Eleven Words Deep

• Anchor Bible Dictionary — “Corinth,” “Corinthians, Second Epistle to the,” “Collection, The,” “Aretas,” “Nabateans,” “Heaven,” “Paradise.”
• Dictionary of Paul and His Letters (IVP) — “Corinthians, Letters to the,” “Collection for the Saints,” “Opponents of Paul,” “Boasting,” “Weakness,” “Visions, Ecstatic Experience.”
• NIDNTT / TDNT — χάρις, κοινωνία, καυχάομαι, ἀσθένεια, σκόλοψ, μετασχηματίζω.
• Commentary on the New Testament Use of the Old Testament (Beale & Carson) — on 2 Corinthians, especially 8:15 (Exodus 16:18), 9:9 (Psalm 112:9 / LXX 111:9), 10:17 (Jeremiah 9:23–24 / LXX 9:22–23) and 13:1 (Deuteronomy 19:15).

2 Corinthians is a New Testament book, so there is no Masoretic column and no Septuagint column to line up. There is still an offset, and it is worth knowing before you chase a cross-reference.

ESV 13:12–13 = Greek 13:12. ESV 13:14 = Greek 13:13. The English tradition divides greet one another with a holy kiss from all the saints greet you; the Greek editions keep them as one verse. Everything before chapter 13 carries the same number in both. Derived here by comparing the MorphGNT verse index against an English chapter-shape file, both held on disk.

Practical consequence: a commentary or a hymn-book heading that reads “2 Corinthians 13:13” may mean all the saints greet you or it may mean the grace benediction, depending on which numbering it follows. Continental European editions and critical Greek texts usually number the benediction 13:13; English Bibles number it 13:14. This page uses the English number and prints the Greek beside it.

Both Old Testament texts quoted in these chapters are in the part of the canon where the Greek and the Hebrew numbering part company.

The Greek Psalter runs one behind the Hebrew from Psalm 9/10 through 147, which is why the psalm about the generous man is 111 in one column and 112 in the other. Greek Jeremiah is differently arranged throughout and one verse out at this point. Both are given in both numbers everywhere they appear on this page.

Chapters 8 and 9 are about a single, specific relief fund: money gathered in Paul’s Gentile churches for believers in Jerusalem. It runs across four letters — Galatians 2:10, 1 Corinthians 16:1–4, these two chapters, and Romans 15:25–31 — and Acts 24:17 has Paul telling a Roman governor he came to Jerusalem to bring alms to my nation.

Paul’s vocabulary for it is measurable and it is not the vocabulary of a tax. Across chapters 8 and 9 he calls the fund χάρις (grace), κοινωνία (fellowship), διακονία (service), λειτουργία (public service, the word for a civic benefaction or a priestly duty), εὐλογία (blessing) and ἁπλότης (single-mindedness, generosity). He never once calls it φόρος, tribute.

And ten of the eighteen occurrences of χάρις in the whole letter are in these two chapters. Counted over MorphGNT: 8:1, 8:4, 8:6, 8:7, 8:9, 8:16, 8:19, 9:8, 9:14, 9:15. The word English readers meet as a doctrine of salvation is, in these chapters, the word for a collection — and at 8:9 the same word carries the incarnation.

What was not consulted. No papyri, no Greco-Roman benefaction inscriptions, no rabbinic material on the temple tax or on almsgiving. The civic-benefaction background of λειτουργία is standard in the lexica and is named here as background, not as a checked source.

The specific charges Paul answers can be read off the text. His letters are weighty and strong but his bodily presence is weak and his speech of no account (10:10). He is ἰδιώτης τῷ λόγῳ, an amateur in speech (11:6). He refused support from Corinth and took it from Macedonia instead (11:7–9), which he phrases with brutal irony as I robbed other churches. His rivals commend themselves and measure themselves by themselves (10:12).

That last complaint is worth measuring. συνίστημι, to commend, to present, stands 15 times in the Greek New Testament. Nine of them are in 2 Corinthians — 3:1, 4:2, 5:12, 6:4, 7:11, 10:12, 10:18 (twice) and 12:11. It is the letter’s running argument with itself about whether a servant of God may present his own credentials, and it ends at 10:18: it is not the one who commends himself who is approved, but the one whom the Lord commends.

12:2–4 is the only place in the New Testament that numbers a heaven as one of a series. ⚠️ It is not the only ordinal standing beside οὐρανός: Revelation 21:1 has ὁ πρῶτος οὐρανός, but that is the former heaven set against a new one, not one floor of a stack. Paul says he knows a man in Christ who was caught up ἕως τρίτου οὐρανοῦ, to a third heaven, and (12:4) into τὸν παράδεισον, paradise, where he heard ἄρρητα ῥήματα, unspeakable words. Whether the third heaven and paradise are two stages or one place said twice is a question the Greek does not settle, and this page does not settle it either.

The measurable part. παράδεισος stands three times in the Greek New Testament: Luke 23:43 (to the thief), 2 Corinthians 12:4, Revelation 2:7 (the tree of life). ὀπτασία, vision, stands four times and this is Paul’s only use of it. ἄρρητος is a New Testament hapax.

⚠️ What this page did not read. The layered heavens of 2 Enoch, the Testament of Levi and the Ascension of Isaiah are the standard comparison for “the third heaven,” and none of them was consulted. No Second Temple text is held on this shelf; the corpora here are the Masoretic Hebrew, the CATSS Greek Old Testament, and the Greek New Testament. The comparison is named as the place to look, not reported as a finding.

These 124 verses contain 40 New Testament hapax legomena — words that occur exactly once in the whole Greek New Testament. Counted over the MorphGNT lemma column across all 27 files.

Chapter 11 — the hardship list, the basket over the wall, the ethnarch of Aretas — carries fifteen of them: ἁρμόζομαι, συλάω, ἀβαρής, ψευδαπόστολος, δόλιος, παραφρονέω, ὑπερβαλλόντως, πεντάκις, νυχθήμερον, βυθός, δίψος, ἐθνάρχης, Ἁρέτας, Δαμασκηνός, σαργάνη.

⚠️ Print the counter-example. Chapter 11 is not a record, and it is not even the densest chapter in its own letter. Over every Pauline chapter of ten verses or more, the densest are 2 Timothy 3 (22 hapax across 17 verses, 1.29 per verse) and Romans 16 (28 across 24, 1.17). Within 2 Corinthians itself, chapter 6 is denser than chapter 11 — 11 hapax across 18 verses, 0.61. So the right claim is the modest one: chapter 11 has an unusual amount of once-only vocabulary for this letter, which is what you would expect of a passage listing shipwrecks and floggings. It is not evidence of a different author, and nothing here is offered as such.

All 124 verses of 2 Corinthians 8:1–13:14, screened against every Greek Old Testament verse unit this shelf screens. 64 CATSS files less the Odes, 30,348 rows. Contiguous runs of whole word forms, never lemmas. Re-run 30 August 2026, when Exodus, Leviticus, Numbers, Joshua and the deuterocanonical books came onto the shelf; re-run again 31 August 2026, when the Odes came back out of the screen.

Nine more rows reach three and one more reaches four. Nothing above four moved at all. The cliff is still a cliff, 9:9 is still alone at eleven, and the corpus that produced those numbers is now a third larger. Where a row’s deepest partner is a deuterocanonical book — Sirach, Wisdom, 1–4 Maccabees, Tobit, Judith, 1 Esdras, Baruch, the Psalms of Solomon — this page names the book and says what it is: a witness, not Scripture.

⚠️ AND ON 31 AUGUST 2026 THE CORPUS GOT SMALLER, WHICH MOVED NOTHING ON THIS PAGE. THE HISTORY IS PRINTED RATHER THAN THE NUMBER SWAPPED. The Odes — the fourteen liturgical canticles bound after the Greek Psalter, 289 verse units, which arrived with the last twenty CATSS files — have been taken out of everything this shelf calls “the Greek Old Testament” for screening. They are still held on the disk; they are no longer screened. Two independent reasons, either sufficient: Ode 9 is Luke 1 (the Magnificat and the Benedictus) and Ode 13 is Luke 2 (the Nunc Dimittis), so a screen holding them can report a Gospel matching itself as an Old Testament contact; and ten more canticles are second copies of passages the corpus already holds — Ode 1 = Exodus 15, 2 = Deuteronomy 32, 3 = 1 Samuel 2, 4 = Habakkuk 3, 5 = Isaiah 26, 6 = Jonah 2, 7 and 8 = Daniel (Th) 3, 10 = Isaiah 5, 11 = Isaiah 38 — so any tie band touching a canticle was doubled. HELD is 30,637 verse units in 64 files. SCREENED is 30,348. Both numbers are correct and this page prints the second one, because a screen is what produced every figure here.

⚠️ What that did to this page: nothing, and the nothing is checkable. Re-run over 30,348 rows the distribution is still 124 / 121 / 72 / 12 / 3 / 2 / 1, the cliff is still a cliff, and every run and every tie band in the table below is unchanged — 9:9 still eleven at Psalm 111:9 in a band of one, 9:10 still six at Isaiah 55:10, 13:1 still five at Deuteronomy 19:15, 11:15 still four in a band of five, 8:15 still three in a band of three. The καθὼς γέγραπται census is still eight rows and δότης still stands exactly once. Thirty-three tie bands narrowed by one to forty-three rows apiece — the largest being 8:4’s 961 to 918 — and not one of them is a number this page prints. A page whose findings are indifferent to the edges of its corpus is a page whose findings were not living on those edges.

⚠️ But twelve of those thirty-three bands had Luke inside them, and that is worth saying out loud rather than filing under ‘nothing moved’. Ode 9 is Luke 1 and Ode 13 is Luke 2. Twelve rows of these six chapters were carrying an Ode 9 or Ode 13 partner inside their tie band — 8:10, 8:16, 9:11, 9:12, 10:1, 10:4, 10:14, 11:5, 11:18, 12:2, 12:3 and 12:8. 2 Corinthians 11:5 alone had seven of them, and 12:2 and 12:3 — the man caught up to the third heaven — each had Odes 9:68, which is the opening of the Benedictus. Every one of these was a run of one or two inside a band of hundreds or thousands, so none of them was ever printed and none of them changed a maximum. That is precisely why it is worth naming: the hazard does not announce itself as a spectacular false hit. It sits in the denominator. A tie band is how this shelf decides whether a run means anything, and for a day these bands were quietly counting the Gospel of Luke as Greek Old Testament. The instrument can do that, it did do it, and the way it was caught was reading what the partners actually were rather than how many there were.

The distribution has a cliff in it. Two rows reach six. Then nothing reaches seven, eight, nine or ten. Then one row reaches eleven. 2 Corinthians 9:9 is not on the same scale as anything else in the passage.

The second-deepest row is not a marked quotation at all. 2 Corinthians 9:10 says God supplies seed to the sower and bread for food and shares six consecutive words with Greek Isaiah 55:10, with a tie band of one — and Paul gives no signal that he is quoting. The page reports six shared words with a band of one and does not call it a citation. Nothing in 9:10 marks itself as scripture, and the direction of dependence is not something a run screen can see.

Paul writes καθὼς γέγραπται twice in these chapters, at 8:15 and at 9:9. On this shelf the two still measure very differently — but for a different reason than this page could give when it was first published.

What this page used to say here, in its own words: “The second line is not a fact about Paul. It is a fact about this shelf.” It was right that it was a fact about the shelf. It was wrong to expect that filling the hole would make the row look like 9:9. Greek Exodus arrived on 30 August 2026, the row moved from 2 to 3, and 3 is still a short run for a marked citation. The gap between Paul’s two καθὼς γέγραπται quotations is real, and it is about Paul.

Here is why the citation only measures three, set out so anyone can check it.

LXX Exodus 16:18  …οὐκ ἐπλεόνασεν ὁ τὸ πολὺ καὶ ὁ τὸ ἔλαττον οὐκ ἠλαττόνησεν
2 Corinthians 8:15  ὁ τὸ πολὺ οὐκ ἐπλεόνασεν, καὶ ὁ τὸ ὀλίγον οὐκ ἠλαττόνησεν

Two changes, of two different kinds. First, Paul puts the subject before the verb where Exodus puts the verb first. Same words, transposed — and a contiguous-run screen is blind to a transposition by design, so four shared words register as a run of three and a run of two. Second, he writes ὀλίγον where Exodus has ἔλαττον. That one substitution destroys the Greek Exodus’ own play on words — ἔλαττον … ἠλαττόνησεν, the adjective and the verb off one root — while keeping the verb that made it.

⚠️ And the word that is not Exodus’ is Numbers’. The three rows that tie at three are Exodus 16:17 and 16:18 — the manna, on ὁ τὸ πολύ — and Greek Numbers 11:32, the quail, which is the one place in the Greek Old Testament that reads ὁ τὸ ὀλίγον. All three rows that reach the maximum are wilderness gathering stories. The page reports that and draws nothing from it: it does not claim Paul blended two wilderness texts from memory, and it does not claim coincidence either. Both books were off this shelf until 30 August 2026 and neither could be searched.

⚠️ Two numbers this page printed have been falsified by the larger corpus, and both are corrected here rather than deleted. (1) It said “only six rows on this whole shelf carry καθὼς γέγραπται at all” — 2 Chronicles 23:18 and 25:4, 4 Kingdoms 14:6 and 23:21 (English 2 Kings), 1 Esdras 3:9, and Theodotion Daniel 9:13. Re-run over 30,348 rows the answer is eight. The two added are Tobit 1:6 in both the BA and the S recension — a witness, not Scripture. (2) The band analysis it printed — of the 602 rows tying at two, 588 matched on καὶ ὁ, four on τὸ πολύ, and the rest on πολὺ οὐ, ὁ τό or τὸ ὀλίγον — no longer describes anything, because the row no longer maxes at two. It is kept on the record because it was the correct reading of a corpus that no longer exists, and because a shelf that silently drops its superseded work teaches readers to trust numbers they cannot audit.

The Hebrew still lines up clause for clause, and still cannot be run. Exodus 16:18 in the Masoretic text: וְלֹא הֶעְדִּיף הַמַּרְבֶּה וְהַמַּמְעִיט לֹא הֶחְסִיר — the one gathering much had no surplus, and the one gathering little had no lack. Paul: ὁ τὸ πολὺ οὐκ ἐπλεόνασεν, καὶ ὁ τὸ ὀλίγον οὐκ ἠλαττόνησεν. Same two clauses, same order, same negation — and Paul’s clause order is the Hebrew’s, not the Greek Exodus’. That is worth noticing now that both columns are on the shelf, and it is as far as this page will go: Hebrew cannot be run against Greek, and no number is claimed from it.

2 Corinthians 10:17 is the most obvious Old Testament reference in the passage. Everyone recognises it. The screen does not.

The deepest partner of this sentence anywhere in either corpus is Paul’s own earlier letter. Jeremiah 9:22–23 in the Greek reads let not the wise boast in his wisdom… but let him who boasts boast in this, to understand and know that I am the Lord — three negative clauses and a long positive one. Paul reduces the whole of it to six words and puts ἐν κυρίῳ where Jeremiah has ἐν τούτῳ. Nothing consecutive survives. He did the same reduction in 1 Corinthians 1:31, where he does write καθὼς γέγραπται in front of it, and here he simply reuses his own sentence.

⚠️ What this row is doing on the page. It is here so that the eleven-word run at 9:9 is not read as more than it is. A deep run proves verbatim contact; a shallow run proves nothing at all. If the screen were the only evidence, 10:17 would look like a verse with no Old Testament behind it — and every commentator who says Jeremiah is standing behind it would still be right.

2 Corinthians 9:7 ends ἱλαρὸν γὰρ δότην ἀγαπᾷ ὁ θεός. CATSS Proverbs carries a row numbered 22:8a which reads ἄνδρα ἱλαρὸν καὶ δότην εὐλογεῖ ὁ θεός.

The Hebrew Proverbs 22 has no line for 22:8a. The CATSS row exists because the Greek Proverbs has a couplet the Masoretic column does not, and the editors gave it a lettered number rather than shifting everything after it. The nearest Hebrew verse in sense is 22:9, טוֹב עַיִן הוּא יְבֹרָךְ, the one good of eye will be blessed — but that is a different sentence, and measured against 2 Corinthians 9:7 its run is one.

⚠️ And here the run screen actively points at the wrong verse.

Sequence says Hosea; stock says Proverbs. The two distinctive words in Paul’s clause are ἱλαρόν and δότην, and Proverbs 22:8a is the only verse in either corpus that has both of them — the ἱλαρ- root stands 15 times across the Greek Old Testament held here (12 when this page was written over 44 files; the three added are 3 Maccabees 6:35 and Psalms of Solomon 4:5 and 16:12, witnesses, not Scripture), and δότης still stands exactly once in each corpus: once in 30,348 Greek Old Testament rows, at Proverbs 22:8a, and once in 7,927 Greek New Testament verses, at 2 Corinthians 9:7 — a negative re-run over the enlarged shelf on 30 August 2026 that did not move. Paul’s verb is ἀγαπᾷ where the Proverbs row has εὐλογεῖ, and his clause has no ἄνδρα and no καί, which is exactly why the contiguous run collapses to two.

What is and is not being claimed. Not that Paul is quoting Proverbs — he gives no citation formula, and he changes the verb. Not that the Greek line is earlier or later than the Hebrew, or that anything was added or lost: this shelf never says which column moved. What is claimed is measurable and narrow: the sentence English readers know as ‘God loves a cheerful giver’ has its closest verbal relative in a Greek Proverbs line that the Hebrew Proverbs does not carry. A reader with only a Hebrew-based Old Testament has no verse to turn to.

124 rows against the 7,671 New Testament verses outside 2 Corinthians.

Both six-word runs are liturgical formulas, and their tie bands say so. The blessing-formula at 11:31 stands in the same six words at Ephesians 1:3 and 1 Peter 1:3; the grace at 13:14 stands in the same six at Philippians 4:23 and Philemon 25. Two is a small band, but a band of two across three different authors’ letter endings is a formula, not a quotation. The page reports the words and does not assert who is drawing on whom.

And the internal bracket, measured the same way. 2 Corinthians 10:8 and 13:10 share six consecutive words, εἰς οἰκοδομὴν καὶ οὐκ εἰς καθαίρεσιν. Screened against the New Testament outside 2 Corinthians, 10:8’s deepest run collapses to three, with a tie band of two. Its own letter is the only place in the New Testament that clause occurs at all. 11:5 and 12:11 share three, τῶν ὑπερλίαν ἀποστόλων, and ὑπερλίαν occurs nowhere else in the New Testament at all.

This section listed three. Two of the three were filled on 30 August 2026, and both of them have been run. The table keeps what it said and adds what happened.

The full list of what is and is not held is printable at any time: python -m dr_tools coverage. It reports 64 CATSS files and 30,637 verse units HELD as this correction is made; it reported 44 files and 23,061 units when the page was written. ⚠️ HELD is not SCREENED. Since 31 August 2026 the Odes — 289 units of the 30,637 — are held on the disk but excluded from every run screen, so the corpus every figure on this page was measured against is 30,348.

No Greek critical apparatus. The SBLGNT apparatus sigla were stripped from the text before it was displayed and before it was measured, so where the SBL edition differs from NA28 or from the Byzantine text this page silently follows SBL. That matters least at 9:9 and most at 12:7, where the variants διό / καὶ τῇ and the placement of the repeated ἵνα μὴ ὑπεραίρωμαι are genuinely contested.

No papyri, no inscriptions, no coins — including the Nabataean and Damascene evidence for Aretas IV at 11:32, which is the one externally datable item in the letter. No patristic commentary. No Vulgate collation for these chapters. No Syriac. The English column is the ESV, and no other English version was collated against it.

And the ESV/Greek verse-count difference at chapter 13 was derived, not looked up. It rests on two files on disk: the MorphGNT verse index and an English chapter-shape file. No printed critical edition was consulted to confirm that the Greek tradition is uniform in numbering 13 verses — it is not, in fact, uniform, since some editions carry 14 — and this page reports the shape of the two files it holds.

Whether chapters 10–13 were originally part of the same letter as chapters 1–9 is the largest open question in the study of 2 Corinthians, and this page does not answer it.

What it can do is print the measurements honestly, and they cut both ways. For a seam: the ἀσθεν- family occurs fourteen times in the letter and every one is after 10:1; the ἀφρ- family occurs eight times and every one is between 11:1 and 12:11; the tone at 10:1 is different by any reading. Against a seam: the καυχ- family runs through both halves — 1:12, 1:14, 5:12 twice, 7:4, 7:14 twice, and then twenty-two more from 8:24 onward; συνίστημι appears in both halves and the argument at 10:18 answers the question raised at 3:1; ἁπλότης spans 8:2, 9:11, 9:13 and 11:3.

Word distribution cannot decide a partition question, and this page says so rather than pretending it can. A change of subject produces exactly the same vocabulary signature as a change of document. The Deep Research method finds verbatim contact; it does not find composition history.

Two more places this page is thin. First, the thorn: every identification of σκόλοψ — illness, eye trouble, malaria, epilepsy, a person, persecution — is a guess from outside the text, and the page offers none. Second, the collection’s outcome: whether the money was accepted at Jerusalem is not narrated in the New Testament, and speculation about it does not belong here.

• Margaret E. Thrall, 2 Corinthians (ICC), 2 vols.
• Murray J. Harris, The Second Epistle to the Corinthians (NIGTC).
• Ralph P. Martin, 2 Corinthians (Word Biblical Commentary 40).
• Victor Paul Furnish, II Corinthians (Anchor Bible 32A).
• Hans Dieter Betz, 2 Corinthians 8 and 9 (Hermeneia).
• Beale & Carson, Commentary on the New Testament Use of the Old Testament — on 8:15, 9:9, 10:17 and 13:1.
• None of these were consulted for the measurements on this page, which were run on the text files named in the footer.

Ten of the eighteen times χάρις appears in 2 Corinthians, it is in chapters 8 and 9, and in those chapters it means a sum of money being raised for hungry people in another country. It is the word for the fund, the word for the churches’ willingness to give to it, and the word for the privilege of being asked — the Macedonians beg for the grace of taking part (8:4).

And in the middle of the appeal, without changing words, Paul writes: you know the grace of our Lord Jesus Christ, that though he was rich, for your sake he became poor (8:9). The verb is one the New Testament never uses again. The argument for the offering is not a principle about stewardship. It is a sentence about the incarnation.

The last verse of the letter is ἡ χάρις τοῦ κυρίου Ἰησοῦ Χριστοῦ… — ESV 13:14, Greek 13:13. The word the letter spent two chapters using for a relief fund is the word it closes on.

The eleven-word run at 9:9 is the deepest thing on this page, and it is worth knowing what it means and what it does not. It means that Paul is reproducing eleven consecutive words of a Greek psalm, and that no other verse in 30,348 Greek Old Testament rows shares eleven consecutive words with him — a claim re-run on 30 August 2026 over a corpus a third larger than the one it was first made on, and unchanged. It does not tell you why he chose that psalm.

For that you have to read Psalm 112 — 111 in the Greek — and notice what kind of poem it is. It is a complete Hebrew acrostic, all twenty-two letters, and the line Paul quotes is the man’s open hand. Paul takes a description of a righteous man and applies it to a church sending money to a famine.

The acrostic, as it can actually be read off the Masoretic text. Psalm 112 has ten verses and twenty-two cola. Verses 1–8 carry two letters each and verses 9–10 carry three, and the opening הַלְלוּ יָהּ stands outside the pattern.

⚠️ And the run of eleven stops exactly where Paul stops. Greek Psalm 111:9 runs nineteen words. Paul writes eleven of them, in order, and then ends his sentence — every word he wrote after καθὼς γέγραπται is a word of the psalm, in the psalm’s order. What he leaves off is τοῦ αἰῶνος τὸ κέρας αὐτοῦ ὑψωθήσεται ἐν δόξῃ, …of the age; his horn shall be exalted in glory. In the Hebrew, τοῦ αἰῶνος is the tail of the Greek rendering of לָעַד, the last word of the צ colon, and everything after it is the ק colon entire. Paul quotes the pe and the tsade and does not quote the qoph. The run is 11 rather than 19 because the quotation ends, not because the two texts part.

How the acrostic was checked, and what the tool could not see. The shelf’s acrostic screen reads the first consonant of the first word of each verse, and on Psalm 112 it returns not an acrostic — it matches eight letters and reports fourteen missing. That is a limit of the screen, not a finding about the psalm, and the tool’s own caveat says so: a colon-internal acrostic is invisible to a verse-initial test. The table above was read off the Masoretic text by hand, colon by colon, and every letter in it is the first consonant of a word standing in the text with no emendation and no prefix stripped.

And the counter-example goes the other way. “Let the one who boasts, boast in the Lord” measures one against Greek Jeremiah. If a run screen were all we had, we would conclude there was no Old Testament behind that sentence — and we would be wrong. The measurement is a floor, not a ceiling. It can prove contact. It cannot disprove it.

Every one of the fourteen weakness-words in 2 Corinthians is in the last four chapters. The first of them is not Paul’s (10:10) — it is the charge against him, quoted back: his bodily presence is weak. He spends three chapters taking the word over.

He asks three times for the thorn to go. The answer he gets is not the removal and not an explanation: my grace is sufficient for you, for my power is made perfect in weakness. The verb at the end of that sentence, ἐπισκηνόω, is a New Testament hapax and it means to pitch a tent over.

The second of the shelf’s two pages on 2 Corinthians, covering the second half of the letter — chapters 8 through 13. The first half is at 2 Corinthians 1:1–7:16. Every chapter in the scope has at least one verse in the codex above.