Thus Saith the Lord,
Approximately

What this page is asking

A man stands up and says God told me this. He is specific. A date, a name, an outcome. The thing does not happen. He is still, a year later, introduced from the same platform as a prophet, and the people who introduce him are not lying, or cynical, or indifferent to truth. They have a reason. They will tell you what it is if you ask.

The reason has a name in the literature. It is the distinction between infallible prophecy and fallible prophets: the claim that a person may genuinely possess the New Testament gift of prophecy, may genuinely receive something from God, and may still report it inaccurately — because the revelation is perfect and the reporting is human. On that view a miss is not disqualifying. It is what the New Testament expects, which is why it tells congregations to weigh prophecies rather than simply to obey them.

The page tests that. It does not assume it. Testing it required looking for the strongest version of the position being examined, which turns out to be considerably stronger than its critics usually allow, and looking for the evidence that cuts against the thesis, which turns out to be real and to come from unexpected places — Aquinas among them.

What the page concludes, stated up front

The prompt behind this page

Everything below was produced by putting the following brief to a frontier language model with web search, which assembled and verified the sources. It is published so you can run it yourself, change it, sharpen it, or reverse it and see what comes back. If you get a materially different answer, that is information, and worth publishing too.

Write a detailed, historically grounded, biblically rigorous and well-sourced
investigation of modern Christian prophets who claim to receive specific
messages from God, make demonstrably false predictions, and are nevertheless
still regarded as legitimate prophets by their followers and defenders.
Avoid electoral politics and political prophecy entirely. Use non-political
examples only: personal prophecies, healings, dates, disasters, deaths,
revivals, marriages, ministries and claims of supernatural knowledge that can
be independently checked.

Test - do not assume - this thesis:
  In Scripture a prophet's authority rests entirely on the claim that the
  message originates with God rather than with the prophet. Parts of modern
  charismatic Christianity have developed a different category: a person may
  genuinely have the gift of prophecy and genuinely receive revelation while
  still delivering inaccurate prophecies, so failure does not disqualify.
  Investigate whether the distinction between INFALLIBLE PROPHECY and
  FALLIBLE PROPHETS solves this problem biblically or merely relocates it.

Cover, in this order:
1. What the Old and New Testaments mean by prophecy. Work from the texts:
   Deut 13:1-5; 18:15-22; Num 12; Num 22-24; 1 Sam 3; 1 Kgs 13; 1 Kgs 22;
   Jer 23; Jer 28; Ezek 13; Joel 2; Matt 7:15-23; 24:24; Acts 5:1-11;
   11:27-30; 13:1-3; 21:10-14; 1 Cor 12-14; 1 Thess 5:19-22; 1 John 4:1-6;
   Rev 2:20-23. Ask directly: can we identify an accepted biblical prophet
   who repeatedly attributes false statements to God and still functions as
   a legitimate prophet? Search seriously for counterexamples. Distinguish
   a prophet's ordinary speech, opinions, mistakes and sins from statements
   explicitly presented as revelation from God. That distinction is critical.
2. What false prophecy meant in the Old Testament - whether Scripture treats
   inaccurate prophecy as innocent error, imperfect interpretation, immaturity
   in a real gift, human contamination, presumption, deception, or proof that
   God did not send the prophet. Read Deut 18 with serious scholarship, and do
   not argue simplistically from the death penalty to modern church practice;
   the question is the EPISTEMIC STANDARD, not the sanction.
3. Whether New Testament congregational prophecy is different in kind.
   Present the strongest scholarly version of the fallible-prophecy position,
   engaging Wayne Grudem directly and other responsible continuationists.
   Do not caricature them. Examine their exegesis of Acts 21, 1 Cor 13-14,
   1 Cor 14:29, 1 Thess 5:19-22, the apostle/prophet distinction and
   Eph 2:20. Then give the strongest critiques.
4. Agabus as a test case. Compare Acts 21:10-11; 21:27-36; 22:22-30; 28:17.
   Consult the Greek. Was Agabus actually wrong? If so, exactly what was
   wrong and what follows? If not, why does this example fail? Follow the
   text, not denominational preference.
5. Prophecy versus impressions and guidance. Compare "I think God may be
   leading me", "I had an impression", "The Lord showed me", "Thus says the
   Lord", "God revealed that X will happen". Does Scripture treat these as
   the same category? Does modern charismatic language collapse the
   distinction?
6. Trace the history: early church, Montanism, medieval movements, the
   Reformation, the Awakenings, restorationism, Irvingism, Holiness, early
   Pentecostalism, Latter Rain, charismatic renewal, Vineyard, the Kansas
   City Prophets, the Third Wave, modern apostolic/prophetic networks,
   social-media prophecy. Identify when an explicit theology of fallible
   New Testament prophecy was first constructed. The chronology is crucial.
7. Then ask the uncomfortable question: did Christians derive fallible
   prophecy from Scripture and then meet fallible prophets - or meet
   repeated failure and then build a theology that could survive it?
   Determine the order of events. If the theology predates the failures,
   that weakens the thesis. Say so if it does.
8. Document actual failed prophecies from primary sources - books, recorded
   sermons, ministry publications, archived pages, audio and video. Do not
   rely on hostile compilations where the original is available. For each:
   what was claimed; whether it was attributed to God; whether it was
   specific enough to be falsifiable; what happened; whether it genuinely
   failed; how the prophet responded; how followers and defenders responded;
   whether it affected the person's standing. Prefer a few excellent cases
   to many weak ones. Limit to persons who are no longer alive. Tone should
   be civil.
9. The sociology of prophetic failure: reinterpretation, spiritualisation,
   retrospective conditionalisation, blaming the audience, insufficient
   faith, moved timelines, partial fulfilment, redefinition, survivorship
   bias, deletion of the record, treating criticism as persecution. Show
   which are actually documented.
10. The asymmetric standard of evidence: a hit confirms the gift, a miss
    does not count against it. If success confirms but failure cannot
    falsify, what could ever show that someone is NOT a prophet? Seek
    explicit answers from continuationist theologians rather than inventing
    answers for them.
11. Institutional consequences: what documented cases show about prophetic
    status and control of followers, marriages, giving, careers, governance,
    appointments, discipline, criticism, and personal wealth. No personal
    attacks of any sort.
12. "God told me" as an authority claim: what happens to deliberation,
    accountability, interpretation and governance when private revelation
    enters institutional decision-making.
13. The wider thesis - revivalism, religious individualism, democratised
    authority, restorationism, anti-creedalism, independent churches,
    celebrity leadership, mass media, prosperity theology, social media and
    algorithmic amplification. Can "the Bible alone" plus no confession, no
    bishop or presbytery, charismatic personal authority and direct
    revelation produce unusually weak external accountability?
14. How major figures in Christian history would judge this. Reconstruct
    from what they actually wrote - Ignatius, Justin, Irenaeus, Tertullian
    (including his Montanism), Augustine, Chrysostom, Aquinas, Luther,
    Calvin, Cranmer, Wesley. Do not invent answers for them.
15. Construct the strongest continuationist case before concluding: that
    prophecy continues; that NT prophecy differs from canonical prophecy;
    that 1 Cor 14 tells us to weigh prophecy precisely because it can be
    imperfect; that "do not despise prophecies" forbids blanket rejection;
    that abuse does not invalidate use; that no one claims these people are
    Isaiah. Give it every reasonable opportunity, then evaluate it.
16. End on the epistemology. A leader says "God told me X". X is false. The
    leader then says "God told me Y". What rational and biblical basis does
    a Christian now have for believing Y?

Legal and editorial constraint - treat this as binding, not advisory:
eliminate legal risk. Publication is in the United States, in a state with
no anti-SLAPP statute, so the cost of defending even a meritless claim is
the real exposure. Therefore: (a) name living individuals ONLY in connection
with their own published books, recorded sermons, signed statements or other
words they themselves put into public circulation, quoted verbatim with a
citation; (b) do not repeat allegations of misconduct made by third parties
against any living person, however well sourced, and do not repeat findings
of private investigations; (c) make no assertion that imputes crime, fraud,
sexual misconduct, dishonesty or professional unfitness to any living person
or currently operating organisation; (d) draw adverse case studies from
figures who are deceased and events whose record is closed; (e) do not place
a living person's words next to unrelated adverse material in a way that
invites an inference about them; (f) label evaluative conclusions plainly as
the author's opinion resting on facts disclosed on the page; (g) do not link
to infringing copies of in-copyright works. Where following these rules costs
evidential force, say so on the page rather than working around them.
Intent is to evaluate arguments and not to malign anyone.

Requirements: quote primary sources verbatim with citations a reader can
check. Distinguish what the record establishes from what it does not.
Include the evidence that cuts against the thesis - Nathan in 2 Samuel 7,
the old prophet of 1 Kings 13, Jonah and conditional prophecy, Aquinas on
prophets mistaking their own spirit, and the Didache's refusal to judge a
prophet speaking in the Spirit. Do not smooth over disagreement.

If you re-run this, the interesting experiment is not whether you get the same conclusions. It is whether you get the same primary sources. Conclusions are cheap; the citations are the part that can be checked.

What do you currently think?

Reading changes very few minds. Noticing that you were wrong about something changes rather more. So before the evidence: seven questions. Answer them honestly — nothing is scored yet and nothing is sent anywhere. You will see your answers again at the end, next to what the sources say.

Your answers stay in this browser tab. Nothing is transmitted.

What follows

What Scripture means by prophecy

Biblical prophecy is not primarily prediction. It is speech on God’s behalf. Prediction is one of the things such speech sometimes contains, and it matters here only because it is the part that can be checked. Everything else a prophet says — the rebuke, the comfort, the summons to justice — may be true or false without anyone being able to demonstrate which. The date, the name and the outcome are the parts that leave a record.

The whole structure of the office rests on a single claim, and it is worth stating as plainly as the texts state it. A prophet has no authority of his own. He is not an expert. He is not a wise man whose counsel is worth weighing. His entire standing derives from the assertion that what he says did not come from him.

“I will raise them up a Prophet from among their brethren, like unto thee, and will put my words in his mouth; and he shall speak unto them all that I shall command him.”

Deuteronomy 18:18, AV. Unless noted, biblical quotations on this page are from the Authorized (King James) Version, which is in the public domain and which any reader can check against a modern translation in a second tab. Where a modern rendering or the underlying Hebrew or Greek changes the argument, both are given.

Put my words in his mouth. That is the claim, and it generates the problem. If the words are God’s, they are as true as God is. If they turn out not to be true, then either God spoke falsely, which no party to this argument will accept, or the words were not God’s. Scripture takes the second branch without hesitation, and it does so in a passage explicitly framed as an answer to the question how are we supposed to tell?

“And if thou say in thine heart, How shall we know the word which the Lord hath not spoken? When a prophet speaketh in the name of the Lord, if the thing follow not, nor come to pass, that is the thing which the Lord hath not spoken, but the prophet hath spoken it presumptuously: thou shalt not be afraid of him.”

Deuteronomy 18:21–22. The Hebrew rendered “presumptuously” is בְּזָדוֹן (bezadon) — from a root meaning to boil up, to act with insolent self-will. It describes a posture, not an accident.

Notice the shape of that sentence. It is not the prophet was mistaken, nor the revelation was garbled in transmission, nor he had a genuine impression and interpreted it badly. It is a statement about origin: the word was not spoken by the Lord. And the inference runs in the direction that matters for this page — from the observable failure back to the unobservable source.

Accuracy is necessary but not sufficient

Five chapters earlier, Deuteronomy makes the complementary point, and it is the one usually left out of popular arguments on both sides.

“If there arise among you a prophet, or a dreamer of dreams, and giveth thee a sign or a wonder, and the sign or the wonder come to pass, whereof he spake unto thee, saying, Let us go after other gods... thou shalt not hearken unto the words of that prophet.”

Deuteronomy 13:1–3.

So the sign came to pass and the prophet is still to be rejected. Deuteronomy 13 and 18 together give a two-part test with an asymmetry built into it: failure is decisive against; success is not decisive for. A hit proves nothing on its own. A miss settles the matter. This is precisely the inverse of the standard that operates in the cases documented in Part X, and it is worth holding onto, because it means the modern asymmetry is not merely different from the biblical one. It is reversed.

What an established prophet looks like

That is how the text says Israel came to know. Not by an inner witness, not by the impressiveness of the man, and not by the number of times he was right relative to the number of times he was wrong. By his words not falling.

Different prophets, different clarity — but not different accuracy

Scripture does distinguish grades of prophetic experience. Numbers 12 sets Moses apart from every other prophet in the sharpest terms available:

“If there be a prophet among you, I the Lord will make myself known unto him in a vision, and will speak unto him in a dream. My servant Moses is not so... With him will I speak mouth to mouth, even apparently, and not in dark speeches.”

Numbers 12:6–8. “Dark speeches” renders חִידֹת, riddles — the ordinary prophet receives something that requires interpretation; Moses does not.

This passage is genuinely important for the argument this page is testing, and it is one of the strongest cards in the continuationist hand. It concedes, in the Torah itself, that ordinary prophetic reception is less clear than Moses’s: mediated, symbolic, requiring interpretation. If reception can be obscure, interpretation can go wrong. That is not a foolish inference.

But notice what Numbers 12 does not say. It does not say that the lesser prophets’ utterances were correspondingly less reliable, or that Israel was to weigh them against a lower standard. The gradation is in the mode of receiving, not in the accuracy of the resulting word. Every prophet in Israel after Moses received in visions and dreams, and Deuteronomy 18 — written of exactly those prophets — still applies the whole standard to them.

The prophets who were sincere and wrong

Two passages are decisive for the question of whether Scripture recognises the category of an honest, non-malicious, mistaken prophet. Both suggest it does — and both classify such a person as a false prophet anyway.

“Woe unto the foolish prophets, that follow their own spirit, and have seen nothing!... They have seen vanity and lying divination, saying, The Lord saith: and the Lord hath not sent them: and they have made others to hope that they would confirm the word. Have ye not seen a vain vision, and have ye not spoken a lying divination, whereas ye say, The Lord saith it; albeit I have not spoken?”

Ezekiel 13:3, 6–7.

“They have made others to hope that they would confirm the word” — in most modern renderings, they wait for the word to be confirmed. These are not cynics. They expect to be vindicated. They are waiting for the fulfilment that will prove them right. Ezekiel’s charge is not that they knew they were lying; it is that they followed their own spirit and did not know the difference. That is the exact psychological profile of the sincere modern prophet — and the text puts it in the false column without qualification.

“Now therefore, behold, the Lord hath put a lying spirit in the mouth of all these thy prophets...” — and Micaiah’s own test: “If thou return at all in peace, the Lord hath not spoken by me.”

1 Kings 22:23, 28. Four hundred prophets, unanimous, confident, and wrong. Micaiah stakes his own standing on a falsifiable outcome and invites the court to apply the test to him.

Micaiah’s sentence is the most striking single thing in the Old Testament for present purposes. He does not say weigh what I have said and keep the good part. He offers a clean falsification condition against himself, in public, in front of a hostile king, and he offers it as though that were simply what a prophet does.

Jeremiah and the two vocabularies

Jeremiah 23 is the most sustained treatment of false prophecy in Scripture, and it repays reading whole. Three of its lines carry most of the weight:

Verse 28 is the one that matters most, and it is regularly misread in both directions. It does not say “dreams are fine, just label them.” The following verse compares God’s word to fire and a hammer, and the chaff-and-wheat contrast is dismissive. But it does concede something real: there is a category of religious experience — the dream — that a man may honestly report as a dream, and the sin is in dressing it as the word of the Lord. Jeremiah’s objection is to the label, not to the experience. Hold that thought; Part VII is built on it.

Hananiah, and the burden of proof

Jeremiah 28 is the closest thing Scripture gives to a controlled experiment. Hananiah prophesies, specifically and datably, that Babylon’s yoke will be broken within two years. Jeremiah’s reply concedes the epistemic difficulty and then resolves it:

“The prophet which prophesieth of peace, when the word of the prophet shall come to pass, then shall the prophet be known, that the Lord hath truly sent him.”

Jeremiah 28:9. Then, to Hananiah personally: “The Lord hath not sent thee; but thou makest this people to trust in a lie” (28:15). Hananiah dies within the year.

Jeremiah does not have a category into which he can put a well-meaning colleague with a partly-accurate gift. He has two categories: sent, and not sent.

The New Testament vocabulary

The New Testament does not soften the warnings; if anything it sharpens them, and it adds a specific and uncomfortable claim: that supernatural performance and divine approval can come apart.

“Many will say to me in that day, Lord, Lord, have we not prophesied in thy name? and in thy name have cast out devils? and in thy name done many wonderful works? And then will I profess unto them, I never knew you.”

Matthew 7:22–23; cf. 24:24, “there shall arise false Christs, and false prophets, and shall shew great signs and wonders; insomuch that, if it were possible, they shall deceive the very elect.”

This is the passage that most directly cuts against the argument-from-fruit. Jesus does not respond to the prophesying, the exorcisms and the wonders by disputing that they happened. He responds by saying they establish nothing about the person doing them. Any argument of the form look what God has done through him has to survive Matthew 7 first, and the honest observation is that this cuts against everybody — it is as much a warning to critics who assume a fraud must be a fraud all the way down as to defenders who assume a wonder must be a vindication.

Alongside the warnings, the New Testament plainly expects prophecy in the congregation. Peter reads Joel 2 as fulfilled at Pentecost (Acts 2:16–18). Prophets travel between churches (Acts 11:27–30), function in the leadership of Antioch alongside teachers (Acts 13:1–3), and appear in every Pauline list of gifts. Paul tells the Corinthians to desire the gift (1 Cor 14:1, 39). Nothing in this page disputes any of that; the disagreement is about what follows from it.

And it plainly expects prophecy to be assessed:

Everything now turns on what those instructions mean. If “judge” means sift the true parts from the false parts within a genuine prophet’s message, the fallible-prophecy position is essentially established by the text. If it means determine whether this is a true prophet or a false one, it is not. Part IV puts the first reading at full strength; Part V answers it.

Interactive: hunt the counterexample

The thesis makes a strong empirical claim about a finite body of text: there is no accepted biblical prophet who repeatedly attributes false statements to God and continues to function as a legitimate prophet. A claim like that is refuted by one clear case. So here are the ten best candidates, including the ones that most embarrass the thesis. Judge each before you read the analysis.

For each: does this passage show a person presenting a statement as revelation from God, the statement being false, and the person retaining prophetic standing? All three, or it is not a counterexample.

What false prophecy meant — and why the sanction is a distraction

The argument from Deuteronomy 18 is regularly conducted badly by both sides. Critics of modern prophecy say “the Bible says false prophets should be stoned” as though that settled something about a Tuesday-night meeting in a leisure centre. Defenders reply, correctly, that nobody is proposing to stone anyone, that the Mosaic civil code is not in force, and that arguing from the penalty to modern church practice is a category error.

They are right about the penalty, and it does not help them, because the penalty is not the interesting part of the passage. Deuteronomy 18:21 is explicitly framed as an answer to an epistemological question: How shall we know the word which the Lord hath not spoken? The verse that follows is a test, and the sanction is what the civil law then does with the result of the test. Remove the sanction entirely — as the church has — and the test is untouched.

Six ways of explaining a failed prophecy, and what the text does with each

When a word attributed to God does not come to pass, six explanations are available. Scripture licenses some and refuses others.

The table is the argument of this part in miniature. Two of the six explanations are biblically licensed. Three are biblically condemned. One — human contamination — is the whole of the modern debate, and it sits uncomfortably close to the one Ezekiel describes.

What Jonah actually shows

Jonah is the counterexample most often produced, and it deserves a straight answer because it is a good one. “Yet forty days, and Nineveh shall be overthrown” (Jonah 3:4) is unconditional in form. Nineveh was not overthrown. Jonah remains a prophet, and the book is in the canon.

Three things are true about this, and they pull in different directions.

A fair test: does the alleged condition have a scriptural warrant of the Jeremiah 18 kind, and was it identifiable before the outcome? If the condition is only discoverable afterwards, and only by the person whose word failed, it is doing the work of an excuse rather than of an exegesis.

The fallible-prophecy case, at full strength

It is easy to write about modern prophecy as though the theology behind it were an afterthought produced by embarrassed publicists. It is not. The scholarly form of the position was worked out in a Cambridge doctoral thesis supervised by C. F. D. Moule, published in 1982 as The Gift of Prophecy in 1 Corinthians, and restated for a general readership in 1988 and again in a revised edition in 2000. Its author, Wayne Grudem, is a Reformed evangelical whose systematic theology sits on more conservative shelves than almost any other book of its kind. He is not a promoter of prophets. He should be read before he is answered.

The thesis in his own words

“Prophecy in ordinary New Testament churches was not equal to Scripture in authority but was simply a very human — and sometimes partially mistaken — report of something the Holy Spirit brought to someone’s mind.”

Wayne Grudem, The Gift of Prophecy in the New Testament and Today, rev. ed. (Crossway, 2000), p. 18.

Elsewhere he defines the gift as “telling something that God has spontaneously brought to mind,” communicated “in the prophet’s own words” (Systematic Theology, 1994, ch. 53). The architecture has two levels rather than two kinds:

“The apostles, therefore, have authority to write words that are God’s own words, equal in truth status and authority to the words of the Old Testament.”

Grudem, The Gift of Prophecy, p. 33; cf. Systematic Theology, p. 1050. Apostles occupy in the New Testament the place Old Testament prophets occupied; congregational prophets occupy a second, lower level.

The five load-bearing arguments

What Grudem himself says should be done about it

It is worth stressing, because it is routinely omitted by both his critics and his popularisers, that Grudem’s position comes with restrictions. He argues that the introductory formula “Thus says the Lord” is inappropriate for congregational prophecy precisely because it claims the wrong level of authority; that prophecies should be reported humbly and tentatively as impressions; that they should never be used to give directive guidance about major life decisions; and that they must always be subject to Scripture and to the assessment of the congregation. The theology, as he states it, is a theology of modest speech.

Who else holds a version of this

A point of intellectual honesty: the cessationist and continuationist scholars listed above agree with each other far more than the popular argument suggests. Both sides hold that modern impressions are not Scripture, that they must be tested, and that no one should stake a decision on them. The disagreement is about what to call the phenomenon and what the New Testament word προφητεία denoted. That is a real disagreement, and it is smaller than the noise around it.

The strongest critiques

The position set out in Part IV has been answered in detail, in print, by scholars who engaged the exegesis rather than the phenomenon. The four best lines of attack are these. Two of them, in my judgement, land; one is a draw; one fails.

1. Ephesians 2:20 — the grammar does not do what it is asked to do lands

Grudem’s reading of “the apostles and prophets” as a single group leans on a single-article construction. The difficulty is that the Granville Sharp rule, which does establish identity of reference, applies to singular personal nouns. With plurals, a shared article establishes that the two groups are being treated together for the purposes of the sentence — not that they are the same people.

And the immediate context tells against it twice over. Ephesians 3:5 speaks of what “is now revealed unto his holy apostles and prophets by the Spirit”; Ephesians 4:11 lists them as separate offices — “he gave some, apostles; and some, prophets” — with distinct article-plus-particle structure; and 1 Corinthians 12:28 ranks them in sequence, “first apostles, secondarily prophets.” If Paul meant one group in 2:20, he distinguishes them in the next chapter and again two chapters later.

The consequence is significant. If Ephesians 2:20 refers to two groups, then New Testament prophets — as such — are foundational alongside apostles, and the two-level structure loses its clearest textual footing.

2. διακρίνω in 1 Corinthians 14:29 a draw, and it matters

Everything turns on whether “let the other judge” means sift this message for its true and false parts or determine whether this speaker is a true prophet. Both readings are lexically available. Two considerations pull against Grudem:

• Paul has already named a companion gift six verses earlier — “discerning of spirits,” διακρίσεις πνευμάτων (1 Cor 12:10) — using the cognate noun. The natural pairing is: prophecy is uttered, and a gift of discerning its source operates alongside it. That is source-assessment, not quality control.
• 1 John 4:1 uses the same conceptual frame — “try the spirits” — and the criterion it supplies is not which parts are accurate but does this spirit confess Christ come in the flesh. Again: a test of origin.

And two pull towards him: 1 Thessalonians 5:21–22 has “hold fast that which is good” followed by abstaining from the rest, which reads more naturally as sorting within a body of material than as excommunicating individuals; and the setting in 1 Corinthians 14 is a series of practical instructions about a meeting, not a disciplinary procedure.

I do not think this one can be settled from the verb. It is a genuine draw, and anyone who tells you the Greek decides it is overstating.

3. The two-level structure is absent from the earliest evidence lands

If the New Testament church operated a category of genuinely-Spirit-given-but-sometimes-mistaken prophecy, one would expect the sub-apostolic writers — who were dealing with wandering prophets constantly, and who wrote practical manuals about them — to describe it. They do not. The Didache, the earliest church order we have, gives tests for prophets, and every one of them is about conduct and doctrine, not accuracy:

“And do not test or judge any prophet who speaks in the spirit; for every sin shall be forgiven, but this sin shall not be forgiven... But not everyone who speaks in the spirit is a prophet, but only if he has the ways of the Lord. So from their ways shall the false prophet and the prophet be known.”

Didache 11:7–8, Roberts–Donaldson translation. Compare 11:12: whoever says in the Spirit “give me money” is not to be listened to.

This passage cuts in both directions and honesty requires saying so. Against the thesis: the Didache forbids judging a prophet who is speaking in the Spirit, which is a startlingly strong protection and looks nothing like Deuteronomy 18 applied at the door. For the thesis: what it substitutes is not a lower accuracy standard but a behavioural one, and it is ruthless — three days’ hospitality, no money, practise what you preach. It builds a wall around the utterance and then judges the man on everything else. What it does not do is create a space in which a prophet may attribute false statements to God and continue.

Christopher Forbes’s study of prophecy against its Hellenistic background (Prophecy and Inspired Speech in Early Christianity and its Hellenistic Environment, Mohr Siebeck, 1995) presses the same point from the comparative side, and concludes that the ancient evidence does not support a two-tier model of inspired speech in which one tier is expected to contain error.

4. The Agabus argument fails, and it is the most important one

See Part VI, which is devoted to it. In summary: Paul himself, in Acts 28:17, describes what happened to him using the very verb Agabus used, in the same voice, with the same prepositional phrase. That is not a detail. It is the apostle’s own retrospective summary of the event, and it matches the prophecy.

Who has made these arguments

Interactive: the Agabus problem

Everything in Part IV that is specific rather than structural rests on Agabus. If the New Testament’s one named, non-apostolic prophet delivers a message under the highest attribution formula available and gets it wrong, the fallible-prophecy category is demonstrated from the text and this page’s thesis is in serious trouble. So the passage deserves to be read slowly, in Greek, against the three later passages that describe what actually happened.

Use the buttons to highlight each of the three claims across all four texts.

So: was Agabus wrong?

The three things worth noticing whatever you conclude

Interactive: five ways of saying it

Much of the confusion in this argument comes from a single sloppiness: treating five quite different speech-acts as one. They are not one. They make different claims, incur different obligations, and are refuted by different things.

Scripture operates this ladder deliberately

The best evidence that these registers are real, and that collapsing them is a modern habit rather than a biblical one, is that an apostle distinguishes between them four times in a single chapter — carefully, explicitly, and with the reader’s ability to weigh him in view.

Read those five together and the picture is unmistakable. Paul — an apostle, writing what the church would receive as Scripture — takes trouble to mark which of his sentences carry the Lord’s authority and which carry his own. He does not think the distinction is pedantic. He thinks his readers are entitled to know which is which.

The same distinction is visible in the Nathan episode. In 2 Samuel 7:3 the prophet tells David “Go, do all that is in thine heart; for the Lord is with thee” — with no attribution formula whatever. That night the word of the Lord comes to Nathan and reverses the advice, and that message is introduced with “Thus saith the Lord” (7:5). The narrative is not embarrassed, because on the text’s own terms nothing went wrong: a prophet gave an opinion as an opinion, and was then given a word as a word.

Sort these fifteen statements

All fifteen are real. Some are Scripture, some are from the case files in Part X, some are advice from theologians on both sides. Assign each a level. Then see where it actually sits — and, more interestingly, where its speaker later said it sat.

Interactive: the timeline

Two questions can only be settled chronologically. First: when did anyone first write down a theology in which a genuine recipient of divine revelation may nevertheless report it wrongly? Second: did that writing come before or after the failures it is now used to explain?

Click any entry for the primary source and what it does and does not establish. Thirty-four entries, from the Didache to the algorithm.

Which came first — the theology or the failures?

This is the part of the brief that most threatened the thesis, and it is right that it did. The uncomfortable question was: did Christians derive fallible prophecy from Scripture and then meet fallible prophets — or meet repeated failure and then build a theology that could survive it? The instruction was to determine the order of events and to say so if the order weakened the thesis.

It does. Here is the finding, stated as flatly as I can manage.

The three genuine antecedents

What survives, and why I think it is the more interesting claim

Abandoning the crude thesis leaves a narrower one that the chronology actually supports, and it is this:

Where the crude version does retain some force

Two qualifications, in fairness to it.

• The specific two-level exegesis — canonical prophecy versus congregational prophecy, grounded in 1 Corinthians 14:29 and Ephesians 2:20 — is genuinely recent as a worked-out scholarly position: Grudem’s Cambridge thesis was completed in 1978 and published in 1982. That is after the healing revival, after Latter Rain, and contemporaneous with the movements that most needed it. The general theology is old; this version of it is not.
• The rapid, wide adoption of that exegesis in the 1980s and 1990s by movements built on named prophetic ministry is a fact about demand as well as supply. Ideas spread when they are useful. That is not an argument against their truth — plenty of true ideas spread because they were useful — but it is a fact about the sociology worth stating.

A note on what this page could not establish. I wanted to trace, from primary documents, how specific prophetic networks of the late twentieth century handled specific failures internally — the memoranda, the disciplinary correspondence, the internal reviews. Much of that material either is not public, or is public only through partisan compilations, or concerns living people and therefore falls outside this page’s constraints. The result is that the 1980s and 1990s are the thinnest part of the historical account here, and the reader should treat the timeline entries for that period as sketching a shape rather than documenting one.

Interactive: the case files

Six cases, spanning eighteen centuries. Every one is drawn from a person who has died and a record that is closed, and every claim below is either quoted from the person’s own published or recorded words or from a contemporaneous published account, with the citation attached. Each is assessed against the same six questions.

What the six cases establish, and what they do not

Interactive: whose words are these?

Ten quotations about prophecy and its failures, spanning eighteen centuries. Some are from people this page treats critically; some are from their sharpest opponents; several are not where you will expect. The exercise is not a trick — it is a check on whether the positions in this argument are as neatly sorted as they are usually presented.

Interactive: what happens after a miss

A large literature describes what religious movements do when a dated prediction fails. Much of it descends from Leon Festinger’s When Prophecy Fails (1956), which coined “cognitive dissonance” in this connection. The literature is worth reading, and it is also worth treating carefully: Festinger’s central prediction — that failure would increase proselytising — has held up poorly under later testing, and the standard survey volumes (Jon R. Stone, ed., Expecting Armageddon, Routledge 2000; Lorne Dawson’s work in Nova Religio) say so.

So the useful question is not which strategies theorists have named, but which ones are documented in the primary sources on this page. Below, eleven named strategies. Match each to the case where the record shows it happening — or to “not documented here”, which is the honest answer for several of them.

The two best-documented instances, in the sources’ own words

What is not documented here. “Blaming the audience’s lack of faith” and “deleting the record” are frequently alleged in this area and are not established by the primary sources examined for this page. The second is intrinsically hard to document — absence of a recording is not evidence of removal — and the honest thing is to say the sources here do not show it rather than to assert it because it is plausible.

Interactive: what would falsify a prophet?

This is the centre of the argument, and it can be stated in one sentence. If a fulfilled prediction is evidence that a person has the gift, and an unfulfilled prediction is evidence of nothing, then no possible observation can count against the claim — and a claim that no observation can count against has stopped being a claim about the world.

That is a strong charge, so it should be tested against what continuationists actually say rather than against what critics assume they say. Two things turn out to be true. First, several continuationist bodies have written down explicit falsification criteria, and one of them is startling. Second, those criteria are frequently not applied to the people in front of them.

The most surprising document on this page

The Assemblies of God is the largest classical Pentecostal denomination in the world. It is continuationist by definition. Its official position paper on Apostles and Prophets says this:

“A predictive prophecy that proves false leads to the conclusion that the person is a false prophet (Deuteronomy 18:19–22).”

Assemblies of God, Apostles and Prophets, official position paper adopted 2001. The same document states that “the content of prophecy itself should always be tested by and responsible to the superior authority of Scripture,” that “all prophecies are to be weighed carefully (1 Corinthians 14:29),” that “a predictive prophecy may be true, but the prophet whose doctrine departs from biblical truth is false,” and that “the New Testament does not make provisions for establishing the prophet in a hierarchical governing structure of the church.”

Read that first sentence twice. A Pentecostal denomination, in its own official theology, applies Deuteronomy 18 to failed predictive prophecy and draws the conclusion that this page’s thesis says nobody draws. So the charge of universal unfalsifiability is simply wrong, and it should be withdrawn. What remains is narrower and, again, harder: the criterion exists and is not enforced, because the bodies that have written it down have no jurisdiction over the independent ministries where the practice is most visible — which is the argument of Part XV.

What the other criteria on offer actually are

Build your own criterion, then see what it catches

Set the standard you would actually be willing to apply. The tool will tell you which of the six documented cases your standard would have identified in advance, and which it would have missed — including the ones that are hard for every standard.

“God told me” inside an institution

Everything so far has been about truth. This part is about power, which is a different question and in practice the more consequential one. A claim to private revelation, once admitted into an organisation’s decision-making, does something specific to how that organisation can think.

What the claim does to deliberation

Ordinary institutional reasoning has a shape. Someone proposes; others test the proposal against evidence, precedent and consequence; the proposal is amended or dropped; a decision is recorded and can later be reviewed against outcomes. Every stage of that is a place where a mistake can be caught.

“God told me” removes all of them at once, and not by rhetorical bullying — by form. Consider what a board member would have to say to disagree.

This is the mechanism, and it is worth being precise about it because it is often described as though it required bad faith. It does not. A sincere leader who sincerely believes God has spoken will experience opposition as opposition to God, and a sincere board member will hesitate for exactly the same reason. The structure produces the outcome regardless of anyone’s motives.

What the closed record actually shows

Three documented historical cases, all long closed, illustrate the range — including one that cuts the other way.

The honest qualification

Structures with external accountability fail too, sometimes catastrophically and on a larger scale, and any account of this subject that pretends otherwise is not serious. Episcopal and presbyterian bodies have covered up gravely and for decades. The claim here is not that hierarchical churches are safer. It is narrower: that a complaint needs somewhere to go, and that the number of places it can go is a structural property of an institution which does not depend on the character of anyone currently in it. A body with a presbytery can fail to use it. A body without one cannot use it at all.

The wider thesis: what happens when four things combine

The brief asked whether “the Bible alone” plus no confession, no bishop or presbytery, personal charismatic authority and direct revelation can produce unusually weak external accountability. I think the answer is yes, and I think the interesting part is that each ingredient is defensible on its own.

The amplifiers

Three modern conditions make the combination more consequential than it was in 1830.

Where the thesis overreaches, and I think it does

Two honest limits. First, this is a structural argument, not a statistical one: nothing here shows that independent charismatic churches produce more failures per capita than any other kind, and no study I could find establishes it. Second, the combination described above is also the engine of enormous good — the same absence of external permission that removes the check also removes the obstacle, and a very large proportion of the global growth of Christianity in the last century has come through movements with exactly this profile. An argument that treats the structure as simply defective has to explain that, and this one does not try to.

A related argument about the ecclesiology behind this, developed at length with its own sources, is in the companion dossier Every Church Has a Theology, Parts XII–XIV.

Interactive: eleven witnesses

The brief asked how major figures in Christian history would judge the practice under examination, and insisted the answers be reconstructed from what they actually wrote rather than invented. That is a real constraint, and it produces a messier answer than either side would like: the ancient witnesses are not cessationists, and the Reformers are not endorsing anything a modern prophetic ministry does either.

Each card gives what the figure wrote, then what does and does not follow from it. Where a judgement is an inference rather than a statement, it is labelled as one.

The strongest continuationist case, given every opportunity

Before concluding, the position deserves its best statement, made without hedging and without the author’s thumb on the scale. Here it is, in seven moves.

Evaluating it

Moves 1, 2, 3, 4 and 5 I think are simply strong, and this page has not disputed them. Move 6 is the crux, and it is where I part company — not because the two-level distinction is unreasonable, but because the practice under examination does not stay on the lower level. A person who says “thus saith the Lord,” names a date and an outcome, and treats the result as a credential is not exercising the modest gift that moves 2 and 3 describe. He has claimed the upper level, and the defence offered when he fails is drawn from the lower one. That is the relocation the thesis named, and I think it survives.

Move 7 is fair, and it is the one that should make the critic uncomfortable. The cessationist inference from 1 Corinthians 13:10 is genuinely weak, and a great deal of confident cessationism rests on it. Someone can accept every word of Part XIII and still be a continuationist; indeed, the Assemblies of God document quoted there shows exactly what that looks like.

The case against this page

A page arguing only one way is worth less to you than one that tells you where it is weakest. Here are the eight strongest objections to what I have written, stated as their proponents would state them, with my honest assessment of each.

God told me Y

Everything on this page reduces to one situation, and it is worth stating without ornament.

Four answers are available, and they are not equally good.

Three questions worth more than any argument on this page

None of those questions requires you to take a side on 1 Corinthians 14. That is deliberate. The exegetical argument in Parts IV to VI is genuinely unsettled, and honest people hold both positions. The three questions above are not unsettled, and a community can answer them whatever it believes about the gift.

Your answers, revisited

Here are the seven questions you answered before reading, with what you said and what the record shows. There is no score. The only useful question is whether anything was different from what you expected — and if so, where you had picked the expectation up.

Check it yourself

This page is only useful if you can get behind it. Almost everything cited here is online without a paywall, and the primary sources are the point. Start with the ones marked read this first; between them they take about an hour and they will let you form your own view of the central question without reference to mine.

The biblical texts, in order of usefulness

• read this first — Deuteronomy 18:15–22 and 13:1–5, read together and in that order. Ten minutes. Everything in Parts I and III is an unpacking of these two passages.
• Jeremiah 23:9–40 and Jeremiah 28 — the fullest treatment of false prophecy in Scripture, and the only place a false prophet is named, dated and answered in real time.
• Ezekiel 13:1–16 — the sincere-but-unsent prophet. The passage that most directly addresses the modern category.
• 1 Kings 22 (Micaiah) and 1 Kings 13 (the old prophet of Bethel) — the two hardest narratives, pulling in opposite directions.
• read this first — 1 Corinthians 12–14 entire, at one sitting. It is shorter than it feels and almost every argument on both sides is drawn from it.
• 1 Corinthians 7:6, 10, 12, 25, 40 — the register ladder of Part VII, in a single chapter, from Paul’s own hand.
• Acts 11:27–30; 13:1–3; 21:1–14; 21:27–36; 22:22–30; 28:17 — the Agabus dossier. Read all six passages before forming a view; almost nobody does.
• Interlinear Greek for Acts 21:11 and 28:17, free: biblehub.com/text/acts/21-11 and 28-17. Compare Acts 2:23 for the idiom.

Primary sources quoted on this page

• 1. Didache 11–12, Roberts–Donaldson translation: earlychristianwritings.com. Five minutes to read; it changes most people’s picture of the early church.
• 2. Eusebius, Ecclesiastical History V.16–19, on the New Prophecy — the anonymous writer, Miltiades and Apollonius: newadvent.org; parallel text at CCEL. Note that these are hostile witnesses and read them as such.
• 3. Justin Martyr, Dialogue with Trypho 82: newadvent.org.
• 4. Irenaeus, Against Heresies V.6.1: newadvent.org.
• 5. John Chrysostom, Homilies on 1 Corinthians, Homily 29 (on 1 Cor 12:1–2): newadvent.org. The clearest patristic statement of what distinguishes a prophet from a diviner.
• 6. Thomas Aquinas, Summa Theologiae II-II q.171 a.5 (with Gregory the Great quoted in the sed contra): newadvent.org/summa/3171. read this first — it is one page, and it is the single most important pre-modern text for this argument, in either direction.
• 7. John Calvin, Institutes of the Christian Religion I.ix (“All the principles of piety subverted by fanatics who substitute revelations for Scripture”): CCEL.
• 8. Robert Baxter, Narrative of Facts, characterizing the Supernatural Manifestations in Members of Mr. Irving’s Congregation (London, 1833). Digitised: Internet Archive (Wellcome Collection copy). Out of copyright, complete, and by a participant writing against himself — which makes it the most valuable single document in Part X.
• 9. William Miller, letter “To Second Advent Believers” (November 1844) and subsequent apology; Josiah Litch’s assessment; Hiram Edson’s manuscript account of 23 October 1844. Digitised Millerite and early Adventist periodicals are collected at the Adventist Digital Library and the Office of Archives, Statistics and Research.
• 10. Oral Roberts’ January 1987 fundraising letter, quoted verbatim in Christianity Today, “Fund Raising: Did Oral Roberts Go Too Far?” (February 1987): christianitytoday.com; and his 28 February 1987 statement in the Washington Post, “Oral Roberts Starts Vigil for $8 Million” (1 March 1987): washingtonpost.com. UPI’s contemporaneous reports of 20 January and 24 March 1987 are in the UPI archive.
• 11. William Branham, An Exposition of the Seven Church Ages, ch. 9 (“The Laodicean Church Age”), and the sermons cited by date and paragraph. The complete sermon archive is published and maintained by his own ministry, Voice of God Recordings, at branham.org, where the sermons are distributed free and are searchable. Quotations here are cited by the ministry’s own sermon number and paragraph (for example 60-1208, para. 233), so any reader can locate them in that archive exactly. Third-party transcript sites also carry this material; because their licensing position is unclear they are not linked here.
• 12. Harold Camping’s letter to Family Radio listeners, March 2012, full text: Christian Post. His 23 May 2011 press conference is widely archived on video.
• 13. Assemblies of God, Apostles and Prophets, official position paper (2001): ag.org, PDF. read this first — twelve pages, and it will surprise most readers on both sides.

The scholarly argument, from all sides

• Wayne Grudem, The Gift of Prophecy in the New Testament and Today, rev. ed. (Crossway, 2000), and Systematic Theology ch. 53. The position at full strength, from its principal architect. Read it before reading anyone against it.
• Wayne Grudem, ed., Are Miraculous Gifts for Today? Four Views (Zondervan, 1996) — Gaffin, Saucy, Storms and Oss arguing with each other in one volume, each with the right of reply. The most efficient book on the subject.
• D. A. Carson, Showing the Spirit: A Theological Exposition of 1 Corinthians 12–14 (Baker, 1987).
• Richard B. Gaffin Jr., Perspectives on Pentecost (P&R, 1979) — the careful cessationist case.
• F. David Farnell, four-part series on New Testament prophecy, Bibliotheca Sacra 149–150 (1992–93) — the most detailed academic reply to Grudem.
• Christopher Forbes, Prophecy and Inspired Speech in Early Christianity and its Hellenistic Environment (Mohr Siebeck, 1995) — the comparative evidence.
• David E. Aune, Prophecy in Early Christianity and the Ancient Mediterranean World (Eerdmans, 1983) — the standard reference work, and neutral between the parties.
• Kenneth L. Gentry Jr., The Charismatic Gift of Prophecy; Sam Waldron, To Be Continued?; Michael John Beasley, The Fallible Prophets of New Calvinism — three replies at descending levels of technicality.
• Karl Rahner, Visions and Prophecies (Herder, 1963) — the Catholic theology of private revelation, and the most developed account anywhere of fallible revelation inside an accountable structure.
• Leon Festinger et al., When Prophecy Fails (1956); Jon R. Stone, ed., Expecting Armageddon (Routledge, 2000) — the second book is the corrective to the first and should be read with it.
• Grant Wacker, Heaven Below: Early Pentecostals and American Culture (Harvard, 2001) — sympathetic, rigorous, and the best account of how the movement actually thought about revelation.
• Nathan O. Hatch, The Democratization of American Christianity (Yale, 1989) — the background to Part XV.

Five ways to test this page’s honesty

1. Read Aquinas II-II 171.5 and the Assemblies of God position paper back to back — a thirteenth-century Dominican and a twentieth-century Pentecostal denomination, both of which this page treats as cutting against its own thesis. If it has represented them unfairly, that is the fastest place to catch it.
2. Read the six Agabus passages in Part VI in a Bible rather than on this page, in the order Luke gives them, and decide for yourself before reading my verdict. If you conclude Agabus was wrong, Part IV gets stronger and Part V gets weaker; the page should be read that way.
3. Take one prophetic claim you have personally heard, at level 4 or 5 on the Part VII ladder, and try to find the original recording. Whether you can, and how long it takes, is a direct measure of the argument in Part XV about permanence without indexing.
4. Ask a continuationist you respect the question in Part XIII: what would have to happen for us to conclude that this person does not have this gift? The quality of the answer is the finding, whichever way it goes — and a good answer is a real refutation of this page’s sharpest charge.
5. Re-run the prompt at the top with the thesis reversed. Ask for the strongest case that the fallible-prophecy distinction is a recovery of the New Testament’s own category, wrongly assimilated to Deuteronomy 18 by its critics. Compare what comes back against the sources listed here.