Did Paul Write the Disputed Letters?

"Many of the objections to Pauline authorship are not individually capable of disproving it but it is their cumulative effect which suggests another author." - Ernest Best, on Ephesians, 1998¹

Summary

Thirteen New Testament letters open with Paul's name.  Nearly all scholars accept seven:  Romans, 1 and 2 Corinthians, Galatians, Philippians, 1 Thessalonians and Philemon.  Six are disputed:  2 Thessalonians, Colossians, Ephesians, 1 Timothy, 2 Timothy and Titus.  This article judges each of the six on its own evidence, without a vote of scholars and without one label for all six.  It tests the strongest case each way, including the case for Paul made by Jonathan McLatchie and by Erik Manning of the Testify channel, and the case against made by Bart Ehrman, Ernest Best and Eduard Lohse.

"Did Paul write it?" is really two questions, and they can have different answers.

1. Is the letter Paul's? That is, did Paul send it in his lifetime, whether he chose the words or a co-worker drafted them for him to approve? For history this is what matters: such a letter is first-hand evidence of Paul's mind.
2. Are the words Paul's own? A letter can be Paul's in the first sense even if a helper chose most of its words.

A letter written in Paul's name after his death is not Paul's in either sense.  It is evidence of what his followers believed.

The verdicts, letter by letter:

1. 2 Thessalonians: probably Paul's, sent in his lifetime with Silvanus and Timothy, as it says. Low to moderate confidence. Marcion's collection had it, Polycarp probably knew it and gives words close to it as Paul's, its vocabulary is Paul's, and one multivariate study of style placed it with Paul. Against it, it follows 1 Thessalonians' wording and layout more closely than Paul's other pairs of letters follow each other, and counts of strings of genitive nouns, Schmidt's and this article's, put it well outside Paul's range. A later imitator of 1 Thessalonians remains a serious alternative.
2. Colossians: more likely than not Paul's, sent in his lifetime, but its wording is probably not his own. Low confidence on the first, low to moderate on the second. Its fit with Philemon is the best evidence for a lifetime origin, and it is evidence of a real but limited kind: it shows that the two letters cohere unobtrusively, not that Colossians got some independently known fact right. Its style is the best evidence against Paul's own wording: several measures of how it builds sentences fall outside the range of his accepted letters. If a helper drafted it, Timothy, the named co-sender, is a candidate, but no evidence picks him out.
3. Ephesians: unresolved. Its vocabulary fits Paul and Polycarp quotes it, apparently as scripture. But much of its wording overlaps with Colossians, its first half is written in a style unlike Paul's, and some phrases sound as if Paul and the readers had not met. Its date does not simply follow from that of Colossians, because the direction of borrowing between the two is itself disputed.
4. 1 Timothy: probably not in Paul's own words; whether it is Paul's, sent through a helper, or written after his death is unresolved. Moderate confidence on the wording. Every 600-word stretch of it uses words absent from Paul's accepted letters at a higher rate than any stretch of those letters long enough to measure. Its rules for church leaders, its "Scripture" quotation that matches Luke, and its missing place in Acts are weaker evidence of a later date than they first look, because each has a reasonable explanation inside Paul's lifetime: a letter written to a delegate, a saying of Jesus known before Luke was written, and a release from prison or a gap in Acts.
5. 2 Timothy: open. Its vocabulary is mostly as far from Paul's as 1 Timothy's, though its opening stretch is not, and it is the most personal of the six letters. The evidence does not choose between Paul writing through a helper near his death, a later writer using real notes, and a later writer with good information.
6. Titus: probably not in Paul's own words; whether it is Paul's is unresolved. Lower confidence than for 1 Timothy, because it has only 659 words. It very probably shares an author with 1 Timothy. A papyrus of about 200 may not have contained it or the two letters to Timothy, but that rests on a reconstruction of the book's missing pages that is now disputed.

Six things hold across the six letters:

1. Early churches named Paul for every one of them. Irenaeus, about AD 180, names Paul as the writer of Titus, Ephesians and Colossians, of 2 Thessalonians through "the apostle" whose Galatians he has just quoted, and of "the Epistles to Timothy" in the plural. Earlier writers use some of the letters without naming them. No early writer reports a doubt about who wrote any of them on grounds of style or history; those who rejected some of them (Marcion, Tatian, and the "heretics" Clement of Alexandria mentions) objected to their teaching, as far as the reports say.
2. The early evidence differs by letter. Polycarp of Smyrna, writing between about 110 and 135, quotes Ephesians in Greek, very probably knew 1 Timothy, probably knew 2 Timothy, and in a passage that survives only in Latin attributes to Paul words close to 2 Thessalonians. Colossians and Titus have no clear use that early. Marcion's collection, by the 140s, held 2 Thessalonians, Colossians and Ephesians but not the three Pastorals.
3. Language separates the six into groups. Counted the same way and against Paul's own spread, 2 Thessalonians, Colossians and Ephesians use Paul's vocabulary, and 1 Timothy, Titus and most of 2 Timothy do not. Colossians and Ephesians nonetheless build their sentences differently from Paul. Van Nes's adjusted rare-word analysis puts all six inside Paul's predicted range. None of these counts is a test of authorship: one writer can step outside his own past range when his subject, audience or helper changes.
4. A helper or co-writer is a real ancient practice, not a universal escape. Paul named co-senders on five of the seven accepted letters, added "all the brothers who are with me" to Galatians, and used a scribe for Romans. A helper can explain different wording. A letter Paul sent and approved can also differ in emphasis because its readers and problems differ. What an authorised letter cannot contain is teaching Paul would have rejected. In the evidence surveyed here and in the sister article, no case is documented of a helper composing a long argued letter in another man's name, but the ancient record of how co-authored letters were made is nearly empty.
5. Coherence is weaker evidence than independent confirmation. The "undesigned coincidences" McLatchie and Manning collect are real. Where a detail in a disputed letter agrees with Acts or Romans, a later writer could have taken it from those books, or from tradition, so the agreement counts for a well-informed writer more than for Paul himself. That possibility lowers their weight; it does not reduce it to nothing, because a possible source is not a demonstrated one.
6. Several popular arguments on both sides are weaker than their users think. On the critics' side: the Pastorals' church officers, their "Scripture" quotation that matches Luke, and Ephesians' "if indeed you heard". On the defenders' side: the claims that the Didache quotes 1 Timothy, that the word "chain" appears only twice, and that Ephesians and Colossians share 32 consecutive words.

Hebrews is not one of the six.  It never names its author, so it cannot be a false claim to be Paul;  the question of who wrote it is treated in Ancient Citation and Authorship Conventions.  What Paul's seven accepted letters say about Jesus is in Paul's Letters as Historical Evidence.

Part 1: What exactly is being decided

Five ways a letter can carry Paul's name

For each letter this article asks which of five accounts fits the evidence best.

1. Paul composed it, dictating to a scribe, as he did Romans: "I, Tertius, who write the letter, greet you in the Lord" (Rom 16:22).² The words are his.
2. Paul authorised it, and a co-worker shaped much of the wording. Five of the seven accepted letters name a co-sender, and the ancient world knew letters composed by a trusted friend in another man's name. The thoughts are Paul's, or approved by him; the words may not be.
3. Genuine fragments were later combined or edited. P. N. Harrison argued in 1921 that the Pastorals contain "five genuine notes" of Paul worked into later letters.³
4. Someone wrote it after Paul's death in his name. This is what critics mean by a pseudonymous or forged letter. Some argue such letters were an accepted convention; Bart Ehrman argues they were meant to deceive.
5. A later reader attached Paul's name to an anonymous work. This applies to Hebrews, not to these six. All six open with Paul's name, and several add "I, Paul" in the body (Col 1:23; 4:18; Eph 3:1; 2 Thess 3:17).

Accounts 1 and 2 both make the letter Paul's in the sense that matters for history:  a first-hand witness of his mind, sent in his lifetime.  This article therefore never counts a letter as "not Paul's" merely because its wording seems to be someone else's;  it says separately whether the wording is his own and whether the letter is his.  Account 4 makes the letter evidence for what Paul's followers believed a generation later.  Account 3 sits between.

What counts as evidence, and what does not

1. The seven accepted letters are the baseline. Every measure of Paul's style and thought in this article is taken from them alone. That is a convenience, not a proof that Paul wrote only seven. But it is the only way to compare without assuming the answer.
2. Votes do not count. Ehrman writes that the scholars who still defend 2 Thessalonians do so because of "theological views that make the presence of literary forgeries in the canon of scripture untenable on principle."⁴ Luke Timothy Johnson, a Catholic scholar who defends the Pastorals, writes that their rejection has become "a matter of dogma."⁵ Each side can explain away the other's majority. So this article weighs arguments, not head-counts.
3. Using a letter, naming its author and dating it are three different things. A quotation without a name shows the letter existed by then. A name tells what one reader believed, and how early. Neither tells when the letter was written, except that it was earlier.
4. A difference from Paul's measured range is a description, not a verdict. Paul's seven letters differ among themselves. A disputed letter that falls outside their spread is unusual for Paul as we have him, but one writer can exceed his own past range when his subject, audience, helper or purpose changes. A difference counts for more when several independent measures agree and when the obvious explanations have been tested.
5. A helper can explain wording; a difference of teaching needs a closer look. A demonstrated contradiction of Paul's teaching would tell against any letter Paul approved. A difference of emphasis, a development, or advice suited to different readers would not. An unapproved departure by a helper would need its own evidence; it cannot explain away a contradiction in a letter Paul supposedly approved. The ancient evidence for secretaries and agents, and its limits, is in Ancient Citation and Authorship Conventions, Part 6, with the evidence that ancient readers condemned letters falsely written in an apostle's name (Part 7) and the limits of computer style tests on short texts (Part 9).
6. Forgers could add personal touches. Ehrman cites Norbert Brox: "Fake personal notes are well known as stylistic devices in ancient pseudepigraphy and do not come as a surprise in the Pastoral epistles either."⁴ So a personal detail counts for authenticity in proportion to how hard it is to explain as a touch added for effect.

Why the six must be judged one by one

Both sides sometimes treat several letters as one block.  McLatchie argues that "the Pastoral epistles must stand or fall together – either all of them come from the hand of Paul or none of them do."⁶  Ehrman says the case against the Pastorals is that "if one of these books is forged, they all are forged."⁴  And Ehrman's case against Ephesians starts from Colossians:  "If Colossians was forged—as it almost certainly was—there can be no thought of Paul himself imitating it in another piece of correspondence."⁴

Johnson objects that grouping distorts the evidence.  Claims made about "the Pastorals" as a whole, he writes, are often false of each letter:  the view that they show a more elaborate church order "is false when applied to either Titus or 2 Timothy. 2 Timothy has nothing on the subject, and Titus has at most a few (not altogether clear) lines."⁵  This article keeps the letters apart.  Where two letters very probably share an author, as 1 Timothy and Titus do, the evidence about one bears on the other, and the article says so.  Where the link is an assumption, it does not.

Part 2: Who used the letters, and who named Paul

The witnesses

1. Polycarp of Smyrna, a bishop in Asia Minor who died about AD 155 aged eighty-six, wrote to the church at Philippi.⁷ His letter may be two letters joined: P. N. Harrison argued that Polycarp "wrote to the Philippians at different times two letters, which were copied out at an early date on to the same papyrus roll, and so fused into one."⁸ On Harrison's view the chapters quoted below come from about 135; on the older view, from about 110. Chapters 10 to 12 survive only in a Latin translation.
2. Marcion, a teacher in Rome in the mid-second century, made a collection of ten letters of Paul. No copy survives; his opponents describe it.
3. Irenaeus of Lyons, about AD 180, quotes widely and names his sources.
4. The Muratorian fragment, a Latin list of accepted books, whose date is disputed between the late second century and the fourth.
5. P46, the oldest surviving collection of Paul's letters, a papyrus book dated by its handwriting to about 200 or the third century.⁹¹⁰

Letter by letter

1. 2 Thessalonians. Polycarp's Latin text says of the Philippians, "who are praised in the beginning of his Epistle. For concerning you he boasts in all the Churches," and a few lines later, "do not regard such men as enemies."⁷ The first is close to 2 Thess 1:4, "we ourselves boast about you in the assemblies of God"; the second to 2 Thess 3:15, "Don't count him as an enemy."² Two echoes of one short letter in one paragraph make it probable that Polycarp knew 2 Thessalonians. And the boast is explicitly Paul's: "he" is Paul, whose letter to the Philippians Polycarp has just mentioned. So if the wording comes from 2 Thessalonians, Polycarp took it as Paul's. Two things weaken this. The words survive only in a Latin translation, which may not keep Polycarp's exact wording. And they are applied to the Philippians, not to the Thessalonians whom 2 Thess 1:4 praises, so Polycarp was either quoting loosely from memory or drawing on some other report of Paul's boasting about his Macedonian churches (2 Cor 8:1-2 praises them). Marcion's collection included 2 Thessalonians: Tertullian complains of an "erasure" Marcion made in it.¹¹ Irenaeus names it, in a passage about "the apostle" whose habits of word order he illustrates first from Galatians: "in the Second to the Thessalonians, speaking of Antichrist, he says".¹² The Muratorian list says Paul "writes once more to the Corinthians and to the Thessalonians".¹³
2. Colossians. No clear echo in Polycarp. Marcion's collection included it.¹¹ Irenaeus names it twice: "in the Epistle to the Colossians: 'Luke, the beloved physician, greets you,'" where the speaker is Paul, and "the apostle, in the Epistle to the Colossians".¹² The Muratorian list names Colossians among Paul's letters to seven churches. P46 contains it.
3. Ephesians. Polycarp quotes it in Greek: "by grace you are saved, not by works" (Eph 2:8-9).¹⁴ And in a Latin passage he joins Psalm 4 to the end of Eph 4:26, introducing both "as it is said in these scriptures, 'Be ye angry and sin not,' and 'Let not the sun go down upon your wrath.'"⁷ The second clause is found only in Ephesians. If "these scriptures" covers both clauses, as the plural suggests, Ephesians was read as scripture by Polycarp's time; it is possible that Polycarp meant only the Psalm and added Ephesians' words from memory. Marcion had it under the title "To the Laodiceans"; Tertullian replies, "We have it on the true tradition of the Church, that this epistle was sent to the Ephesians, not to the Laodiceans."¹¹ Irenaeus names it: "the blessed Paul declares in his Epistle to the Ephesians".¹² P46 contains it.
4. 1 Timothy. Polycarp writes, "the love of money is the beginning of all troubles. Knowing therefore that we brought nothing into the world, but neither have we anything to carry out."¹⁴ These are 1 Tim 6:10 and 6:7, in reverse order, with eight Greek words shared in sequence; his rules for deacons ("not double-tongued, not lovers of money") also echo 1 Tim 3:8.¹⁴ Each saying could be proverbial, but together they make his knowledge of 1 Timothy very probable. Irenaeus opens his great work with the "vain genealogies" which, "as the apostle says," minister questions (1 Tim 1:4), his usual way of quoting Paul; and he writes that Linus is mentioned by Paul "in the Epistles to Timothy", in the plural, though the verse he means is in 2 Timothy.¹²¹² Manning says the Didache "clearly quotes 1 Timothy 3:4."¹⁵ It does not. Didache 15.1 asks for "bishops and deacons worthy of the Lord, meek men, and not lovers of money, and truthful and approved,"⁷ which shares two qualifications with 1 Tim 3:3 and 3:8-10 ("not lovers of money", "approved") but nothing from 3:4, which is about ruling one's household. Two shared words in a list of virtues for church leaders could come from a common stock of such lists as easily as from 1 Timothy, so the resemblance gives little evidence of direct use. It does not establish a quotation, and it cannot securely date 1 Timothy.
5. 2 Timothy. Polycarp says the martyrs "did not 'love this present world'", which echoes Demas, who left Paul "having loved this present world" (2 Tim 4:10), and he writes "we shall also reign with him" (2 Tim 2:12).⁷²¹⁴ These are short phrases, so his knowledge of 2 Timothy is probable rather than certain. Irenaeus names the letters to Timothy as Paul's, as above, and quotes 2 Tim 4:10-11 as Paul's own words about Luke.¹²
6. Titus. No clear echo in Polycarp. Irenaeus: "these men Paul commands us, 'after a first and second admonition, to avoid'" (Titus 3:10).¹² Jerome reports that Tatian, a second-century teacher who "himself rejected several epistles of Paul," held that Titus above all was the apostle's.¹⁶

The Muratorian list adds that Paul wrote "one to Philemon, one to Titus, and two to Timothy" which "are held sacred in the esteem of the Church catholic for the regulation of ecclesiastical discipline."  It rejects letters "to the Laodiceans" and "to the Alexandrians" as "forged in Paul's name".¹³

Marcion and the Pastorals

Tertullian, about AD 208, wonders that Marcion kept Philemon, "this letter which was written but to one man," yet "rejected the two epistles to Timothy and the one to Titus, which all treat of ecclesiastical discipline."¹¹  Jerome, two centuries later, lists 1 and 2 Timothy, Hebrews and Titus among the letters Marcion and others threw out, and complains that they gave no reasons:  "if they gave reasons why they think they are not by the Apostle, we would try to answer them".¹⁶  Clement of Alexandria, about 200, says that "the heretics reject the Epistles to Timothy" because 1 Tim 6:20 condemns "knowledge falsely so called";¹⁷ Ehrman adds Origen's report that some rejected 2 Timothy for naming the magicians Jannes and Jambres.⁴  Where a motive is supplied, it concerns the letters' content.  Jerome's complaint also leaves some rejections unexplained.

"Rejected" is Tertullian's word.  Marcion may have rejected letters he knew, as Tertullian thought, because their approval of the Jewish scriptures, marriage and food did not suit his teaching.  Or his collection may simply not have contained them.  The evidence does not decide which.  What it does show is that by the mid-second century one widely used collection of Paul held the three prison letters and 2 Thessalonians but not the Pastorals.

P46 and the missing pages

P46 now ends in 1 Thessalonians.  Its pages were numbered, and some of its leaves are still joined in pairs, which shows that it was made as a single stack of folded sheets.  Seven leaves are lost from the front, so, on the assumption that the numbering began at the first leaf, seven matching leaves are lost from the back, and Frederic Kenyon, who first edited it, reconstructed the whole as 104 leaves.¹⁸  The University of Michigan, which holds part of it, gives the contents by page:  Ephesians 146-158, Galatians 158-168, Philippians 168-176, Colossians 176-184, 1 Thessalonians 184-191, and lost pages 192-205.  It judges that 2 Thessalonians very likely filled pages 192-195 and that "The remaining ten pages are too small a space to hold the Pastoral Epistles".⁹

That claim can be checked roughly.  Measured in Greek letters of the modern critical text, and taking each letter's page span from the Michigan table, the scribe averaged between 986 and 1,108 letters to a page across the last five surviving letters.  At the highest of those averages 2 Thessalonians fills 3.7 pages, which matches the four lost pages assigned to it.  1 and 2 Timothy and Titus together need about 17 pages at that rate, against ten left.  To fit, the scribe would have had to write about 1.7 times as densely as his average in any of the last five letters.¹⁹  The figures are approximate:  the page spans count shared pages crudely, and the scribe's text abbreviated sacred names and differed in its readings.  But the shortfall is far larger than those errors.  Kenyon made the same check in 1936 from printed lines and reached the same result:  "The space required is therefore nearly twice as much as is available, and by no compression of writing is it conceivable that these four Epistles could have been included."  He left two possibilities open:  "The last five leaves may have been left blank, or some additional leaves may have been attached at the end so as to take the Pastoral Epistles".¹⁸

So the conclusion is conditional, and its premise is now disputed.  The space argument assumes that the book ended at 104 leaves, a number reached from the page numbers.  Jeremy Duff argued in 1998 that the scribe was squeezing more text onto each page toward the end and probably added leaves to fit all fourteen letters.  That part of his case has not held up:  as Brent Nongbri summarises later work by Eldon Epp and Edgar Ebojo, "the phenomenon that Duff emphasized (more text in the second half of the codex) is real, the trend is not as pronounced or as regular as Duff made it appear," and Nongbri found that the ancient books Duff cited do not show leaves being added in this way.  But Nongbri, surveying better-preserved single-quire books, raised a different doubt.  Several begin with blank, unnumbered leaves used as flyleaves or pasted into the cover, so "Page numbering of single-quire codices is not necessarily an exact guide to the total number of bifolia that originally made up the quire."  If P46 had such leaves at the front, it had more leaves at the back than Kenyon counted.  Nongbri therefore allows that Kenyon's blank final leaves are "plausible", but also "that we have underestimated the size of the quire", and concludes:  "We cannot say, for instance, that the Beatty-Michigan codex is secure evidence for the circulation of a ten-letter collection of Paul's letters".²⁰²⁰²⁰

So P46 is weak evidence.  On Kenyon's reconstruction, one Egyptian copy of about 200 was planned without the Pastorals;  on Nongbri's caution, the number of lost pages, and so what they held, is unknown.  Even on Kenyon's reconstruction it would be evidence about how Paul's letters were collected, not about when the Pastorals were written:  P46 may also have lacked Philemon, which nobody doubts and which would have fitted.  Letters to individuals may have circulated separately and joined the church collections later, or the Pastorals may have been later letters.  P46 cannot tell these apart.

What reception proves

1. For all six letters, naming is unanimous once naming begins. By about 180 every one of the six was quoted as Paul's. No surviving early writer gives a reason of style or history for doubting any of them. Some reported rejections concern content; others come without a reason.
2. The naming is not early enough to settle the question by itself. Irenaeus wrote about 115 years after Paul's death. Ehrman points out that forgeries could gain acceptance quickly: ancient authors complained of forgeries "circulating in their own names, even within their own lifetimes".⁴ He adds that early use "scarcely can be counted as evidence for Pauline authorship", because a reader a generation later would take a letter's claim at face value.⁴ So unanimous naming by 180 is compatible with letters written in Paul's name after his death.
3. But silence about doubts of authorship is not empty. Ancient readers did test authorship. The same Muratorian list that accepts the six rejects two letters as forged in Paul's name, and ancient readers doubted Paul's authorship of Hebrews and Peter's of 2 Peter on grounds of style. Ernest Best, who thinks Ephesians was not written by Paul, puts the question plainly: "why was it that the non-Pauline authorship of Ephesians was not detected and the letter rejected as 3 Corinthians and the Epistle to Laodicea were?"¹ Johnson makes the same point for the Pastorals: they "were never challenged on the basis of style in the ancient church, even though criticism of Hebrews on this very point was not lacking."⁵
4. Each letter has its own bound. Ephesians, 1 Timothy and probably 2 Thessalonians and 2 Timothy were in use by about 135, which rules out an origin in the later second century for those four. Colossians is attested by the 140s. Titus is attested directly only by about 170, unless it shares 1 Timothy's author, as it very probably does. None of these bounds says how much earlier a letter was written.
5. The collections divide the six into two groups. 2 Thessalonians, Colossians and Ephesians are in Marcion's collection and P46 (2 Thessalonians very probably, in its lost pages). The Pastorals are absent from Marcion's collection and may have been absent from P46, though Polycarp probably knew 1 and 2 Timothy. Marcion's pattern counts somewhat for the first three and somewhat against the Pastorals, without deciding either; P46's adds little.

Part 3: What the language shows

Simple counts, run on all thirteen letters

Arguments about style usually trade examples.  This article instead ran a few simple counts on the Greek text of all thirteen letters, using only the seven accepted letters as the yardstick and never fitting anything to the disputed six.²¹  The counts are descriptive.  They show where each letter falls against Paul's own spread;  they are not a statistical test, and a letter outside Paul's spread is unusual, not proved to be someone else's.

1. Connecting words. Walter Bujard's 1973 study of Colossians, which Ehrman summarises as "widely thought to be unanswerable," rests largely on how often a writer uses conjunctions: words such as "but", "for", "so that" and "therefore".⁴ Ehrman reports Bujard's totals: Galatians 239 (10.7 percent of its words), Philippians 138, 1 Thessalonians 126, Philemon 28, "but Colossians only 63 (4 percent)."⁴ This article counted every word the computer tagging marks as a conjunction, leaving out "and" (καί). That is a related measure, not Bujard's own: his book was not read here, and he may classify some particles differently. The totals come out close to his: Galatians 249, Philippians 132, 1 Thessalonians 129, Philemon 29, Colossians 72. So do the infinitives (Galatians 32, Philippians 39, 1 Thessalonians 45, Colossians 11, against his 32, 39, 48 and 11) and the relative pronouns, which track his relative clauses to within two-tenths of a percentage point in every letter. Two of his counts do not reproduce, because the definitions differ: the number of different conjunctions (Galatians 22 here, 33 in Ehrman's report) and the articular infinitive (Philippians 11 here, or 16 by a looser search, against his 16; Galatians 3 against his 5).⁴ On every one of these counts, though, Colossians stands apart from Galatians, Philippians and 1 Thessalonians, the letters Bujard compared it with, and like Bujard this count finds no articular infinitive in it. So the pattern Ehrman reports is not an artefact of one scholar's counting, though this is a partial check, not a reproduction of Bujard's study.
2. Vocabulary outside Paul. For each letter, the count is the share of its words whose dictionary form never appears in Paul's accepted letters. For an accepted letter the yardstick is the other six, so no letter is measured against itself. Proper names are left out. To compare long and short letters fairly, each letter is cut into consecutive 600-word blocks, and every possible 600-word stretch is also measured.
3. Strings of genitives. Ehrman reports Daryl Schmidt's finding that Colossians, Ephesians and 2 Thessalonians pile up strings of nouns in the genitive ("of ... of ...") far more than Paul does.⁴ This article counted a simpler version: a noun followed by at least two genitive nouns or pronouns.

"Words" is every Greek word;  the vocabulary rates leave out proper names, so their denominators are slightly smaller (Titus, for example, has 641 words that are not names).  Philemon's genitive figure rests on six strings in 334 words and says little.

What the vocabulary count shows

The six fall into two groups, with 2 Timothy mostly in the second.

1. 2 Thessalonians, Colossians and Ephesians use words Paul uses, at rates inside the spread of his accepted letters, whether measured whole or stretch by stretch. The conclusion changes little when each is measured against only six of the accepted letters at a time, so that it faces a yardstick the same size as each accepted letter does (2 Thessalonians 4.3 to 5.1, Colossians 6.1 to 7.6, Ephesians 6.0 to 7.2, against accepted letters' 4.2 to 7.5). Colossians then edges just above that whole-letter range in one comparison.
2. 1 Timothy and Titus do not. Every 600-word stretch of them, from 12.7 to 17.3 words in a hundred, lies above every 600-word stretch of the six accepted letters long enough to have one, whose highest is 12.3, in Romans. The margin between 1 Timothy's lowest stretch and Romans' highest is small; the difference between the letters as wholes (15.3 and 14.5 against at most 7.5) is not.
3. 2 Timothy lies in between, but closer to 1 Timothy. Its opening 600 words (9.8) fall inside Paul's range, but 499 of its 567 possible 600-word stretches lie above Paul's highest, and its whole-letter rate (11.4) is well above every accepted letter.

This describes the text we have, not every passage Paul ever wrote.  A writer who turned to a new subject, wrote to a different kind of reader or used a new helper could exceed the range of his surviving letters.  That is the defenders' strongest reply, taken up below.

Other studies show why the definition of the measure matters.  Jermo van Nes counted words or senses confined to one of the thirteen letters, even when they occur there more than once, and corrected for length with a regression on the seven accepted letters.  His initial analysis put Colossians, Ephesians, 2 Thessalonians and Titus inside Paul's predicted range, with only 1 and 2 Timothy above it.²²  But that was not his final result.  After excluding proper names, words in explicit quotations, and compounds formed by adding prepositions to words used elsewhere in the corpus, all six disputed letters fell inside the predicted range.²²  The last adjustment rests on the reasonable possibility that a writer can form such compounds without becoming a different author.  Van Nes's finished analysis therefore gives less support to a different writer for the Pastorals than his initial count does.

Luke Timothy Johnson, defending the Pastorals, judged that "The vocabulary in 2 Timothy, taken as a whole, is closer to that of the undisputed letters (taken as a whole), whereas the vocabulary of 1 Timothy and Titus diverges more dramatically."⁵  The count here agrees with that relative ordering.  It differs from van Nes's adjusted result because it counts occurrences of words absent from the seven accepted letters, including words shared among the Pastorals, and does not remove quotations or prepositional compounds.  It describes a real difference in vocabulary;  it does not overturn his result or establish a different author.  Taken together, the studies show that the strength of the vocabulary argument depends partly on which differences are treated as relevant.

The old form of the argument, that the Pastorals use second-century words, is weaker than the plain difference from Paul.  P. N. Harrison found that of 848 words in the Pastorals other than names, "306 or over 36 per cent. are not to be found in any one of the ten Paulines."³  He also argued that their vocabulary was that of second-century Christian writers.  Donald Guthrie replied, in the passage McLatchie quotes, that "Before such parallels can prove common vintage, it is necessary to show that the words in question could not have been used in the first century," and that almost all of them were in use earlier.²³  That reply is sound for the date of the words.  It does not touch the distance from Paul, which is what the counts measure.

McLatchie's asymmetry argument

McLatchie argues that likeness and unlikeness do not weigh the same:  "whereas a high level of verbal similarity between two pieces of writing does indeed provide strong evidential support for a literary relationship ... verbal dissimilarity between two writings offers only comparatively weak evidence".⁶  He points out that Ehrman uses similarity to prove that one person wrote 1 Timothy and Titus, then counts difference against Paul.  And he calls it "heads I win; tails you lose" to treat likeness to Paul as a forger's craft and unlikeness as proof of forgery.⁶  Manning adds that Ehrman himself warns, "Probably not too much stock should be placed in mere numbers."²⁴

McLatchie's principle is right, and the counts here do not overturn it.  Shared rare wording is hard to get by chance;  missing wording has many innocent causes.  What the counts add is a scale:  they show that the distance of 1 Timothy and Titus from Paul's vocabulary is larger than the spread among his seven letters, measured on stretches of equal length.  That makes a different author a reasonable explanation.  It does not make it the only one, because the measurement cannot separate a different author from the same author writing on new subjects, to delegates rather than congregations, or through a different helper.  Those causes are confounded with authorship here, and no count of words can pull them apart.

His second point cuts both ways.  A critic who explains every likeness to Paul by imitation and every difference by forgery cannot be refuted;  neither can a defender who explains every likeness by Paul and every difference by a secretary.  The way out for both is to ask what each explanation predicts that the other does not, and to look for measures that agree with each other.

What the conjunction count shows

Colossians uses connecting words at 4.6 percent of its words.  Paul's accepted letters range from 8.1 to 12.6.  The usual reply is that subject matter changes style:  a prayerful or ethical letter needs fewer "for"s and "therefore"s than an argument.  A simple test of that reply goes against it for Colossians.  Paul's own exhortation sections stay high (Romans 12-16, 8.7;  Galatians 5-6, 10.7;  1 Thessalonians 4-5, 9.6), while both halves of Colossians are low (chapters 1-2, 4.1;  chapters 3-4, 5.2).  The test is crude.  Chapter ranges do not match Colossians' hymn-like passage in chapter 1 or its traditional lists of duties, and three sections of Paul are a small sample.  It weakens the simple explanation "Colossians is low because it is ethical and devotional";  it does not rule out every explanation from genre.

Every other New Testament letter was counted too:  Hebrews 7.9, James 7.4, 1 Peter 9.4, 2 Peter 6.1, 1 John 7.3, Jude 5.9, and the two short letters of John 5.7 and 6.4.  So the measure does not separate Paul from other writers well, since 1 Peter matches him.  Colossians is lower than every one of them.  Ephesians' low overall figure comes from its first three chapters (2.9), a long prayer and praise;  its last three (7.8) are just below Paul's range.  2 Thessalonians (7.8) and the Pastorals (5.9 to 7.4) sit a little below Paul's range, among the other New Testament writers.

What the genitive count shows

On this measure 2 Thessalonians (17.1), Colossians (12.7) and Ephesians (19.0) are above every accepted letter except Philemon, whose figure rests on six cases.  Without Philemon, Paul's letters run from 3.5 to 11.7.  The Pastorals (6.5 to 9.1) are inside Paul's range.  This agrees in direction with Schmidt's figures as Ehrman reports them (Paul 7.4 to 15.2 strings per thousand words, Colossians 29.7, Ephesians 31.7, 2 Thessalonians 26.7), though the two counts are defined differently.  Ehrman himself adds the caution:  "In isolation this kind of stylistic demonstration can carry little weight. Authors can and do vary their style".⁴

So the measures do not all agree.  For 2 Thessalonians, vocabulary fits Paul, connecting words are just below his range, and genitive strings are well above it.  A larger study that combined twenty-five measures, by David Mealand in 1995, found the reverse of a clear separation:  as Paul Foster reports it, "When 2 Thessalonians was specifically subjected to discriminant analysis as a doubtful sample it was decisively classed with Paul."²⁵  Single features can mislead;  so can a combination trained on a few long letters.  For Colossians, by contrast, connecting words, infinitives, relative pronouns and genitive strings all depart from Paul in the same direction.

What language cannot settle

1. It cannot name the writer. A count can show that a letter does not write like Paul's other letters. It cannot tell Paul writing through Timothy from a disciple writing after his death. Kenneth Neumann, whose statistical study placed Ephesians, Colossians and 2 Thessalonians with Paul but the Pastorals elsewhere, said so himself: such letters might come from "Paul's amanuenses or co-senders writing predominantly under or with the apostle's guidance, or they may be later Pauline students."²⁶ Nor can a different rate of connecting words identify which co-worker, if any, chose the words.
2. Short letters are noisy. Titus has 659 words and 2 Thessalonians 820, so each gives a single block and only a narrow band of overlapping stretches.
3. Subject matter matters. Johnson notes that Paul uses "law" seventy-two times in Romans and three in Philippians.⁵ The 600-word stretches control for length, not for subject, audience or helper. The Pastorals' subjects, church leaders and false teachers, and their readers, single delegates, are not Paul's usual ones. Paul's own most unusual stretches still stay below them, but Paul's surviving letters contain no letter to a delegate to compare.

Part 4: 2 Thessalonians

What the letter says about itself

It opens with the same three senders as 1 Thessalonians:  "Paul, Silvanus, and Timothy, to the assembly of the Thessalonians".²  It answers a claim "that the day of Christ has already come," which had reached the readers "either by spirit or by word or by letter as if from us".  It reminds them, "Don't you remember that when I was still with you, I told you these things?"  And it ends:  "I, Paul, write this greeting with my own hand, which is the sign in every letter. This is how I write."²

The case against

Ehrman, in his scholarly study Forgery and Counterforgery, sets aside weaker arguments and presses two:  "(1) the impressive parallels to 1 Thessalonians ... which, when examined closely, seem virtually inexplicable on grounds other than that a second author (not Paul) has used Paul's letter as a model for his own; and (2) the substantive differences from 1 Thessalonians in precisely the passage ... where the parallels evaporate."⁴

1. Copying, of words and of layout. The openings are almost identical, the closing blessings share nine words, and some sentences repeat word for word: 1 Thess 2:9 and 2 Thess 3:8 both have "working night and day, that we might not burden any of you".² Ehrman lists phrases found in these two letters and nowhere else in Paul, and he asks: "Is it likely that Paul remembered to the very word what he said at times in his earlier letter, creating verbatim agreements ... extending sometimes to sequences of nine, ten, or more words?" Paul's other pairs of letters, Galatians and Romans, 1 and 2 Corinthians, show "nothing like this phenomenon".⁴ The likeness goes beyond wording: "Both, unlike any other Pauline letter, have two thanksgivings (instead of one), one at the outset and one in the body of the letter," and every major theme of 2 Thessalonians is already in 1 Thessalonians.⁴²
2. Style. Ehrman adds Schmidt's counts, which find a more complex sentence structure in 2 Thessalonians than in Paul, its supposed model included (Part 3).⁴
3. The end. 1 Thessalonians says "the day of the Lord comes like a thief in the night."² 2 Thessalonians says it will not come "unless the rebellion comes first, and the man of sin is revealed". For Ehrman, "for this other author, the end will not come suddenly, without advance warning, like a thief in the night. There will in fact be plenty of warning".⁴
4. The signature. Ehrman compares 3:17 with Paul's own handwritten greetings. Its opening words "are precisely parallel to 1 Cor. 16:21, and are matched still more famously by Gal. 6:11. But in neither of these instances is Paul's final signing off ... said to be a mark of authenticity." So "If this was the way Paul always finished his letters, it is more than a little strange that he does not end all his letters this way, and stranger still that he never refers to the practice as a prophylaxis against forgery." And the claim is "an odd claim to make in a (Christian) culture where letters circulated more commonly in copies rather than autographs."⁴²
5. A counterforgery. Ehrman reads 2:2 and 2:15 together. The letter the readers should hold to is 1 Thessalonians, which the writer accepts and copies; the letter "as if from us" is some other letter in Paul's name that taught the end was at hand. Ehrman does not claim to know that such a letter existed: "this lost letter—whether it ever existed or not cannot be known". The writer said it did: "Or so the forger averred."⁴ On this reading one writer, after Paul's death, both imitated 1 Thessalonians and warned against forgeries in order to make his own letter believable: "He countered a forgery by producing a forgery."⁴

Ehrman's account is economical.  It needs one later writer, who knew 1 Thessalonians, disagreed with an expectation of an immediate end, and used the tools forgers used.  It does not need a second forger or a real forged letter, since the writer may have invented the threat.

The case for, and how the objections fare

1. The copying favours imitation, but not decisively. This article's sister study counted the five-word sequences that pairs of letters share, leaving out formal openings and closings. 2 Thessalonians shares 12.4 per thousand of its words with 1 Thessalonians. Romans and Galatians, both certainly Paul's, share 8.0; Colossians and Ephesians 14.8; 1 and 2 Corinthians 2.1. And 1 Timothy and Titus, which both sides agree share an author, share none at all.²⁷ So a high rate is not what one author normally produces, but Paul can repeat himself when he writes on the same themes. That is one measure. Ehrman's argument also rests on the shared layout, the two thanksgivings and the matching themes, which a phrase count does not capture. On the traditional view one team wrote twice to one church about one crisis within weeks, perhaps with a copy of the first letter at hand. That would explain repeated themes and layout, but the close verbal reuse is still more than Paul shows elsewhere. The overlap is better predicted by imitation than by authenticity; how much better depends on how likely it is that the senders reread their first letter, which nobody can measure.
2. The end: sudden is not the same as unannounced. 1 Thessalonians itself goes on: "But you, brothers, aren't in darkness, that the day should overtake you like a thief."² Ehrman grants that "It may fairly be objected that the thief image of 1 Thessalonians speaks not of an absolutely imminent appearance of the Lord, but of a sudden appearance," and replies that the image still clashes.⁴ Manning answers from the Gospels: Jesus' sayings in Matthew 24 combine the thief image with a list of signs that come first, so one teacher could hold both.²⁸ The point does not need Matthew's date. Mark 13 already sets "the end is not yet" beside "of that day or that hour no one knows".² 2 Thessalonians says it is correcting a specific claim that the day "has already come", and a correction of that claim is what a pastoral letter from the same team would say. Foster, who thinks Paul wrote both letters, calls it a real change, Paul "reformulating his eschatological scheme" when his first teaching produced idleness.²⁵ The tension fits a development within one teacher; it also fits Ehrman's later writer. It does not decide.
3. The signature had a known use, and Paul's own letters show the habit. In 1 Corinthians and Galatians Paul writes the close of the letter in his own hand, and in Philemon a pledge; 1 Corinthians 16:21 uses the same opening words as 2 Thess 3:17.² Chris Keith's study shows that a sender's own hand was an ordinary way to authenticate a letter, and that its absence sometimes had to be explained.²⁹ Bishop Cyprian of Carthage, in the third century, sent back a suspect letter with the request: "ascertain whether the writing and subscription are yours".³⁰ Three distinctions matter here. First, the check worked for the first readers, who held the original; later copies could not show it, as Ehrman says, but a genuine letter was written for its first readers. Second, "the sign in every letter" need not mean that every letter mentions the handwriting. Foster suggests Paul may have added his greeting by hand at the end of every letter and mentioned it only when it mattered.²⁵ Third, a writer who has just warned against a "letter as if from us" has a reason to mention it here.
4. Foster's argument from the signature. Paul Foster, of Edinburgh, argues that a later writer would have needed to know Paul's habit across many letters to claim it for "every" letter, and that copies of those letters would not have shown it. He notes that "the only Pauline epistle that appears to be known by 2 Thessalonians is in fact 1 Thessalonians, where there is no reference to Paul making a sign in his own hand," and concludes that a forger "would not have had access to the resources of a Pauline letter collection until the second century on which to base the claim".²⁵ The argument has force, but it rests on two inferences: that the writer knew no other letter of Paul because none is visibly echoed, and that a forger needed evidence before asserting a habit. A forger could know 1 Corinthians without echoing it, and Ehrman's point is precisely that a bold, uncheckable claim served a forger well. Foster's conclusion, "it is difficult to explain the letter signature except on the grounds that it is a genuine Pauline feature", is stronger than these premises.²⁵
5. The setting. Manning asks why a writer decades later would address idle Thessalonians still waiting because of a letter Paul sent long before: "Wouldn't it be more logical for Paul to clear up any confusion himself shortly after writing his initial letter to them?"³¹ Foster quotes Karl Donfried, who rejects Paul's authorship, making the same point.²⁵ Ehrman's answer is that the address is part of the fiction; the real readers were elsewhere and later. Both readings explain the letter's contents. The specific, local character of the problem fits a real letter more naturally, but a skilled imitator would aim at exactly that.
6. "I told you these things". William Paley argued in 1790 that the obscure passage about "what withholdeth" makes sense only to people who heard Paul's earlier teaching: "No man writes unintelligibly on purpose."³² That argument is weaker than it looks. Apocalyptic writers often were deliberately cryptic: Mark 13:14 adds "let the reader understand," and Revelation sets a riddle in a number.² The obscurity fits a genuine letter, but an imitator writing in the style of apocalyptic would produce it too.
7. Language. The vocabulary is Paul's, on this article's count and on van Nes's. The connecting-word rate is a little below Paul's lowest (7.8 against 8.1), and genitive strings are well above his range (Part 3). Mealand's combined measure classed the letter with Paul. So the language is mixed: it gives no clear reason to doubt the letter, and no clear reason to trust it.
8. Reception. Marcion's collection held it, Irenaeus names it, and Polycarp probably knew it and took its words as Paul's. It very likely filled the lost pages of P46.

Verdict on 2 Thessalonians

The two accounts both explain the letter.  Ehrman's later writer explains the close copying better;  the traditional account explains the early reception, Polycarp's attribution and the specific setting more naturally, and needs only that Paul's team wrote again soon on the same subject.  The language does not choose between them, and the change in teaching about the end fits either.  Foster, a critical scholar, and James Dunn, who sides with the critical majority on Ephesians and the Pastorals, both regard 2 Thessalonians as Paul's.³³²⁵  Paul's in the historical sense, sent in his lifetime with Silvanus and Timothy:  probably.  Confidence:  low to moderate.  Whose wording, Paul's or a co-sender's, the evidence cannot say.  A later imitation of 1 Thessalonians remains a serious alternative.

Part 5: Colossians

The Philemon network

Colossians and Philemon, which nobody doubts, present themselves as written at the same moment.  In both Paul is a prisoner and Timothy is co-sender (Col 1:1;  Phlm 1).  In both Onesimus is going back:  Philemon is Paul's appeal for him, and Colossians sends him with Tychicus as "the faithful and beloved brother, who is one of you."  Archippus is greeted in Philemon and given a charge in Colossians:  "Tell Archippus, 'Take heed to the ministry which you have received in the Lord, that you fulfill it.'"  Five of the six men who send greetings in Colossians send them in Philemon:  Aristarchus, Mark, Epaphras, Luke and Demas.²

Paley drew the argument in 1790 that McLatchie and Manning repeat.  Philemon never says where Philemon lived.  Colossians never mentions Philemon.  Only by joining Onesimus to Philemon, Philemon to Archippus in Philemon's greeting, and Archippus to Colossae in Colossians 4:17, does a reader learn that Onesimus was indeed "one of you".  Paley:  "this result is the effect, either of truth which produces consistency without the writer's thought or care, or of a contexture of forgeries confirming and falling in with one another by a species of fortuity of which I know no example."  He then argues that design is ruled out:  the purpose of such a design, to confirm that Onesimus was "one of you", "would be lost upon ninety-nine readers out of a hundred," and the means are "too circuitous to have been the subject of affectation and contrivance.  Would a forger, who had this purpose in view, have left his readers to hunt it out".³²  McLatchie concludes that "This undesigned coincidence supports the Pauline authorship of the epistle to the Colossians."³⁴  Manning adds a detail of his own:  Colossians calls Aristarchus "my fellow prisoner," where Philemon gives that title to Epaphras, and "A forger would've made sure there was more of an alignment in terms."³⁵

The critics' answer is that Colossians was written with Philemon to hand.  Eduard Lohse, who thinks a disciple wrote Colossians around AD 80, says:  "Only in the case of the letter to Philemon is literary dependence obvious, for Phlm's list of greetings was used and expanded in Col."³⁶  Ehrman quotes Victor Furnish:  either one author wrote both, or Colossians was written by someone imitating Paul "with the letter of Philemon to hand."⁴

Weighing the two:

1. The coincidence shows coherence, not an independently confirmed fact. Nothing outside Colossians says where Philemon lived. The town comes from Colossians itself. So a later writer using Philemon did not have to "guess right": whatever town he chose for Onesimus and Archippus, the two letters would agree, because Philemon names no town. The coincidence shows that the two letters fit together, not that Colossians knew something true that its writer could not have got from Philemon.
2. Paley's argument, fairly stated, is about design, and it has force against a borrower too. Paley's point is that the fit is too roundabout to have been planted to confirm "one of you". A modern critic does not think the writer of Colossians was trying to confirm Philemon, so that form of the point misses him. But a borrower working from Philemon made choices that a writer aiming at a convincing Pauline setting would not obviously make. He left out Philemon himself, the man his source is written to, and Apphia, greeted beside him. He sent Onesimus home as a "faithful and beloved brother" without a word of the trouble Philemon is written to settle. He added details that are not in Philemon: that Mark is "the cousin of Barnabas", that the readers had received instructions about him, a sixth greeter, "Jesus who is called Justus", and a charge to Archippus.² A real writer sending both letters by one carrier explains all of this without effort: Philemon got his own letter, and the private matter stayed private. A borrower can explain it too, as a writer careful not to copy too obviously. The real writer explains it more simply.
3. The variations add little by themselves. The moved title "fellow prisoner" and the sixth greeter look like a second letter written at the same moment, as Manning says. But Ehrman holds that forgers added such touches, and a careful imitator would not copy a list word for word. These small differences are consistent with a lifetime origin; they do not establish it.
4. Lohse's "obvious" dependence is an inference. The two lists share names and a couple of titles, applied differently, not sentences. Shared names are what both accounts predict. What would show copying is shared wording or a borrowed mistake, and the lists show neither.

The Philemon network is the best evidence that Colossians belongs to Paul's lifetime, but it is evidence of coherence and of choices that fit a real writer better than a borrower.  It is not an independent check, and it is not proof.

Style

Here the case against Paul's own wording is strong.  The figures Ehrman reports from Bujard come out close on this article's related count, and several measures depart from Paul in the same direction:  few connecting words, few infinitives, many relative pronouns and long strings of genitives (Part 3).  The simple reply that subject matter explains them does not survive a first test:  both the teaching half and the ethical half of Colossians are lower in connecting words than any section of Paul tested or any other New Testament letter.  Ehrman puts the conclusion bluntly:  "This book is not written in Paul's style."⁴  Its vocabulary is Paul's;  its way of building sentences is not.  Neumann's broader statistical study, though, assigned Colossians to Paul's group, if far from Paul's centre,²⁶ a reminder that style measures do not all agree;  the sister article, Part 9, sets out that study's limits.

Teaching

Lohse finds that "Pauline theology has undergone a profound change in Col, which is evident in every section of the letter and has produced new formulations in Christology, ecclesiology, the concept of the apostle, eschatology, and the understanding of baptism."³⁶  The clearest example is the resurrection.  Romans 6 says that the baptized were buried with Christ "that ... we also might walk in newness of life," and "we will also be part of his resurrection," still in the future.  Colossians says "you were also raised with him through faith" (2:12) and "If then you were raised together with Christ, seek the things that are above" (3:1).²  Ehrman:  "Maybe he changed his mind. But given the stylistic differences—and the other matters of content to be discussed—it seems unlikely."⁴

The difference is real but narrower than it first looks.  Colossians keeps a future hope:  "When Christ, our life, is revealed, then you will also be revealed with him in glory" (3:4).  Romans already speaks of believers as "alive to God" now (6:11) and as already "glorified" in God's purpose (8:30).²  Colossians says in so many words what Romans says only in part.  That could be a disciple pushing Paul's thought further, or Paul, or a helper he approved, applying it to readers tempted by a "philosophy" that offered higher experiences.  It is a development, not a contradiction, so it does not rule out a letter Paul approved.  The teaching alone does not decide.

Who chose the words?

Since 1857 some scholars have proposed that Timothy, named as co-sender, wrote Colossians for Paul.  Dunn is the best-known:  he regards Colossians "as probably written by Timothy before Paul's death".³³  Ehrman calls the proposal "completely implausible," because the letter twice says "I, Paul" and claims Paul's own hand (1:23;  4:18), and because Paul would have had to sign off on its content, "Yet it is precisely the contents that dissociate the letter from Paul."⁴

Neither objection is decisive.  A letter drafted by a co-worker and approved and signed by the sender is exactly what "I, Paul, write this greeting with my own hand" (4:18) would look like.  And the content, as just shown, develops Paul's teaching rather than contradicting it.  A helper's draft would explain both the strongest evidence for a lifetime origin and the strongest evidence against Paul's own wording.  But that is a hypothesis fitted to the evidence, and its weaknesses should be stated plainly.  No text says anyone drafted Colossians for Paul.  Timothy's place as co-sender makes him the obvious candidate, but being named as co-sender is not evidence that he chose most of the words, and other helpers are possible.  And in the evidence surveyed here, no ancient case is documented of a helper composing a long letter of this kind in another man's name.

Verdict on Colossians

The external evidence is good:  Marcion, Irenaeus, the Muratorian list and P46 all have it.  The vocabulary is Paul's.  The links with Philemon cohere in ways that fit a real writer of Paul's lifetime better than a borrower.  The style is not Paul's, and the teaching has developed.  Paul's in the historical sense, sent in his lifetime:  more likely than not, with low confidence.  Paul's own wording:  probably not, with low to moderate confidence;  if the letter is his, a helper probably shaped it, and Timothy is a candidate.  A disciple writing after Paul's death, as Lohse and Ehrman hold, remains a serious alternative, and the style is its strongest support.

Part 6: Ephesians

A letter to whom?

Most printed Bibles begin "to the saints who are at Ephesus".²  But the words "at Ephesus" are absent from P46, the oldest copy, and from the first hands of Codex Sinaiticus and Codex Vaticanus, and Origen's text lacked them.³⁷  Marcion's collection called the letter "To the Laodiceans";  Tertullian insisted it was to the Ephesians.¹¹  Colossians, meanwhile, tells its readers to read "the letter from Laodicea" (4:16).²

Some phrases read as if Paul and the readers had not met.  He writes "having heard of the faith in the Lord Jesus which is among you," and "if it is so that you have heard of the administration of that grace of God which was given me toward you".²  Acts has Paul at Ephesus for two to three years (19:10;  20:31).²  How much these phrases show is debatable:

1. "Having heard of your faith" does not imply strangers. Paul writes the same thing to Philemon, his own convert: "hearing of your love and of the faith which you have toward the Lord Jesus".²
2. "If indeed" (εἴ γε) need not express doubt. Paul uses the same words for things he takes to be true or hopes are true: "if it is indeed in vain" (Gal 3:4), "if indeed being clothed" (2 Cor 5:3), and Colossians has "if it is so that you continue in the faith" (1:23).² Ephesians uses it twice of what the readers "heard" (3:2; 4:21). It can be read as "surely you have heard". It reads more naturally to readers who had heard about Paul's commission rather than received it from him, but it does not prove they had never met him.
3. The address is uncertain. The oldest copies lack a place name. A letter for several churches in the province, as many conservative and critical scholars propose, would explain both the missing name and the general tone. So would a letter whose original address was lost or altered in copying. Ehrman himself sets aside the general address as an argument: "Was Paul for some reason not allowed to write a circular letter if he so chose?"⁴

So a general or circular letter is a reasonable reading, not an established fact.  The address bears on who the readers were;  it does not show who wrote the letter.

Ephesians and Colossians

Among the pairs of letters in the sister study, Colossians and Ephesians share the most five-word sequences (14.8 per thousand words of the shorter, leaving out openings and closings).²⁷  Andrew Lincoln, as Ehrman reports him, finds that 34 percent of the words of Colossians reappear in Ephesians.⁴  The closest passage is the note about the carrier.  Of the 34 Greek words of Colossians 4:7-8, 32 reappear in the same order in Ephesians 6:21-22;  only "and fellow bondservant" is missing, and Ephesians adds "how I am doing".²¹  Manning, following Craig Blomberg, calls this "thirty-two consecutive words that are identical";³⁵ they are identical and in order in the SBL text, but not consecutive:  the longest unbroken run is 19 words, because "and fellow bondservant" interrupts them.  The correction is small and does not weaken Blomberg's real point, which is that the two notes are nearly identical.

How the two letters are related is a separate question from whether they are related.  There are four main possibilities, and each has defenders:

1. Ephesians used Colossians. This is the commonest view.
2. Colossians used Ephesians. Ernest Best, who thinks Paul wrote neither, judged that the case for the first direction is not proved: "There is a relationship with Colossians but it cannot be proved that AE used that letter; the customary argument that his use of it would imply he was not Paul cannot then be sustained." On the Tychicus passage itself he thought "a slightly better case can be made" for Colossians using Ephesians.¹¹
3. Two writers from one Pauline circle, which Best thought "more probable than the use of one letter by the author of the other".¹
4. One sender writing both at one time, as the shared carrier, Tychicus, suggests. Ephesians 6:21 even begins "that you also may know my affairs,"² which some interpreters, as Best reports, read as "you, my readers, as well as the Colossians": a slip by someone who had just written Colossians. Best thinks that reading unlikely, but it shows that the overlap fits common authorship as easily as copying.¹

The verdict on Ephesians does not simply follow from the verdict on Colossians.  If Ephesians used a Colossians written after Paul's death, Ephesians is later still.  But if Colossians is Paul's, Ephesians could still be a later imitation of it, as Ehrman's own logic allows for 2 Thessalonians.  And if Colossians used Ephesians, Ephesians is the earlier letter, whatever is decided about Colossians.  Each conditional depends on a direction of borrowing that the evidence has not settled.

Paley's chain

Paley argued that Ephesians 6:20, where Paul is literally an ambassador "in a chain" (ἐν ἁλύσει), fits Paul's Roman custody exactly, chained to one soldier (Acts 28:16, 20), while the parallel in Colossians uses a general word for being bound.  A forger who planted the detail in one letter, he asked, "why did he not do it in the other?"³²  Manning repeats the argument and adds that the word for "chain" (ἅλυσις) appears in only two places, Ephesians 6:20 and Acts 28:20.³⁸  That is wrong.  The word occurs eleven times in the New Testament, in ten verses.  The singular, a single chain, is used of Paul at Acts 28:20, Ephesians 6:20 and 2 Timothy 1:16 ("not ashamed of my chain"), and also of the man in the tombs (Mark 5:3) and of Satan (Revelation 20:1).³⁹  In a 2023 version Manning gives the Greek word as "desmios", which is a different word, "prisoner", used of Paul in Ephesians 3:1 and 4:1, in Philemon and in 2 Timothy.³⁵  What survives the correction is Paley's narrower point:  Ephesians uses the singular "chain" in a letter from custody, as Acts describes Paul's Roman custody.  That fits Paul's situation, but Paley's own claim that the singular is used of "no other kind of custody" is too strong, and the detail is slight evidence about who wrote the letter.

Style and teaching

1. Long sentences. Ephesians 1:3-14 is one sentence in the Greek, and Ehrman reports Morton and McLeman's count that nine of a hundred sentences in Ephesians exceed fifty words, against one or three in passages of Paul of similar length.⁴ Those counts depend on modern editors' punctuation, a weakness set out in the sister article. The connecting-word count avoids that problem: Ephesians 1-3, a blessing and prayers, is very low (2.9 percent); Ephesians 4-6, the instructions, is just below Paul's range (7.8). That pattern fits a change of genre inside one letter as well as a change of author. Genitive strings, by contrast, are high throughout (Part 3).
2. Vocabulary. Inside Paul's range on the count here and on van Nes's. Best: "the number of hapax legomena in Ephesians is not exceptional and therefore tells us nothing in respect of its authorship".¹
3. The resurrection. God "raised us up with him, and made us to sit with him in the heavenly places in Christ Jesus" (2:6).² This goes a step beyond Colossians. The same answer applies as for Colossians: the future is still there (1:14; 4:30), and Romans speaks of believers as already alive to God.
4. Paul's past. In his popular book Forged, Ehrman finds it odd that "Paul" includes himself among those who once lived "in the lusts of our flesh" (2:3): "This doesn't sound like the Paul of the undisputed letters, who says that he had been 'blameless' with respect to the 'righteousness of the law'".⁴⁰² Manning replies with Romans 7, where Paul describes covetousness at work in him under the law.³⁸ That reply depends on reading the "I" of Romans 7 as Paul's own past, which interpreters dispute. A firmer reply uses passages where Paul plainly includes himself: "Are we better than they? No, in no way. For we previously warned both Jews and Greeks that they are all under sin" (Rom 3:9), and "We, being Jews by nature and not Gentile sinners, yet knowing that a man is not justified by the works of the law ... even we believed" (Gal 2:15-16).² "Blameless" in Philippians 3:6 is about keeping the law's requirements, not about freedom from desire. So Ephesians 2:3 does not contradict Paul; its "we all" is a general confession joining writer and readers, as Paul's accepted letters also make.
5. The church and the apostles. The church here is the universal church, "built on the foundation of the apostles and prophets" and the gospel "revealed to his holy apostles and prophets".² To many readers "holy apostles" sounds like a later generation looking back. Paul calls whole churches "saints," holy ones, so the phrase is not impossible for him; but it is unusual.
6. The household rules (5:21-6:9). Best judges that, unlike the teaching on the church, the household code "does not display Paul's awareness of the real situation of believers."¹ That is a judgment of tone, and other readers find the code's addresses to slaves and children close to Colossians 3-4.

Verdict on Ephesians

The external evidence is among the best of the six:  Polycarp quotes it and probably counts it as scripture, Marcion had it, and Irenaeus names it.  Its vocabulary is Paul's.  Against that, its sentence style is unlike Paul's, some phrases are easier to read as addressed to people he had not taught, and it is closely bound to Colossians in a relation whose direction is unsettled.  Best states the critical case fairly:  "The vocabulary and style vary slightly from Paul's in such a way as to suggest, but not compel, rejection of Pauline authorship," and the objections work by "their cumulative effect".¹¹  Dunn sides with the majority against Paul.³³  Paul's in the historical sense:  unresolved.  Paul's own wording:  doubtful, because of its style.  If Colossians is a letter of Paul's lifetime and Ephesians was sent with it by Tychicus, the shared carrier note and its "you also" fit well.  If Ephesians drew on a later Colossians, it is later still.  The evidence specific to Ephesians does not settle which.

Part 7: The Pastoral letters

Do the three share an author?

On 1 Timothy and Titus the two sides agree.  Ehrman finds "numerous clear and specific parallels and overlaps that are virtually inexplicable apart from a literary relationship of some kind, either an abject borrowing of one author by another or, far more likely in this case, joint authorship."⁴  McLatchie accepts this and adds the greetings, whose address no other New Testament letter shares:  "to Timothy, my true child in faith" and "to Titus, my true child according to a common faith".²  Both letters give rules for choosing church leaders, warn against teachers of Jewish myths, and use the formula "the saying is faithful".  One author for 1 Timothy and Titus is very probable.  Oddly, the two share no five-word sequence outside their greetings,²⁷ which shows how little shared wording by itself says about authorship in either direction.

2 Timothy is different in kind:  a farewell from prison, almost silent on church order, full of names and errands.  It shares the greeting and some phrases, and most of its vocabulary sits with the other two (Part 3).  Most scholars on both sides take all three as one author.  But McLatchie notes that some have held 2 Timothy alone to be Paul's,⁶ and Johnson insists that each letter "must therefore be considered individually".⁵  So the evidence on 1 Timothy bears directly on Titus, and only partly on 2 Timothy.

Where would they fit in Paul's life?

The letters give these settings:²

1. 1 Timothy: Paul has gone to Macedonia, leaving Timothy at Ephesus (1:3). Acts 20:1 has Paul leave Ephesus for Macedonia, but in Acts Timothy had already been sent ahead to Macedonia (19:22) and travels on with Paul (20:4), so 1 Timothy does not fit that departure without strain. Some defenders place it in a gap in Acts' account of the Ephesian years.
2. Titus: Paul has worked in Crete and left Titus there (1:5), and plans to winter at Nicopolis (3:12). Acts has Paul touch Crete only as a prisoner on the way to Rome (27:7-13). No point in Acts fits.
3. 2 Timothy: Paul is in chains in Rome, has had a "first defense" at which all deserted him, and expects death (4:6, 16-17). That could in principle be the Roman custody with which Acts ends, which Acts does not see through to its outcome. Paul expected death in other letters from prison too (Phil 1:23). What strains that setting are the travel notes: Trophimus left sick at Miletus and the cloak "left at Troas" (4:13, 20) imply a recent journey through Asia, but in Acts Trophimus went on to Jerusalem with Paul (21:29) and the voyage to Rome did not pass Miletus or Troas. Onesiphorus also had to search hard to find Paul in Rome (1:17), while Acts ends with Paul receiving all comers in his own lodging (28:30-31).²

So the difficulty is real for Titus and 1 Timothy and partial for 2 Timothy.  Philip Towner, who leans to Paul's authorship, sets out both ways of meeting it:  a release from the Roman custody of Acts 28 followed by more travel and a second imprisonment, or a fit inside gaps in Acts.  Of the first he writes that "the basic hypothesis ... remains a possibility. The greatest challenge is probably the uncertainty of the evidence on which the release from his first Roman imprisonment is grounded."⁴¹  The evidence for a release is thin but early.  Paul expected release in two of his prison letters ("I trust in the Lord that I myself also will come shortly";  "prepare a guest room for me"), and Romans planned a journey to Spain.²  1 Clement, written from Rome about the 90s, says Paul "reached the limits of the West" before his death;⁷ the Muratorian list says Acts omits "the departure of Paul from the city [of Rome] when he journeyed to Spain".¹³  Both may rest on Romans 15 rather than on separate knowledge.  Ehrman grants that the chronology argument is weak:  "our hard information on Paul's ministry is extremely sparse: eight of the twelve years spanning 50 and 62 CE are summed up in the book of Acts in four lines."⁴

McLatchie goes further:  a forger "would be far more likely" to set his letters "at some recognizable point within the travels in Acts", so the misfit with Acts "is an argument in precisely the opposite direction."⁶  That argument holds against one kind of forger, one who built his settings from Acts:  such a writer would probably have placed them where Acts could confirm them.  It does not hold against a writer who did not know Acts, as Ehrman thinks this one did not,⁴ or one who drew on a tradition that Paul travelled again after Rome.  Nor did an imitator who wanted a dying Paul's last words have to set them after Acts:  Acts ends with Paul in Roman custody and his fate untold, so a farewell could be set at that point.  The misfit with Acts is therefore evidence against an Acts-based forgery, which few critics propose, and neutral between Paul and a writer working from other traditions.  A later release is possible, not shown.

Timothy's youth is a smaller point.  Ehrman asks how Timothy, already a disciple about AD 49, could still be warned "Let no man despise your youth" (1 Tim 4:12) some fifteen years later.⁴  McLatchie answers that a man of about thirty-two could still be called young.⁶  The answer works.  The same phrase is used of Timothy in 1 Corinthians:  "let no one despise him" (16:11).²  Youth relative to Paul and to church elders is all the phrase needs.

Church order

Ehrman's strongest historical argument concerns church leaders.  In Corinth, he says, there were no officers:  "Each member of the community had a spiritual gift".  The Pastorals, by contrast, "presuppose that a church hierarchy already is in place, to the extent that there are bishops, deacons, and widows who are to be enrolled."⁴  And on the spiritual gifts of every member:  "Nothing."⁴  Manning and McLatchie answer that Paul's accepted letters already have leaders:  Philippians greets "the overseers and servants" (in Greek ἐπισκόποις καὶ διακόνοις, the same words as 1 Timothy 3), 1 Thessalonians asks respect for "those who ... are over you in the Lord," and 1 Corinthians lists "governments" among God's gifts.²  Ehrman's reply is that the Philippian overseers are never addressed or described:  "nothing is said about them at all, giving us no way of knowing whether they are comparable to the figures addressed in the Pastorals or not."⁴

That leaves the real question:  what is new in the Pastorals is not the titles but written qualifications for leaders, rules for which widows the church supports, and a delegate who appoints elders.  How much of that is expected in a letter to a delegate, whatever its date?

1. The genre explains much of it. 1 Timothy and Titus are written to Paul's representatives, telling them what to do in churches where they act for him. Six of Paul's accepted letters address congregations; the seventh, Philemon, is a personal appeal, not instructions to a delegate. Towner, following Michael Wolter and Johnson, compares 1 Timothy and Titus to "mandates" in which a ruler instructed a delegate and the community receiving him, and he is careful about what the comparison shows: it "does not settle the authorship question (anyone, Paul or a pseudepigrapher, could have formulated the letters), neither does the unique shape of the letters in comparison with the undisputed letters to churches preclude the possibility that Paul could have designed such letters to his subordinate delegates."⁴¹ Paul did send delegates elsewhere: Timothy to Thessalonica, Corinth and Philippi.² No letter of instruction to such a delegate survives among the accepted letters, so there is no direct control.
2. The handing over to Satan is not a new procedure. 1 Timothy 1:20 says "whom I delivered to Satan". In 1 Corinthians 5 Paul also writes "I ... have already ... judged him", and then tells the gathered church to carry out the same handing over.² A one-line look back at a past discipline need not mention the assembly; as Towner says, the past tense "does not rule out the church's participation in the event."⁴¹ This is no evidence of a new, more authoritarian structure.
3. Elders in every town match Acts. Titus is to "appoint elders in every city" of Crete (1:5). The accepted letters never use the word "elders" for church leaders, which counts for something. But Acts has Paul and Barnabas appoint "elders for them in every assembly" on their first journey (14:23).² Whoever treats Acts as historical evidence for Paul elsewhere cannot treat this practice as unthinkable in his lifetime; a critic who dates Acts late can say that Acts reflects the same later practice. So elders in every town fit Acts' picture of Paul and differ from the vocabulary of his own letters.
4. The widows are a support rule, not an order. 1 Timothy 5 tells the church to support widows who have no family, and their families to support the rest, with sixty as the age for being put on the list (5:9). Care for widows was a church duty from the start (Acts 6:1). William Mounce, defending Paul's authorship, argues that the chapter does "not show an order of widows with entrance requirements and prescribed tasks"; it settles "which widows should be cared for by the church and which widows should be cared for by their families."⁴² The rule implies a church old enough to have many widows and a fund to support them, but nothing that requires the second century.
5. The structure is still simple. Mounce points out that the Pastorals have overseers and deacons, not the three ranks under a single bishop found in Ignatius of Antioch early in the second century, and that Timothy and Titus are not local officers but "itinerant, apostolic delegates sent with Paul's authority to deal with local problems".⁴² Towner adds a point the other way: Titus asks a delegate to choose leaders in newly founded house churches, which "would have been completely irrelevant" years after Paul, when leaders were already in place.⁴¹ A critic replies that a later writer used exactly such instructions to give an existing order apostolic backing. Both readings fit.

So the church order of 1 Timothy and Titus is more settled than anything in Paul's letters to congregations, but most of the difference has a ready explanation in the letters' purpose and readers.  Taken alone it is weak evidence of a later date.  Johnson's caution applies:  2 Timothy has almost none of it.

Teaching

1. The law. 1 Timothy says "we know that the law is good if a person uses it lawfully," and "law is not made for a righteous person, but for the lawless".² Ehrman argues that for Paul "the Law is good unconditionally, regardless of how a person uses it, since it was given by God."⁴ But the condition in 1 Timothy attaches to how a teacher uses the law, not to whether the law is good: the next lines say what it is for, which is to restrain the lawless. The first words, "the law is good", echo Romans 7:16, where Paul agrees "that it is good" (1 Timothy καλὸς ὁ νόμος; Romans σύμφημι τῷ νόμῳ ὅτι καλός), and Paul too says the law "was added because of transgressions" (Gal 3:19) and that against the fruit of the Spirit "there is no law" (Gal 5:23).² The emphasis and the vocabulary around it are not Paul's usual ones; a contradiction of Paul is not shown.
2. "Faith" as a body of teaching. Ehrman says that in the Pastorals "faith" is no longer trust in Christ: "It is now the content of the religion".⁴ Manning replies with Galatians 1:23, where people say Paul "now preaches the faith that he once tried to destroy".³⁸² Philippians 1:27 speaks of "striving for the faith of the Good News". And 2 Timothy still uses faith as personal trust: "I know him whom I have believed".² The difference is one of frequency, not a new idea.
3. Opponents. The false teachers forbid marriage and certain foods, teach "Jewish fables" and want to be "teachers of the law," say "the resurrection is already past", and 1 Timothy closes against "what is falsely called knowledge" (gnosis).² F. C. Baur took them to be second-century Gnostics,⁴ and Irenaeus later took that phrase as the title of his attack on the Gnostics of his day. But the opponents described are Jewish-leaning teachers inside the churches, not the elaborate systems Irenaeus fought, and Paul already faced Corinthians who prided themselves on "knowledge" (1 Cor 8:1) and denied the resurrection of the dead (1 Cor 15:12).² Baur and others once read the warning against "oppositions" (ἀντιθέσεις) in 1 Timothy 6:20 as an attack on Marcion's book of that name, a view Mounce reports is "rightly rejected today";⁴² if Polycarp used 1 Timothy by about 135, as his echoes make likely, the letter is older than Marcion's work in Rome. The opponents do not require a second-century date.

1 Timothy 5:18 and Luke

1 Timothy 5:18:  "For the Scripture says, 'You shall not muzzle the ox when it treads out the grain.' And, 'The laborer is worthy of his wages.'"  The second saying is Luke 10:7 word for word in Greek, less Luke's "for";  Matthew has "food" instead of "wages".²  Ehrman takes this as a sign of a date "at the end of the first century at the very earliest, when Gospel texts were assuming a position of authority".⁴  McLatchie replies that it is evidence for an early Luke instead.⁶  Three questions decide what it shows, and none has a sure answer.

1. Does "Scripture says" cover both sayings? Mounce argues it "need only apply to Paul's first citation from the OT", as one formula in Mark 1:2-3 introduces several quotations but names only Isaiah.⁴² Towner agrees that the formula probably does not cover the saying of Jesus.⁴¹ The "and" can be read either way.
2. Was the saying taken from written Luke? Paul already pairs this very verse of Deuteronomy with a command of Jesus about workers' pay in 1 Corinthians 9:9-14, a letter nobody doubts: "the Lord ordained that those who proclaim the Good News should live from the Good News."⁴¹² And Paul's account of the Last Supper in 1 Corinthians 11:24-25 already agrees with Luke's wording against Mark's ("Do this in memory of me"; "This cup is the new covenant in my blood"), a point Mounce credits to Gordon Fee.⁴²² So Paul knew sayings of Jesus in a form close to Luke's before Luke was written, and a Lukan wording in 1 Timothy does not need a written Gospel.
3. When was Luke written? If 1 Timothy quoted written Luke as scripture, its date would depend on Luke's, which is argued elsewhere. An early Luke would make that compatible with Paul; a late Luke would not. Neither date is settled by this verse.

The verse fits a writer who knew a Lukan form of Jesus' saying.  It is evidence for a late date only on the readings that make "Scripture" cover a written Gospel, and then only as late as Luke itself.

Undesigned coincidences

McLatchie and Manning both build much of their case for the Pastorals on "undesigned coincidences":  small details in two independent texts that fit together without either seeming to intend it.  McLatchie defines the test well:  details that "interlock in a way that is unexpected if (a) the story is being made up out of whole-cloth; (b) one account is borrowing from the other; or (c) both documents are copying from a common source."⁶

The test is a matter of degree.  A coincidence counts for Paul to the extent that it is more expected if Paul wrote than if a later writer did.  Where a later writer could have drawn the detail from Acts, Romans or tradition, the agreement is less surprising on the later-writer view, and so it counts for less.  But "could have drawn" is not "did draw".  Ehrman himself thinks the writer "probably did not know" Acts.⁴  On Ehrman's own view, then, agreements with Acts are agreements between two independent sources, and they show that the writer of the Pastorals (often called "the Pastor") had accurate information about Paul's early travels and Timothy's family.  That favours a writer close to Paul's circle over a distant one;  it does not by itself favour Paul over a well-informed disciple.  Critics who think the Pastor did know Acts can explain the same agreements by borrowing.

1. Timothy's family. 2 Timothy praises the faith of Timothy's "grandmother Lois and your mother Eunice," and says he knew the scriptures "from infancy". Acts says he was "the son of a Jewess who believed, but his father was a Greek," and that Paul circumcised him because everyone knew his father was a Greek.² McLatchie's point is that the two accounts fit at an angle: Acts implies an unbelieving Greek father and a believing Jewish mother, and 2 Timothy praises the faith of the women of the family and says nothing of the father, while adding names and a grandmother that Acts lacks.⁶ The fit of an uncircumcised son raised on the scriptures by a Jewish mother is real. A writer who knew Acts 16 could produce it by naming the mother and adding a grandmother; a writer who did not would need accurate family tradition. Either way the detail counts for good information. It counts for Paul only as much as the names Lois and Eunice are unlikely to be invented, which cannot be measured.
2. Antioch, Iconium, Lystra. 2 Timothy recalls "persecutions, and sufferings—those things that happened to me at Antioch, Iconium, and Lystra."² Acts 13-14 records trouble in exactly those three towns, in that order, and none in Derbe, the fourth town of that journey.² McLatchie and Manning add why these three: Timothy came from Lystra, and was most likely converted on that very journey, so these were the persecutions he had seen or heard of at home, and Paul names them to him rather than the many later ones.⁶⁴³ That is a genuine point about selection, and it is the strongest of these examples. A writer drawing on Acts would have had to select the persecution towns and leave out Derbe; possible, but an unforced choice. A writer with independent tradition would have to have it this exact. Paul writing to a man from Lystra explains the choice most simply.
3. Erastus. "Erastus remained at Corinth" (2 Tim 4:20). Acts 19:22 pairs Erastus with Timothy; Romans 16:23, written from Corinth, sends greetings from "Erastus, the treasurer of the city".² A pavement found at Corinth names an Erastus who paid for it as aedile, a city magistrate, about the middle of the first century. Andrew Clarke, who studied the inscriptions, concludes only that identifying him with Paul's Erastus "should remain therefore a significant possibility."⁴⁴ The fit of an Erastus who stays at Corinth with an Erastus of Corinth in Romans is a contextual fit, not an independent confirmation, and the pavement may or may not be the same man.
4. Alexander the coppersmith. McLatchie, following Lydia McGrew, argues that a forger placing Timothy at Ephesus would have chosen Demetrius the silversmith of Acts 19 as Paul's enemy, and notes that the man 2 Timothy names instead, a coppersmith, belongs to the same kind of trade as the "workmen of like occupation" whom Demetrius gathered (Acts 19:25).²⁶ The point has some force against a writer building from Acts, who would more likely have reached for the famous name. It has little against a writer who did not use Acts or who chose an obscure name to avoid looking contrived, and forgers did invent names. Acts 19:33 also mentions an Alexander at the Ephesian riot, which a borrower could have used. Weak evidence, but not none.

The critic Kipp Davis put the objection sharply, as Manning quotes him:  Manning "demonstrates that he is just as clever as second-century Christian writers were in crafting their highly convincing forgeries."⁴⁵  Davis assumes what he needs to show.  The cumulative case is real but limited.  The examples are not independent of one another, since most are checked against the same book, Acts, and their combined force depends on one unresolved question, whether the Pastor knew Acts.  What they establish with some confidence is that the writer had accurate, detailed information about Paul's circle.  That makes an ignorant or distant forger unlikely.  It gives modest support to Paul, or a helper writing for him, over a disciple a generation later, because accurate incidental detail is more expected from someone who lived it;  but it does not rule the disciple out.

Personal notes and the fragment theory

2 Timothy is full of notes that no argument needs:  "Bring the cloak that I left at Troas with Carpus when you come—and the books, especially the parchments."²  1 Timothy advises "a little wine for your stomach's sake".  Defenders find such details hard to explain as invention.  Ehrman, citing Brox, answers that fake personal notes were a known device.  Harrison proposed a middle path:  real notes of Paul built into later letters.  Ehrman reports that the theory "was finally demolished by Cook, who showed that the alleged differences between source and redaction required by the theory in fact do not exist."⁴  The personal notes are the best evidence for some real link to Paul in 2 Timothy.  How strong they are depends on how far one thinks ancient forgers went, and the sister article's survey shows that ancient readers who caught such forgers condemned them.

The secretary proposal: Luke

C. F. D. Moule proposed in 1965 that "Luke wrote all three Pastoral Epistles. But he wrote them during Paul's lifetime, at Paul's behest, and in part (but only in part), at Paul's dictation."  Ben Witherington develops the idea:  "the voice is the voice of Paul, but the hand is the hand of Luke."  He cites Stephen Wilson's count of "some 37 (or more accurately, 34) words" found in the New Testament only in the Pastorals and Luke-Acts.⁴⁶⁴⁶  Mounce argues the same from the setting of 2 Timothy, where Paul is alone in harsh custody except for Luke, and suggests that "Luke was Paul's amanuensis for all three letters and was given considerable freedom in writing to two men who, most likely, were also Luke's friends."⁴²  McLatchie adopts it as a plausible explanation of the language.⁶  2 Timothy itself says "Only Luke is with me" (4:11).²

This is the strongest form of the case for Paul, because it accepts the language evidence instead of denying it.  Its difficulties are two.  First, it has no direct evidence:  no text says Luke wrote them.  Johnson, who defends the Pastorals, explains the shared words with Luke differently, as both writers drawing on "a broader sample of Koine Greek".⁵  Towner, who also leans to Paul, leaves the question of how the letters were written open:  "Here I choose to allow open questions to remain open."⁴¹  Second, Ehrman's objection applies:  of the co-authored papyrus letters Michael Prior collected, "we literally have not a single clue" how they were composed,⁴ and in the evidence surveyed here no case is documented of a helper composing long instructional letters for another man.  That absence cuts both ways:  it means the Luke proposal has no ancient parallel to lean on, and also that nothing shows such letters were never made.  The Luke proposal is possible.  It is a hypothesis built to fit the evidence, not something the evidence points to by itself.

Verdicts on the three

1. 1 Timothy. Its vocabulary lies further from Paul's than any stretch of his own letters. Its church order, its teaching on the law and faith, its opponents and its saying of Jesus each have a reasonable explanation inside Paul's lifetime, so together they add only a little to the vocabulary. Polycarp very probably knew it by about 135, Irenaeus names Paul, and the Luke proposal can absorb the language. Paul's own wording: probably not, with moderate confidence. Paul's in the historical sense, sent through a helper in his lifetime, or written after his death: unresolved.
2. 2 Timothy. Most of its vocabulary is as far from Paul's as 1 Timothy's, though less so as a whole. It has almost nothing on church order. Its personal notes are the most numerous and least necessary of the six letters, and its list of persecutions is the strongest undesigned coincidence in the Pastorals. Early writers knew it, and Irenaeus names it. Against it, it shares phrasing with 1 Timothy and Titus, Marcion's collection and perhaps P46 lacked it, and its travel notes fit no point in Acts. Possible accounts range from Paul writing near his death through a helper, to a later writer using genuine notes, to a later writer with good information. Paul's own wording: doubtful. Paul's in the historical sense: open. The evidence does not choose among the accounts.
3. Titus. It very probably shares an author with 1 Timothy, and its vocabulary is as far from Paul's. It is so short (659 words) that every measure of it is uncertain, and van Nes's measure puts it just inside Paul's range. Its setting fits no point in Acts. Tatian, who rejected other Pauline letters, accepted it. Paul's own wording: probably not, with lower confidence than for 1 Timothy. Paul's in the historical sense: unresolved.

Part 8: McLatchie and Manning, checked

Jonathan McLatchie, a Christian apologist who writes on biology and the New Testament, wrote the fullest defence of the Pastorals examined here (2021) and an earlier note on Colossians and Philemon (2017).  Erik Manning, who runs the Testify channel and the isjesusalive.com site, has answered Ehrman letter by letter in articles from 2019 to 2024 and in videos.  Both rely heavily on William Paley's Horae Paulinae (1790) and on Lydia McGrew's revival of the undesigned-coincidence argument.  Manning's videos mostly repeat his written articles, so the articles are cited here;  where a video was checked, it was checked in YouTube's automatic captions, which give the gist but not reliable wording.

Their strongest arguments, as they present them, are the early and unanimous naming of the letters, the Philemon-Colossae link, the coincidences between 2 Timothy and Acts, the personal details of 2 Timothy, Foster's argument from the signature in 2 Thessalonians, and the reconciling of the "thief" and the signs.  Their main claims, with what checking found:

Manning's satirical video applying the critics' arguments to Philippians makes a fair methodological point:  some critical arguments, applied with equal force, would cast doubt on letters no one doubts.  In it the narrator, speaking as a critic, points to the dozens of words in Philippians found nowhere else in the New Testament.⁴⁷  The counts in Part 3 answer part of that worry:  applied the same way to every letter, they leave Philippians inside Paul's range and put 1 Timothy and Titus outside it.  They do not answer all of it, because a measure that separates letters still cannot say why they differ.

Objections and responses

1. "Most critical scholars call all six forgeries. Who are you to disagree?" The scholarly picture is not one vote on six letters. Dunn, no conservative, reports "a roughly even split among critical commentators on Colossians and 2 Thessalonians," and himself holds 2 Thessalonians to be Paul's.³³ Foster, another critical scholar, argues the same for 2 Thessalonians. On the wording of 1 Timothy and Titus this article agrees with the critical majority; on whether those letters are Paul's in the historical sense it leaves the question open. Towner leans toward Paul while leaving the manner of composition open. In each case the reason is an argument the reader can check, not a count of names.
2. "Unanimous early naming proves nothing; forgeries were accepted quickly." Agreed that it does not prove authorship. It does set a latest date for each letter: before about 135 for Ephesians and 1 Timothy, and probably for 2 Thessalonians and 2 Timothy; by the 140s for Colossians; by about 170 for Titus, or earlier if it shares 1 Timothy's author. And it is not empty, because ancient readers did test claims of authorship and rejected some letters in Paul's name.
3. "A secretary or co-writer can explain anything. It is an escape hatch." Used as a general rule, it is, and this article does not use it as one. A helper predicts a particular pattern: different style, with Paul's teaching and Paul's situation. Colossians shows a pattern close to that (Paul's vocabulary, a different sentence style, developed but not contradictory teaching), which is why the helper hypothesis earns some weight there. For 1 Timothy and Titus a helper would also have to account for a new vocabulary. The Luke proposal does that, but with no direct evidence, so it is left as possible, not probable.
4. "A pseudonymous letter would be a lie, and the church would never have accepted one. So all thirteen must be Paul's." The first part is supported by the ancient evidence: the sister article shows that ancient readers who detected false apostolic writings condemned them. But that is evidence that such letters were wrong, not evidence that none was ever accepted. The historical question has to be answered first, letter by letter. On the evidence here, 2 Thessalonians and Colossians lean toward letters of Paul's lifetime, and 1 Timothy, 2 Timothy and Titus could be letters Paul sent through a helper. But this article cannot honestly call that last account probable; it can only say it is not ruled out.
5. "Vocabulary depends on subject. Paul never wrote to a delegate about church leaders before." That is the strongest reply, and the article accepts its force. The counts control length, not subject, audience or helper, and Paul's surviving letters include no letter to a delegate for comparison. The reply cannot be tested with the letters we have. It leaves the 1 Timothy and Titus figures standing as a large difference, but it weakens what the difference proves. Van Nes's adjusted rare-word analysis likewise shows how the choice of measure changes the result (Part 3).
6. "Many undesigned coincidences add up, even if each is weak." They add up only as far as they are independent and each is more expected if Paul wrote. The Pastorals' examples are mostly checked against one book, Acts, and their weight turns on whether the writer knew it. They add up to strong evidence that the writer knew Paul's history well, and modest evidence for Paul over a later disciple. They do not add up to proof.
7. "Bujard's style evidence settles Colossians. The Timothy idea is special pleading." The style evidence is strong, and several measures agree. But Colossians names Timothy as co-sender, so a helper is not invented to save the letter, and a helper predicts what the evidence shows, a change of style without a change of vocabulary. That is why this article judges Paul's own wording improbable but the letter's lifetime origin still more likely than not. Being named as co-sender does not show that Timothy chose the words.
8. "The overlap between 1 and 2 Thessalonians has no parallel in Paul." The measured verbal overlap is higher than between Romans and Galatians, a little below that between Colossians and Ephesians, and far above 1 and 2 Corinthians; with the shared layout it is unusual. This article counts it as evidence for imitation, outweighed, on balance, by the reception and setting. It is unusual, not unparalleled.
9. "P46 shows the Pastorals did not exist about 200." On Kenyon's reconstruction of the book's missing end, one Egyptian collection was planned without them; Nongbri has shown that the reconstruction itself may undercount the lost leaves. Either way, Polycarp already knew 1 and probably 2 Timothy, and P46 may also have lacked Philemon.
10. "Marcion threw out the Pastorals because they contradicted his theology, so his collection says nothing about their date." That may be so; it is Tertullian's view, and Clement reports heretics rejecting them for their content. It is also possible that his collection never had them. The article counts Marcion's collection only as weak evidence either way for the Pastorals.
11. "If 1 Timothy's wording is not Paul's, it is a forgery." Not necessarily. A letter Paul authorised and a helper drafted is Paul's in the sense that matters for history, and this article keeps that possibility open for all three Pastorals. Equally, it cannot be assumed: a letter whose wording is not Paul's may also be one written after his death. The evidence here does not decide between those for 1 Timothy and Titus.

Issues and unanswered questions

1. Several standard commentaries were not available: Andrew Lincoln on Ephesians, Dunn on Colossians, I. Howard Marshall on the Pastorals, Abraham Malherbe on Thessalonians, and the Oxford study of how the Apostolic Fathers used the New Testament (Gregory and Tuckett, 2005). Their views appear here only where another source reports them. Walter Bujard's and Daryl Schmidt's style studies were not read; their figures are given as Ehrman reports them. Jeremy Duff's 1998 article on P46 and Edgar Ebojo's 2014 study of the manuscript are known here only through Brent Nongbri's 2022 account of them.
2. Ehrman's Forgery and Counterforgery was read in a copy whose page numbers differ from the printed book, so it is cited by chapter and section. Paul Foster's 2012 article was read in the author's accepted manuscript, whose pages differ from the journal's.
3. How many leaves did P46 have? If its page numbering began after some blank leaves, as in other books of its kind, it had more room at the end than Kenyon allowed, and its evidence about the Pastorals is weaker still.
4. The counts are descriptive. They have no error bars, short letters give few stretches, and subject, audience and helper are not controlled. The measures do not all agree, especially for 2 Thessalonians, and van Nes's adjusted rare-word test puts all six inside Paul's predicted range. A proper test would model each letter's uncertainty, use letters by other ancient writers of similar length and kind as controls, and fix the method before looking at the disputed letters.
5. Who knew Acts? The weight of the Pastorals' agreements with Acts turns on whether their writer had read Acts. Neither side has shown this.
6. Was Paul released from his first Roman imprisonment? 1 Clement 5 and the Muratorian list say he reached the West or Spain, but both may rest on Romans 15.
7. The original address of Ephesians is uncertain, and with it the letter's first readers.
8. Which way did the borrowing between Colossians and Ephesians run? The verdict on Ephesians depends on it, and the evidence has not settled it.
9. Where did Philemon live? Only Colossians places his household at Colossae. An independent witness would turn the Philemon network from evidence of coherence into evidence of accurate knowledge.
10. Erastus. Whether the aedile of the Corinth pavement is Paul's Erastus remains, in Clarke's words, a "significant possibility".
11. The Testify videos were read in YouTube's automatic captions with time stamps, not heard. Where a video repeats a written article, the article is cited instead.
12. What follows for scripture if a letter is pseudonymous, or written by a helper, is a theological question this historical article does not try to settle.

Verdict

The six disputed letters do not stand or fall together, and "Did Paul write it?" has two parts:  whether the letter is Paul's, sent by him in his lifetime, and whether the words are his own.

1. 2 Thessalonians is probably Paul's, sent with Silvanus and Timothy; the confidence is low to moderate, and a later imitation of 1 Thessalonians remains a serious alternative.
2. Colossians is more likely than not a letter of Paul's lifetime, but its words are probably not his own; a helper, perhaps Timothy, may have shaped them. The first judgment carries low confidence, the second low to moderate.
3. Ephesians is unresolved. Its style tells against Paul's own wording, and its date depends on a relation to Colossians whose direction is unsettled.
4. 1 Timothy and Titus are probably not in Paul's own words. Whether he sent them through a helper or a follower wrote them after his death, the evidence does not decide.
5. 2 Timothy is open.

The strongest evidence for Paul is the early, unanimous naming of all six once naming begins, Polycarp's use of four of them, the coherence of Colossians with Philemon, the fit of 2 Thessalonians, Colossians and Ephesians with Paul's vocabulary, and the well-informed details of 2 Timothy.  The strongest evidence against is the distance of 1 Timothy, Titus and most of 2 Timothy from Paul's vocabulary, larger on this article's measure than any stretch of his accepted letters shows, the sentence style of Colossians and Ephesians, and the close imitation of 1 Thessalonians in 2 Thessalonians.  Van Nes's adjusted rare-word analysis also puts all six letters inside Paul's predicted range, so the vocabulary result depends on the measure chosen.  Several arguments often used against the Pastorals, from church order, a "Scripture" quotation and Acts, prove much less than they seem.  McLatchie and Manning are right that several popular arguments against the letters are weak, and that the six should not be written off together.  The evidence does not show, as they hold, that Paul wrote all of them;  for four of the six it leaves the question open.

Sources