They Bought
Tickets to China

What this page is asking

There is a way of arguing about the gift of tongues that never gets anywhere, and almost everybody has sat through it. One side reads 1 Corinthians 14 and says the gift is real and continues. The other side reads the same chapter and says it ceased. Both sides produce their verses, neither side can show the other anything the other cannot absorb, and the argument ends where it began, because nothing either side believes is the sort of thing that can be checked by anyone standing outside it.

This page is not that argument. It is about a period of roughly seven years in which the claim being made was checkable, was checked, and failed — and about what was done afterwards.

The earliest Pentecostal movement understood its restored gift substantially as the gift described in Acts 2: the supernatural bestowal of real, identifiable human languages. That understanding had a consequence that almost no religious claim has. It generated a prediction about the physical world, in a form that any bystander who happened to speak the language could settle in about fifteen seconds.

Missionaries sailed on that proposition. They reached the populations whose languages they believed they had been given. And the communication failed — not subtly, not in a way requiring theological interpretation, but in the plainest sense available: they could not be understood, and they said so themselves. Within a decade the movement’s theology of tongues had moved to forms of the gift in which the speech is not required to constitute any identifiable human language at all.

So the question this page puts is not whether tongues are real. It is a narrower and, I think, a more answerable one:

Three propositions, which are not the same proposition

The subject is impossible to discuss clearly without separating three things that the word tongues is made to carry. This page uses these labels throughout.

This page grants B and C for the sake of argument. It does not attempt to establish that they are false, it does not adjudicate between cessationist and continuationist readings of 1 Corinthians, and it makes no claim about anybody’s experience. Grant both. The historical puzzle survives the granting, and that is precisely what makes it interesting.

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, source-first historical investigation of the following
question, in British English, for a general reader who is willing to work.
Quote scripture from the Authorised Version throughout, so that every
citation can be checked free of charge. Date and attribute every other
quotation: give the publication, the issue, the date and the page.

  The earliest Pentecostal movement interpreted its restored gift of
  tongues substantially as the Acts 2 gift of supernatural human
  languages. That interpretation produced testable missionary claims.
  When missionaries reached the populations whose languages they believed
  they had supernaturally received, the claims repeatedly failed objective
  communication tests. Pentecostal theology subsequently shifted toward
  forms of tongues that did not require the speech to constitute an
  identifiable human language.

  The original proposition was not obscure. It was:

      Pentecost has returned -> Acts 2 has returned ->
      missionaries can receive actual languages.

  That proposition generated a test. The test failed.

Distinguish three propositions throughout and use the labels consistently:

  A. ACTS 2 XENOLALIA. A Galilean miraculously speaks a real human
     language such that a Parthian hears the message in his native
     speech. Verification is built into the claim.
  B. 1 CORINTHIANS CONGREGATIONAL TONGUES. An utterance requiring
     interpretation.
  C. MODERN PRIVATE PRAYER LANGUAGE. Unintelligible speech principally
     directed toward God, sometimes understood as an angelic or heavenly
     language.

GRANT B AND C FOR THE SAKE OF ARGUMENT. Do not attempt to establish that
either is false. Do not adjudicate cessationism versus continuationism.
Take no position on whether the gifts continue. The enquiry is strictly
historical: what was claimed, what test it generated, what the test
returned, and what the movement did next.

Even granting B and C, there is a historical puzzle to be explained. The
phenomenon with the strongest unmistakable New Testament attestation and
the easiest objective verification - A - becomes vanishingly rare. The
phenomenon whose supernatural character is almost impossible for an
outsider to verify - C - becomes extraordinarily common. And modern
Pentecostalism originally claimed A; when A was subjected to obvious
real-world testing, the movement increasingly emphasised B and C. That is
not by itself proof that modern tongues are false. It is a substantial
piece of evidence that a historical investigation should explain.

Work through, at minimum:

 1. THE PROPOSITION AS PRINTED, 1899-1901. Establish that A was the
    claim, from contemporary secular newspapers rather than from later
    hostile retellings: the Topeka State Journal of 7 January 1901, the
    Kansas City Times of 27 January 1901, and the Hawaiian Gazette of
    31 May 1901 (headline and all). Show that it was framed as a
    missionary programme that made language study unnecessary. Build an
    interactive that presents statements about tongues-as-languages and
    asks the reader to date each one, mixing genuine 1901, 1906 and 1916
    material with modern examples composed by the page, so that the
    reader discovers how unfamiliar the original claim now sounds.

 2. WHY ACTS 2 IS TESTABLE AND THE OTHER PASSAGES ARE NOT. Work through
    Acts 2:1-12 clause by clause and show that the verification in the
    narrative is carried by the HEARERS, not by the speakers: "every man
    heard them speak in his own language", "how hear we every man in our
    own tongue, wherein we were born", and the catalogue of fifteen
    peoples. Contrast 1 Corinthians 14:2 ("no man understandeth him"),
    14:13 and 14:27-28. Build an interactive in which the reader must
    say, for each element of the Pentecost narrative, whether it is
    asserted by the speaker or confirmed by an outside hearer.

 3. THE PROGRAMME IN PRINT AND THE PEOPLE WHO ACTED ON IT. The Apostolic
    Faith (Los Angeles), vol. 1 no. 1, September 1906: the gift of
    languages given with the Great Commission, and the list of languages
    reportedly received. Then the departures of late 1906 for Liberia,
    Angola, India and China. Name only the dead, and treat them with
    respect: several died within weeks of landing. Build an interactive
    that hands the reader the manifest and asks him to predict the
    outcome for each party before revealing it.

 4. THE TEST AND THE RESULT. The 1908 account in Confidence (Sunderland)
    by a missionary who had gone to India believing he had been given
    Bengali: what he expected, what he found, what he concluded, and the
    sentence in which he detaches his faith from the empirical claim.
    Then the contemporary count of eighteen missionaries to Japan, China
    and India who could not do it. BE HONEST ABOUT THE STRENGTH OF EACH
    CITATION: say which witnesses are hostile, say where a source reaches
    us only through a critic, and say what evidence would overturn it.
    Build an attribution interactive in which the reader must say, for
    each quotation, whether it comes from the movement's founder, from
    the movement's own paper, from a hostile critic, or from the
    movement's own denominational paper a decade later - and design it so
    that readers of both persuasions are caught out.

 5. THE WITHDRAWAL, IN THE MOVEMENT'S OWN PAPERS. The Weekly Evangel of
    22 April 1916 and 3 June 1916; the Pentecostal Evangel of 1920; the
    1916 Statement of Fundamental Truths and its separation of the
    "initial physical sign" from the gift of tongues; and the later
    internal verdict of the movement's own historians. Build a timeline
    in which the reader must guess the year of each move before it is
    revealed, including the year the claim was withdrawn in print.

 6. SORTING THE TEXTS. Give the reader a set of biblical passages and
    later descriptions and make him assign each to A, B or C. Show how
    few belong unambiguously to A, and how thin the textual basis for C
    actually is - principally one clause of 1 Corinthians 13:1 in a
    sentence whose point is that without charity tongues are nothing.

 7. THE VERIFICATION ASYMMETRY, made explicit. For each of A, B and C,
    have the reader work out who could check the claim, with what, and
    in how long. Then set the answers against the historical trend in
    prevalence. This is the core of the argument and must be the part
    the reader constructs himself rather than the part he is told.

 8. THE OBJECTIONS, ANSWERED INSIDE THE ARGUMENT. Do NOT include a
    separate "case against this page" section and do NOT include a "what
    this page cannot settle" section. Instead, build an interactive that
    states each objection in the strongest form its proponents would use,
    asks the reader to weigh it BEFORE the page replies, and then replies.
    Cover at least: that xenolalia was always a minority reading; that it
    was the founder's error rather than the movement's; that the failure
    was of expectation rather than of the gift; that 1 Corinthians 14
    always described something other than Acts 2, so nothing was lost;
    that doctrines are legitimately refined by experience; and that the
    movement's subsequent growth is itself evidence. Concede what is
    true in each.

 9. WHAT WOULD CHANGE YOUR MIND. Close by asking the reader to state, in
    advance, what evidence would settle each of the three propositions
    for him - and then show him which of those items are obtainable and
    which are not. Apply the same test to this page's own thesis and say
    what would refute it.

CONSTRAINTS. Name no living person, church, ministry, author or
organisation. Everyone quoted must be dead and cited from a dated,
published source. Where an example of a modern formulation is needed,
compose it and label it as composed. Make no claim about anyone's
salvation, sincerity or motives. Treat the missionaries as people acting
in good faith on what they had been taught, because that is what they
were. Label every evaluative conclusion as opinion and rest it on facts
disclosed on the page. Where a citation is weaker than the others, say so
in the body text at the point of use, not in a footnote.

Output a single self-contained interactive HTML page with at least eight
interactive sections that make the reader commit to an answer before the
sources are revealed. At least two must be capable of showing the reader
that his instinct is wrong, and at least one must be capable of
embarrassing a reader who already agrees with the page. Publish this
prompt on the page.

Three instructions did most of the shaping. Granting B and C at the outset is what keeps the page from collapsing into the usual argument, and it is why the conclusion is narrower and harder to dismiss than it would otherwise be. Requiring the strength of each citation to be stated in the body text rather than in a footnote is what produced the paragraph in Part IV about the count of eighteen, which is the most important paragraph on the page for anyone deciding how much weight to put on it. And requiring that the objections be answered inside the argument, with the reader weighing each one before the reply, is what turned Part VII from a list of rebuttals into the part most likely to change somebody’s mind — in either direction.

What do you currently think?

Eight questions, two or three minutes. Answer from what you believe now, not from what you suspect this page wants. Your answers come back at the end alongside what the sources say. Most readers miss the same three, and one of the three is a quotation that nearly everybody attributes to the wrong side of the argument.

What follows

The claim, as it was printed

The first thing to establish is that Proposition A really was the claim, and that we do not have to take a critic’s word for it. The best evidence is the secular press, because the reporters had no stake in the question and were writing before there was a movement with a reputation to protect.

On 1 January 1901, at a small Bible school in Topeka, Kansas, a student named Agnes Ozman spoke in what those present took to be Chinese. Within a week the claim was in the local paper, and it was not phrased as a mystical experience. It was phrased as a solution to a logistical problem:

“The Lord will give us the power of speech to talk to the people of the various nations without having to study them in schools.”

Topeka State Journal, 7 January 1901

Three weeks later, in a larger paper, the practical corollary was drawn out:

“A part of our labor will be to teach the church the uselessness of spending years of time preparing missionaries for work in foreign lands when all they have to do is ask God for power.”

Kansas City Times, 27 January 1901

That is an unusually clean specimen of a religious claim. It is not a claim about an inner state. It says that a particular institutional practice — years of language training before a missionary is sent — is unnecessary, and it says why. It is the sort of claim that has consequences you can count.

By May of that year the story had travelled as far as Honolulu, where a paper ran it under a headline that puts the whole thesis in nine words — New Kind of Missionaries: Envoys to the Heathen Should Have Gift of Tongues — and reported that those who sought the gift in faith would

“be made able to talk to the people whom they choose to work among in their own language, which will, of course, be an inestimable advantage.”

Hawaiian Gazette, 31 May 1901, p. 10

A note on the pre-history: the expectation did not begin in 1901. A woman named Jennie Glassey was reported in 1899 to have received a dozen African dialects and several other languages, and sailed for Sierra Leone in December 1895 on that basis; the report appeared in the Kansas paper Apostolic Faith of 3 May 1899. The idea was in circulation in holiness and missionary periodicals through the 1890s. What changed in 1901 was that it acquired a doctrine, a school, and a man prepared to teach it as the ordinary expectation of every believer who received the baptism.

Interactive — when was this written?

Six statements about tongues. Three are genuine and dated; three were composed for this page in the register of a period they do not belong to. Place each one. You will not be told which is which until you have committed to all six.

0 of 6 answered.

Who does the verifying at Pentecost?

Acts 2 is a strange passage to read carefully, because almost everything load-bearing in it is reported from outside the group experiencing the miracle. The disciples are described as doing one thing. The crowd is described as doing everything else.

“And they were all filled with the Holy Ghost, and began to speak with other tongues, as the Spirit gave them utterance. Now when this was noised abroad, the multitude came together, and were confounded, because that every man heard them speak in his own language. And they were all amazed and marvelled, saying one to another, Behold, are not all these which speak Galilaeans? And how hear we every man in our own tongue, wherein we were born?”

Acts 2:4–8, Authorised Version

Then comes the catalogue, and the catalogue is the point. Parthians, Medes, Elamites, and the dwellers in Mesopotamia, Judaea, Cappadocia, Pontus, Asia, Phrygia, Pamphylia, Egypt, the parts of Libya about Cyrene, strangers of Rome, Jews and proselytes, Cretes and Arabians — fifteen named populations, and then the verdict they deliver: “we do hear them speak in our tongues the wonderful works of God” (Acts 2:11).

Notice the grammar of the miracle. The disciples do not announce that they have been given Parthian. A Parthian announces it. The claim is made by people who had every reason to be sceptical, who did not share the speakers’ language, and whose competence to judge is the only competence in the room that matters. Luke even records the alternative explanation being offered on the spot — “others mocking said, These men are full of new wine” (2:13) — which is to say the narrative includes a control group who heard the same thing and drew a different conclusion.

Now set beside it the passage that governs Proposition B:

“For he that speaketh in an unknown tongue speaketh not unto men, but unto God: for no man understandeth him; howbeit in the spirit he speaketh mysteries.”

1 Corinthians 14:2, Authorised Version

“No man understandeth him.” Whatever else is true, that is a different situation from the one in Acts 2, where the entire evidential weight rests on men understanding. Paul’s remedy is an interpreter — “let him that speaketh in an unknown tongue pray that he may interpret” (14:13); “if there be no interpreter, let him keep silence in the church” (14:28) — and an interpreter is a member of the same community, which is a weaker check than a Parthian who has just walked in off the street.

The word unknown, incidentally, is supplied by the Authorised Version’s translators and is printed in italics in modern settings of the Authorised Version to show that it has no equivalent in the Greek. That does not damage the argument here, and this page mentions it because a reader who discovers it elsewhere should not think it was being hidden. The contrast that matters is not carried by that word. It is carried by no man understandeth him, which is in the Greek, and by the fifteen nations, who are also in the Greek.

Interactive — asserted, or confirmed?

Eight elements of the Pentecost narrative. For each one, say whether it is something the speakers assert about themselves, or something the outside hearers confirm. Most readers get six of eight and are surprised by which two they miss.

0 of 8 answered.

The manifest

In September 1906 the Azusa Street mission in Los Angeles began publishing its own paper. The first issue states the doctrine without qualification, and connects it directly to the Great Commission:

“The gift of languages is given with the commission, ‘Go ye into all the world and preach the Gospel to every creature.’”

The Apostolic Faith (Los Angeles), vol. 1 no. 1, September 1906

The same issue prints a list of what had reportedly been received — Greek, Latin, Hebrew, French, German, Italian, Chinese, Japanese, Zulu and languages of Africa, Hindu and Bengali and dialects of India, Chippewa and other languages of the Indians, Esquimaux, and more. It also contains a sentence that turns out, in hindsight, to be the first hairline crack in the doctrine:

“The gift of tongues can only be used as the Spirit gives utterance. It cannot be learned like the native tongues, but the Lord takes control of the organs of speech at will. It is emphatically, God’s message.”

The Apostolic Faith (Los Angeles), vol. 1 no. 1, September 1906

Read that as a missionary would. It grants the language, and then withdraws control of it. You have Bengali, but you cannot use it when you choose to. Nobody at the time seems to have noticed that this makes the gift useless for the purpose it was being advertised for — you cannot conduct a conversation in a language that switches itself on and off at will — and it is the sentence a careful reader in 1906 would have wanted to press on.

Nobody pressed on it. Within about three months, people sailed.

Interactive — predict the outcomes

Six parties who left Los Angeles for the mission field in the closing weeks of 1906 and the first days of 1907, in the confidence that the languages had been given. For each, say what you think happened. Then see.

0 of 6 answered.

The test, and who is telling you about it

This is the part of the page where the evidence has to be handled carefully, because the temptation is to take the most damning witness and put him first. The most damning witnesses here are hostile ones, and a page that leads with them deserves to be disbelieved. So this Part is arranged the other way round: friendly witnesses first, hostile witnesses afterwards and labelled as hostile, and the weakest citation identified as weakest before anybody has to rely on it.

The witness who sailed

Alfred G. Garr, a Kentucky-born pastor who had been running a mission in Los Angeles, received the experience at Azusa Street in 1906 and left for India with his wife Lillian at the turn of that year, believing he had been given Bengali. They reached Calcutta on 13 January 1907 and were in Hong Kong by the following October. He was not a marginal figure — he was among the first from the mission to go abroad on the doctrine. In 1908 he set down an account of what had happened, in a Pentecostal paper published in Sunderland, for a Pentecostal readership. Nothing about the venue makes it easy to dismiss.

“I know for some time I was saying the word, ‘Bengalee,’ and when I reached India, I found myself in the Bengal Province.”

Confidence: A Pentecostal Paper for Great Britain, May 1908, Special Supplement, p. 2

That is the expectation, in his own words, and it is worth pausing on how reasonable it sounded from inside. A man finds a word forming in his mouth for weeks; he goes where the word points; he arrives in the province the word names. Then:

“However, before leaving for America I noticed that the languages changed and I was talking quite a different tongue, and after reaching Calcutta I noticed another change but could not understand the words.”

Confidence, May 1908, Special Supplement, p. 2

And then the sentence that does more work than anything else on this page, because it is not a critic’s summary of what the movement did next. It is a participant doing it, in real time, in print:

What has happened in that sentence is that the empirical content of a claim has been severed from the claim, and the claim has been kept. In 1901 the proposition was: you will be able to talk to the people in their own language. In 1908 it is: whether it was a language or not does not matter, because God was speaking. Those are not the same proposition, and the second cannot be checked by anybody.

He also, to his credit, reported the general result rather than only his own. Writing from Hong Kong, where he had gone on from India:

“So far I have not seen any one who is able to preach to the natives in their own tongues with the languages given with the Holy Ghost. Here in Hong Kong, we preached the word to the Chinese through an interpreter.”

Confidence, May 1908, Special Supplement, p. 3

Not one person, in two countries, two years after the doctrine was printed with a list of languages attached. And the working arrangement on the ground — an interpreter, the ordinary tool of ordinary missions — is exactly what the 1901 claim had said would no longer be needed.

The count of eighteen, and why this page marks it as its weakest citation

The most frequently quoted piece of evidence in this whole subject is a report attributed to S. C. Todd of the Bible Missionary Society, who is said to have investigated eighteen Pentecostals who went to Japan, China and India “expecting to preach to the natives in those countries in their own tongue”, and to have found that by their own admission “in no single instance” had they been able to do so.

It is a striking figure, and it is used everywhere. Here is the problem with it, stated plainly rather than buried. The earliest printing this page has been able to locate is not Todd’s own publication. It is an article by A. E. Seddon, “Edward Irving and Unknown Tongues”, in The Homiletic Review (New York: Funk and Wagnalls, 1909), p. 109 — a critical piece, in a mainstream Protestant journal, written by somebody unsympathetic to the movement. Todd’s original report, its date, its method, and the identities of the eighteen are not given there and this page has not found them.

The hostile witnesses, labelled as hostile

Two more sources belong here, and both need their prejudices declared before they are read.

The first is Charles Parham, the man who originated the doctrine at Topeka. By late 1906 he had fallen out comprehensively with the Azusa Street mission, and his account of visiting it is the account of a rival. Read with that discount applied, it is still remarkable:

“I sat on the platform in Azusa Street Mission, and saw the manifestations of the flesh … one who was a hypnotist (I had seen him lay his hands on many who came through chattering, jabbering and sputtering, speaking in no language at all).”

The Life of Charles F. Parham (1930), p. 163 — a hostile witness, writing about a mission he had broken with. Weigh accordingly.

What makes it worth quoting is not that it is damning — a rival calling a rival’s meeting fleshly is worth very little. It is the criterion he applies. He is not saying the utterance was insufficiently reverent or insufficiently orderly. He is saying it was no language at all, and he is treating that as decisive, because on Proposition A it is decisive. The founder of the doctrine is here functioning as its first and most exacting tester, and reporting a negative. He never abandoned Proposition A, and held to it for the rest of his life while the movement he had started moved away from it — which tells you that the shift was not forced by the logic of the doctrine but by what happened when people tried to use it.

The second is a holiness paper openly opposed to the Pentecostal movement, which reported in November 1906 that a man sent to check had “come back stating he could find no one who could really speak in any of the languages of India”, and in September of the same year that a claimed Arabic “sounded like a lot of turkey gobblers” (Rocky Mountain Pillar of Fire, 12 September and 14 November 1906). That is a hostile witness in a hostile paper, and a reader is entitled to weight it at close to nothing on its own. It is included because it establishes something that does not depend on the reporter’s honesty: that in 1906 the claim was understood by everybody, friend and foe, to be the kind of claim you check by going and asking a speaker of the language. Nobody at the time thought that was a category error. That understanding is what has since been lost.

Interactive — who said it?

Eight quotations from the period. For each, say where it comes from: the movement’s founder, the movement’s own paper at the height of the revival, a hostile critic, or the movement’s own denominational paper a decade later. This one is built to catch out readers on both sides, and it usually does.

0 of 8 answered.

The year it was withdrawn

Doctrines usually change slowly and without anybody putting a date on it. This one has a date, because the movement had a weekly paper and the paper printed the change.

On 22 April 1916, The Weekly Evangel carried the replacement doctrine:

“This is not a gift of different languages as some have believed, but is an emotional or heavenly language, in which the speaker speaks only to God.”

The Weekly Evangel, no. 136, 22 April 1916, p. 6

Six weeks later, on 3 June 1916, the same paper printed the retraction of the old one:

Nineteen words. No explanation of how the belief had arisen, no account of who had taught it, no mention of the people who had sailed on it ten years earlier and in several cases had not come back. Some have believed. The passive voice is doing a great deal of work.

In the same year the General Council of the Assemblies of God adopted a Statement of Fundamental Truths which fixed the replacement in confessional language, where it remains. The baptism in the Holy Spirit is “witnessed by the initial physical sign of speaking with other tongues as the Spirit of God gives them utterance” — and then a careful sentence performs the separation that the previous decade had made necessary: this speaking “is the same in essence as the gift of tongues, but is different in purpose and use.”

Read that sentence as a piece of engineering and it is rather elegant. The sign and the gift are now two things. The sign has no communicative function to fail at — its purpose is to be a sign — and the gift retains the regulation of 1 Corinthians 14, which requires an interpreter and therefore never has to survive contact with a Parthian. Between them they occupy the whole ground that Proposition A used to occupy, and neither of them can be falsified by anything that happens on a mission field.

By 1920 the private use was being commended in the movement’s paper in terms that would have been unintelligible in 1906 — “the private use of the gift of tongues as a medium of expressing the heart’s deepest emotions” (Pentecostal Evangel, nos. 336–337, 1920, p. 7). And by the 1940s the movement’s own historian, writing a short history for British readers, described the missionary-tongues expectation of the pioneers as mistaken and unscriptural. The internal verdict was in, and has not been disturbed since.

It is worth saying what this does not show. It does not show a cover-up. A movement correcting a mistake in its own newspaper is a movement behaving reasonably, and the 1916 sentence is a correction, printed where its own people would see it. What is odd is not the correction but its afterlife: the episode has almost entirely vanished from the movement’s account of its own origins, so that a member today can know the story of Azusa Street in some detail and not know that the doctrine which sent its first missionaries abroad was withdrawn in print ten years later.

Interactive — guess the year

Ten moves, from 1899 to the 1940s. For each, commit to a year before you open it. The one most readers get badly wrong is the retraction itself — almost everybody places it far too late.

0 of 10 answered.

Which proposition is this text about?

Most of the confusion in this subject is produced by a single habit: quoting a text that belongs to one of the three propositions in support of a different one. It is done in both directions and usually without any intent to deceive, because the word tongues sits in all three and the English translation does not distinguish them.

Before the exercise, the textual position, stated as fairly as this page can manage.

The upshot is that the distribution of textual support runs almost exactly opposite to the distribution of modern practice. A has the clearest attestation and is rare. C has the thinnest and is close to universal. That is not an argument that C is false. It is an observation that wants an explanation, and “the texts led us here” is not available as one.

Interactive — sort ten texts

For each, say which proposition it most naturally describes: A, B, C, or none of them clearly. Two of these are traps and one of them catches nearly everybody.

0 of 10 answered.

Who could check it, and how long would it take?

This is the centre of the argument, and it is deliberately left for you to construct. The page will not tell you the asymmetry; you are going to work it out by answering the same four questions about each of the three propositions and then looking at what you have produced.

The four questions are the ordinary questions anybody asks about any claim, religious or not. Can a person outside the community check it? Does checking it require any special access or equipment? Could a single observation settle it? And is there any result that the claim’s holders would accept as counting against it?

That last one is the one that matters. A claim which no possible observation could count against is not thereby false. It is, however, a claim about which evidence has nothing further to say, and anyone who holds it should know that he holds it on grounds other than evidence.

Interactive — the four questions, three times

Answer for each proposition. The page adds up what you said afterwards and shows you the shape of it.

And the fact it has to be set against

Whatever you concluded above, set it against the historical trend, which is not in dispute and is accepted by the movement’s own historians. In 1906 the claim being made in public, in print, with a list of languages, was A. Today the overwhelming majority of what is practised and defended under the word tongues is C, with B in a regulated congregational role. The movement grew by orders of magnitude across that period. The proposition that could be checked went out of use; the proposition that could not went everywhere.

There are innocent explanations of that, and Part VIII puts the best of them. But notice that the innocent explanations all have to account for the direction. Revisions under evidence normally run in both directions over a century: some claims get narrower, some get bolder as confidence accumulates. Here every revision has run the same way, from the checkable toward the uncheckable, and none has ever run back.

The replies, weighed before they are answered

Six replies are made to the argument of this page, and some of them are good. Each is stated below in the strongest form its proponents would use. You are asked to weigh each one before the page answers it, because a reader who reads the reply and the answer together will simply absorb the answer, and the point of the exercise is to find out where you actually stand.

Where a reply is partly right, the page says so and says which part. Two of the six are, in my judgement, substantially correct and require the thesis of this page to be narrowed. You will not be told which two until you have weighed all six.

0 of 6 weighed.

What would change your mind?

One question decides how much of this page you should have accepted, and it is not a question about Pentecostalism. It is a question about you.

Below are twelve pieces of evidence. Tick the ones which, if you encountered them, would genuinely move you — not the ones you think you ought to tick. Then the page will tell you something about the list you have made, which is that some of the items on it are obtainable in an afternoon and some of them are not obtainable by anybody, ever, and that the split between those two groups is not random.

And the same test, applied to this page

It would be a poor sort of argument that demanded falsifiability of everybody else and offered none itself. So, explicitly, here is what would refute the thesis of this page.

Items two and three are the ones a serious opponent should go after, and both are ordinary historical work — archives, mission board records, periodical runs — rather than theology. If you do that work and it comes out the other way, publish it. That is the whole point of putting the prompt at the top of the page.

Your answers, revisited

Here is what you said before you read the page, next to what the sources show.

Check it yourself

Thirteen items. An unusual amount of this subject is free: the early Pentecostal periodicals have been digitised and are searchable, the 1901 newspapers are in a national digital archive at no charge, and the Authorised Version is out of copyright, which is why it was used throughout. If you read only four, read the ones marked Read this first — between them they contain the entire evidential spine of this page, and two of the four are single issues of a newspaper. Modern scholarly works below are listed by title rather than by author, in keeping with this page’s rule of naming no living person.

1. 1. Acts 2:1–21, and 1 Corinthians 12–14 entire. Read this first. Forty minutes, free. Read Acts 2 once for the story and a second time marking who says what — the exercise in Part II is just that, done for you. Then read the three Corinthian chapters straight through rather than in the usual fragments; 12 and 13 change the sense of 14 considerably, and chapter 14’s tone toward the gift is more restrictive than either its friends or its enemies usually let on.
2. 2. The Apostolic Faith (Los Angeles), September 1906 onward. Read this first. An hour for the first three issues, free. Digitised and searchable at the Consortium of Pentecostal Archives (pentecostalarchives.org), with a run also at Asbury Theological Seminary’s ePLACE repository. This is the movement speaking for itself at its most confident, and it is the single best corrective to any account of this period — including this one. Read the September 1906 issue in full rather than the extracts quoted here. Then read forward through 1907 and 1908 and watch how the vocabulary changes without announcement.
3. 3. Confidence: A Pentecostal Paper for Great Britain, May 1908, Special Supplement. Read this first. Twenty minutes, free at the Consortium of Pentecostal Archives. The Garr account, in full and in context. Everything this page says in Part IV comes from pages 2 and 3 of that supplement, and a reader who thinks the quotations have been arranged unfairly should go and check, because the surrounding material is as interesting as the sentences taken out of it. Read the whole supplement, not the quoted lines.
4. 4. The Weekly Evangel, 22 April and 3 June 1916. Read this first. Fifteen minutes, digitised by the Flower Pentecostal Heritage Center (ifphc.org) and searchable there. Two issues, six weeks apart, containing the replacement doctrine and the retraction of the old one. If you read nothing else on this list, read these two, and notice that neither of them mentions the missionaries.
5. 5. Topeka State Journal, 7 January 1901; Kansas City Times, 27 January 1901; Hawaiian Gazette, 31 May 1901. Thirty minutes, free at Chronicling America, the Library of Congress newspaper archive, which carries the Topeka State Journal and the Hawaiian Gazette in full-page scans. Read the whole articles and not the quoted sentences: the reporters’ tone is worth having, and so is the fact that they thought this was a story about missionary logistics rather than about religious experience.
6. 6. A. E. Seddon, “Edward Irving and Unknown Tongues”, The Homiletic Review (New York: Funk and Wagnalls, 1909), p. 109. The source of the count of eighteen, and the weakest load-bearing citation on this page. It is out of copyright and findable in digitised runs of the journal. Go and read the page it sits on, and notice that it is a critical article and that Todd’s own report is not reproduced. If you can find that original report, you will have advanced the subject; this page could not.
7. 7. The Statement of Fundamental Truths adopted by the General Council of the Assemblies of God in 1916, and its current text. Twenty minutes, free from the denomination’s own site and from digitised General Council minutes at the Flower Pentecostal Heritage Center. Read the articles on the baptism and on the initial physical evidence, and read the sentence separating the sign from the gift twice. This is the settlement that the previous fifteen years produced, and it is still in force.
8. 8. The Life of Charles F. Parham (1930). Several hours; in print in reissue and findable secondhand. A wife’s memoir of her husband, and therefore not neutral in either direction — but it is where the account of the Azusa visit at p. 163 comes from, and it is the best source for the fact that the originator of the doctrine never let go of Proposition A. That obstinacy is evidence: he had every incentive to follow the movement and did not.
9. 9. The Pentecostal Movement: A Short History and an Interpretation for British Readers (1941, revised 1949). Four hours; in libraries and secondhand. The movement’s own historian, writing for its own people, describing the missionary-tongues expectation as mistaken and unscriptural. Read it if you suspect this page has got the internal verdict wrong. It has not, and the internal verdict is older and more candid than most outsiders assume.
10. 10. With Signs Following (1926; revised 1946). Four hours; secondhand and in some digitised collections. This one rewards a peculiar kind of reading. The early chapters are full of accounts of miraculous foreign languages; the closing material treats tongues as speech between a man and God. The two positions sit in the same book without the join being acknowledged. Whatever you conclude from that, it is a document of the transition rather than a report on it.
11. 11. Vision of the Disinherited: The Making of American Pentecostalism (Oxford University Press, 1979). Ten hours; in libraries, and lendable through the Internet Archive. The standard social history, and the route by which the count of eighteen reached most modern discussions. Read it for the wider account of the movement’s origins, which is more sympathetic and more interesting than its title suggests.
12. 12. “Shortcut to Language Preparation? Radical Evangelicals, Missions, and the Gift of Tongues”, International Bulletin of Missionary Research (2001), by the Assemblies of God historian Gary B. McGee (1945–2008); and the article “The Origin of T. B. Barratt’s Concept of ‘Missionary Tongues’” in the Cyberjournal for Pentecostal-Charismatic Research. Three hours; both free online. These are the serious scholarly treatments of exactly the question this page asks, written from inside the tradition rather than against it, and they trace the expectation back into the 1880s and 1890s. If you want one item that shows the episode is not a polemical invention, this is it.
13. 13. On the other side of the argument, and on what came after: Tongues of Men and Angels: The Religious Language of Pentecostalism (Macmillan, 1972), by the linguist William J. Samarin (1926–2020), and Speaking in Tongues: A Cross-Cultural Study of Glossolalia (1972) by the anthropologist Felicitas Goodman (1914–2005). Both are studies of Proposition C rather than of Proposition A, and they are listed here precisely because this page granted C and did not examine it. A reader who wants to go on from the historical question to the linguistic one starts here, and should be warned that both authors’ conclusions are contested and that neither settles anything this page argued. Read also, on the other side of the historical question, the account of Pentecostal missions expansion after 1920 in any standard history: whatever the origins, what followed was not a small thing, and an argument about 1906 is not an argument about 1970.