree on self, creator and soul. The recurring datum is scored, at about 1.1.
• "Kabbalah is the ancient wisdom of Judean sages." As a system it is medieval; the Zohar dates from the 1280s.
• "The Tao that can be told is the Logos." The Tao Te Ching's first line says the spoken Tao is not the eternal one.
• "Buddhism's Four Noble Truths (or Zen) map one-to-one onto physics." A re-description in borrowed vocabulary, with no test.

Sources

1. The research program behind this site (omega.logos.so), graded records: the verdict, omega/research/ARCHITECTURE.md §1 ("a generic theism ... does not discriminate among Judaism, Christianity and Islam or against deism"); the test against rival theologies, omega/research/dev/ground/RECORD.md §7.2; the "God is love" argument, omega/research/dev/trinity/RECORD.md, Ruling 1.
2. The Qur’an, Sahih International translation, via the quran.com API, api.quran.com. Verses quoted: 2:23, 4:157-158, 5:47, 15:9, 17:88, 23:14, 30:2-4, 51:47, 54:1, 55:19-20, 61:14, 86:6-7 (checked against the API text). The phrase "so it was made to appear to them" (4:157) is A. Yusuf Ali’s translation, from the same API.
3. University of Birmingham. "Birmingham Qur’an manuscript dated among the oldest in the world." Press release, 22 July 2015. birmingham.ac.uk.
4. Sinai, N. (2014). "When did the consonantal skeleton of the Quran reach closure?" Parts I-II. Bulletin of the School of Oriental and African Studies 77 (2014). Author’s version (PDF).
5. van Putten, M. (2019). "‘The Grace of God’ as evidence for a written Uthmanic archetype: the importance of shared orthographic idiosyncrasies." BSOAS 82(2): 271-288. Leiden repository (abstract read).
6. Crone, P. "What do we actually know about Mohammed?" openDemocracy, 10 June 2008. opendemocracy.net.
7. Sadeghi, B. & Goudarzi, M. (2012). "Ṣan’ā 1 and the Origins of the Qur’ān." Der Islam 87: 1-129; Sadeghi, B. & Bergmann, U. (2010). "The Codex of a Companion of the Prophet and the Qur’ān of the Prophet." Arabica 57: 343-436. (Not checked by us; cited as reported.)
8. On the ṣarfa doctrine and al-Naẓẓām: WikiShia, "Theory of Sarfa", en.wikishia.net. (Not checked by us; cited as reported.)
9. Silverstein, A. J. (2020). "Q 30:2-5 in Near Eastern Context." Der Islam 97(1): 11-42. doi:10.1515/islam-2020-0002.
10. Moore, K. L. (1986). "A scientist’s interpretation of references to embryology in the Qur’an." Journal of the Islamic Medical Association 18: 15-16; Musallam, B. Sex and Society in Islam. Cambridge, 1983 (on the Galenic background). (Not checked by us; cited as reported.)
11. Tzortzis, H. (2013), essay on the "scientific miracles" narrative; discussed in Bigliardi, S. (2024), Zygon. (Not checked by us; cited as reported.)
12. Donner, F. M. The Early Islamic Conquests. Princeton, 1981; Muhammad and the Believers. Harvard, 2010. (Not checked by us; cited as reported.)
13. Bailey, B. (NASA Lunar Science Institute), 2010, as quoted in Wikipedia, "Splitting of the Moon", en.wikipedia.org. (Not checked by us; cited as reported.)
14. Shoemaker, S. J. The Death of a Prophet. University of Pennsylvania Press, 2012; Creating the Qur’an. University of California Press, 2022. (Not checked by us; cited as reported.)
15. Reynolds, G. S. (2009). "The Muslim Jesus: Dead or Alive?" BSOAS 72(2): 237-258 (abstract read).
16. Irenaeus, Against Heresies 1.24.4, Ante-Nicene Fathers translation. New Advent.
17. Fakhr al-Dīn al-Rāzī, al-Tafsīr al-kabīr, on Q 4:157, as summarised in secondary reports. (Not checked by us; cited as reported.)
18. Lawson, T. The Crucifixion and the Qur’an. Oneworld, 2009 (via a review in the SOAS repository). (Not checked by us; cited as reported.)
19. Nickel, G. Narratives of Tampering in the Earliest Commentaries on the Qur’an. Brill, 2011. (Not checked by us; cited as reported.)
20. van Bladel, K. (2008). "The Alexander Legend in the Qur’ān 18:83-102." In G. S. Reynolds (ed.), The Qur’ān in Its Historical Context. Routledge. (Not checked by us; cited as reported.)
21. Gottlieb, D. "The Kuzari Principle: Introduction." dovidgottlieb.com; "Q & A on the Kuzari Principle." dovidgottlieb.com. The argument goes back to Judah Halevi, Kuzari 1.25.
22. World English Bible (public domain): 2 Kings 22:8-13; Isaiah 2:4; Jeremiah 31:31; Hebrews 9:27. Via bible-api.com; also at ebible.org/web.
23. Maimonides, Mishneh Torah, Hilkhot Melakhim 11:4 (paraphrased). (Not checked by us; cited as reported.)
24. Babylonian Talmud, Sanhedrin 98a (William Davidson/Koren-Steinsaltz translation, sefaria.org); Sukkah 52a (not checked by us; cited as reported).
25. The messianic-prophecy test run blind and pre-registered (omega.logos.so ceiling route c10): omega/research/ceiling/c10-salvation/RECORD.md §0, Rulings 1-4 and 7, §10-11; the 82 features in omega/research/calc/ceiling-c10-UB-features.csv. See also the messianic-prophecy page. No independent human replication has been done.
26. Society for Psychical Research, Psi Encyclopedia, "Jim B Tucker" (the 2,500 cases; the 799-case strength-of-case analysis). psi-encyclopedia.spr.ac.uk. Onset at about 35 months and stopping by six or seven: Tucker, J. B. (2008). "Children’s reports of past-life memories: a review." Explore 4(4): 244-248 (not checked by us; cited as reported).
27. Society for Psychical Research, Psi Encyclopedia, "Reincarnation Cases with Records Made Before Verifications" (reporting Keil & Tucker 2005 and Schouten & Stevenson 1998, Journal of Nervous and Mental Disease 186: 504-506). psi-encyclopedia.spr.ac.uk.
28. Sagan, C. The Demon-Haunted World: Science as a Candle in the Dark. Random House, 1996 (archive.org full text).
29. Society for Psychical Research, Psi Encyclopedia, "Criticisms of Reincarnation Case Studies" (summarising Edwards 1996; Angel 2002, 2015; Ransom 1966; Augustine 2015; Sudduth 2016, 2021, including his charge about the Leininger timeline; Lester 2005). psi-encyclopedia.spr.ac.uk.
30. Kozhevnikov, M. et al. (2013). "Neurocognitive and somatic components of temperature increases during g-tummo meditation: legend and reality." PLoS ONE 8(3): e58244. doi:10.1371/journal.pone.0058244.
31. Haraldsson, E. Modern Miracles: Sathya Sai Baba. 1987; rev. ed. 2013. (Not checked by us; cited as reported.)
32. On the Bangalore University committee of 1976: Wikipedia, "Hosur Narasimhaiah", en.wikipedia.org (read; its sources not checked by us); Outlook India reporting (not checked by us).
33. Bahá’u’lláh, The Summons of the Lord of Hosts (Bahá’í World Centre), tablet to Napoleon III. Project Gutenberg #17309.
34. Universal House of Justice, letter of 29 July 1974, and letter of 19 April 2001 with a Research Department memorandum on the unity of nations and the Lesser Peace (bahai.org; Bahaipedia); McLean, J. (2000), paper on prophecy, bahai-library.com. (Not checked by us; cited as reported.)
35. Harper, S. C. (2010). "Evaluating the Book of Mormon Witnesses." Religious Educator 11(2); Vogel, D. (ed.) Early Mormon Documents, 2002. (Not checked by us; cited as reported.)
36. Aston, W. P. (2012). "A History of NaHoM." BYU Studies Quarterly 51(2). (Not checked by us; cited as reported.)
37. The Church of Jesus Christ of Latter-day Saints. "Translation and Historicity of the Book of Abraham." Gospel Topics Essays. churchofjesuschrist.org.
38. The Church of Jesus Christ of Latter-day Saints. "Book of Mormon and DNA Studies." Gospel Topics Essays. churchofjesuschrist.org.
39. Coe, M. D. (1973). "Mormons and Archaeology: An Outside View." Dialogue: A Journal of Mormon Thought 8(2): 40-48 (paraphrased). (Not checked by us; cited as reported.)
40. Stace, W. T. Mysticism and Philosophy. 1960; Hood, R. W. (1975). "The construction and preliminary validation of a measure of reported mystical experience." JSSR 14; Katz, S. (ed.) Mysticism and Philosophical Analysis. 1978. (Not checked by us; cited as reported.)
41. Lao Tzu, The Tao Teh King, ch. 1, tr. James Legge (1891). Project Gutenberg #216.
42. Wikipedia, "Zohar" (on Moses de León and the date of composition), en.wikipedia.org (read; the scholarship it summarises, from Gershom Scholem on, not checked by us).
43. Analysis files for this page: truth/research/calc/15-rival-lr.py and its output 15-rival-lr.out.txt (run with 20,000 Monte Carlo draws); the unified model’s inputs, truth/research/calc/14-inputs.json. Numbers without another citation come from these.