The Book of the Heavenly Luminaries

• Anchor Bible Dictionary — its articles on the books of Enoch and on calendars.
• Eerdmans Dictionary of Early Judaism — its entries on 1 Enoch, the Astronomical Book and the calendars.
• Dictionary of New Testament Background (IVP) — its entries on the Enoch literature and on calendars. These are subjects to look up; the exact entry titles were not checked, and none of these works is held on this disk.
• On the text — any article that says which Ethiopic text it reads (Dillmann, Flemming, Charles, Knibb, the Ethiopian Orthodox printed Bible) and whether it is reading the Geʻez or the Aramaic. In these chapters the numbers differ between editions and even between a printed page and its digitisation; an article that quotes ‘the Ethiopic’ without saying which is not usable for them.

• R. H. Charles, The Book of Enoch, or 1 Enoch (Oxford: Clarendon Press, 1912) — the English quoted here, and the notes this page quotes from an OCR scan.
• A. Dillmann, Liber Henoch aethiopice (Leipzig, 1851) — the Geʻez measured here, and the printed pages read for this study.
• J. T. Milik, The Books of Enoch: Aramaic Fragments of Qumrân Cave 4 (Oxford, 1976) — named in the shelf’s acquisition record as the first edition of the Aramaic; not held, not consulted.
• M. A. Knibb, The Ethiopic Book of Enoch (Oxford, 1978) — not held, not consulted.
• James C. VanderKam, Calendars in the Dead Sea Scrolls: Measuring Time — on the 364-day year; not held.

Jubilees 6 argues the same year in other words. In Charles’s English Jubilees the four memorial days stand at the new moons of the first, fourth, seventh and tenth months, and “each had thirteen weeks” (6:29) — thirteen weeks are 91 days, the length 1 Enoch 82:15 and 82:18 give each leader’s season. Moses is to command Israel to keep “three hundred and sixty-four days” (6:32), and the moon is blamed because it “cometh in from year to year ten days too soon” (6:36) — 364 less 354, the yearly gap that 1 Enoch 74:14–16 counts in fives and eights. The two books agree on 364, on 91, on four days that divide the year into quarters and on ten — though Jubilees sets its four days at new moons and 1 Enoch sets its four by gates (75:2). Measured in Geʻez, they share almost no wording — Stage 4.

Jubilees also says what Enoch wrote. “who wrote down the signs of heaven according to the order of their months in a book” (4:17), and the angels showed him “the rule of the sun” (4:21). Jubilees’ claim about Enoch, reported; no line of descent between the two books is drawn here.

And a Second-Temple book that reckons the feasts by the moon. Sirach 43:7, CATSS Greek: ἀπὸ σελήνης σημεῖον ἑορτῆς φωστὴρ μειούμενος ἐπὶ συντελείας — KJV Apocrypha: “From the moon is the sign of feasts, a light that decreaseth in her perfection.” The Psalter gives the moon the appointed times (Psalm 104:19, below). 1 Enoch’s moon has a full course of its own, but the year it is measured against is the sun’s 364.

• Psalm 104:19 (Masoretic; LXX Psalm 103:19) — עָשָׂ֣ה יָ֭רֵחַ לְמוֹעֲדִ֑ים, “He made the moon to mark the seasons; the sun knows its time for setting.” (ESV). CATSS Greek: ἐποίησεν σελήνην εἰς καιρούς ὁ ἥλιος ἔγνω τὴν δύσιν αὐτοῦ. Set beside 1 Enoch as Scripture’s own sentence on the two lights; no dependence either way is asserted.
• The Qumran Aramaic Astronomical Book — 4Q208, 4Q209, 4Q210, 4Q211: 336 manuscript lines in 89 fragments or fragment columns (80 fragments when a fragment’s columns are counted once), 4,144 letters preserved. Only 5 of those units (four fragments; 4Q210 frg. 1 counts twice, for its columns ii and iii), 970 preserved letters (23%), could be placed against verses of 72–82 in the shelf’s placement file; the rest is described in the acquisition record as mostly a synchronistic lunar calendar with no Ethiopic counterpart. Most of it has no verse of the Ethiopic to match.
• R. H. Charles, 1912 — argued that 72–82 came “From a Hebrew original”, partly from the word-play on the quarters in 77:1–2 (Stage 4), and that the eight-year totals of 74:13–16 show the writer knew the Greek eight-year cycle. His arguments, reported; the page tests only what the Aramaic letters preserve.

• 154 verses, 2,601 Geʻez word forms, 210 of them carrying a numeral sign. Per chapter, 72 to 82: 56·7·37·10·16·10·39·3·0·3·29.
• Six gates east and six west for the sun (72:3), twelve doors in the sun’s chariot (75:4), twelve portals at the ends of the earth (75:6), twelve gates of the winds (76:1) — and twelve leaders of the months (82:11).
• Two names for the sun and four for the moon (78:1–2), and four leaders of the quarters named at 82:13, of whom 82:15–20 describes two.
• Three holy ones or seven? At 81:5 Charles reads those seven holy ones; Dillmann’s page has ወእሙንቱ ፡ ፫ቅዱሳን, three, and so does the Ethiopian Orthodox edition. Charles’s note gives seven from the group of manuscripts he calls α and three from the group he calls β.

39 statements in Charles’s English — the sun’s months and day-lengths, and the yearly and multi-year totals — were recomputed from the chapter’s own numbers (what was not checked is named below the table): the sun’s year of 364, and the moon’s year of 354 that the text never states but builds from three months of thirty days and three of twenty-nine, twice over (78:15–16).

35 of 39 close. Three do not, and they are exactly the three words Charles daggered in 72:10–32 (72:12, 72:18, 72:28): a step of one part called two twice, and called a ninth once. The fourth, the six-day overplus of 74:11, closes on a year of 360, not 364 — as Charles’s own note to 74:10–11 observes (“360 days. Thus the difference in this case is six days”). Charles judged the system “most perplexing”; measured on his own English, the sums checked here close everywhere except at the four places named. Not checked: the moon’s sevenths and fourteenths (73:3–8, 74:3, 78:6–8), her courses of seven and eight days (74:6–8), her twenty days (78:17), and the months of twenty-nine days and once twenty-eight (78:9) — a length 78:15–16, with its months of thirty and twenty-nine, does not use; Charles’s 1912 note refers the twenty-eight to the cycle of Callippus.

Dillmann’s Ethiopic parts from Charles’s text at two of these verses. At 72:28 Dillmann’s page reads one part (ዝውእቱ ፡ ክፍል ፡ ፩ወትከውን), which closes; Charles’s note gives ninth from three of his manuscripts. At 72:10 the digitisation reads ትነውኅ ፡ ካዕበተ ፡ ዕለት ፡ እምሌሊት and Dillmann’s page (p. 47, read for this check) has the same words; Charles’s 1912 note renders the reading of the manuscripts “the day is longer by twice as much than the night” and sets aside for the ninth part that it finds in three manuscripts (g, q, u); read as twice as much, it does not close (ten parts against eight). At 72:12 and 72:18 Dillmann prints two, as Charles’s text does — and the chapter’s own numbers need one.

The Geʻez writes the year in figures six times in these chapters: 72:32, 74:10, 74:12, 75:2, 82:6, 82:11. Dillmann’s printed page of 1851 has 364 at all six. The free digitisation the shelf’s Geʻez screens read — Jerabek’s of 1995, as cleaned by HaCohen in 2011 — has 364 at three: 72:32 = 364, 74:10 = 364, 74:12 = 310, 75:2 = 312, 82:6 = 312, 82:11 = 364. At 74:12 the sign for sixty has become the sign for six; at 75:2 and 82:6 it has become the sign for eight; at 74:14 the same slip turns the moon’s three years from 1,062 to 1,010, and the 62 of the bracketed sum to 10.

24 numerals were read on the page — the six years; the two totals of 74:14 and the eight years of 74:13–16; every numeral of chapter 72 where the digitisation and Charles’s English disagree; 81:5; and controls where the two agree. 10 differ from the digitisation. The other five are in chapter 72: the sun’s six western gates keyed as seven (72:3), those six portals as sixty (72:6), the day’s step of two as twenty (72:12), the night of six parts as sixty (72:14) and the day of six as seven (72:26). Nine of the ten break a count that holds on the printed page; at 72:12 the page’s own two already misses by one (§1), and the digitisation makes it twenty. The same slips stand in HaCohen’s own pages and in the copy Beta maṣāḥəft carries, so they belong to the digitisation, not to the shelf’s cutting of it. The other numerals of these chapters were not re-read one by one; ten is a floor.

The normalisation, declared — the same as the Jubilees pages. Words are split on the Geʻez divider and on spaces; the full stop and other punctuation and any brackets are removed; every word form is kept as written, numerals glued to their word as the digitisation glues them. A run is consecutive identical forms; 0 means the verse shares no two consecutive forms with any verse.

Against all 1,308 rows of VanderKam’s Geʻez Jubilees, the 154 verses: 75 share no two consecutive word forms, 76 reach exactly two, 3 reach exactly three, and nothing reaches four. Median 2. With the laryngeal and sibilant letters folded together: 74 share no two consecutive word forms, 77 reach exactly two, 3 reach exactly three, and nothing reaches four.

Against Jubilees 6:23–38, the calendar chapter, alone: 151 of the 154 verses share no two consecutive forms, and the other 3 share two (until eternity at 72:1 and 82:2, your children that at 81:5; band 1, 1, 1 within those sixteen verses). Against all 1,308 Jubilees rows the same three verses (72:1, 81:5, 82:2) still reach only two, with bands of 29, 50, 21 — two-word runs shared that widely do not by themselves show contact. From Jubilees’ side, its sixteen calendar verses against all 1,062 rows of Geʻez 1 Enoch: 9 share no two consecutive forms with any verse of the book; none goes beyond 2. The one number the two books most insist on cannot even be seen by this screen: 1 Enoch writes 364 in figures, the Geʻez Jubilees in words.

Four runs reach three, and none is deeper than the others. One is a list of three nouns: 82:8 and Jubilees 2:8 share ፀሐየ ፡ ወወርኀ ፡ ወከዋክብተ, the sun and the moon and the stars — and Jubilees 2:8 is the fourth day of creation. The other three are men, for they (75:2, Jubilees 10:8), all who walk (82:4, Jubilees 42:20) and in heaven that (82:8, Jubilees 30:23) — the page’s literal glosses.

Against all 1,526 rows of the Geʻez Genesis: 84 share no two consecutive word forms, 66 reach exactly two, 4 reach exactly three, and nothing reaches four. Median 0. The four runs of three, each with its band on the Enoch side (among Genesis rows) and on the Genesis side (among all 1,062 Enoch rows): 72:36 ↔ Genesis 7:16, as the Lord commanded, band 1 and 1; 76:13 ↔ Genesis 24:62, which is toward the south, band 1 and 1; 80:7 ↔ Genesis 36:20, dwell upon the earth, band 1 and 3; 82:19 ↔ Genesis 34:28, and all that is in, band 1 and 1 (the page’s literal glosses). Folded: 83 share no two consecutive word forms, 67 reach exactly two, 4 reach exactly three, and nothing reaches four. The page’s counter-example: Genesis 1:14–19, the making of the lights, against these chapters — 1:14: 2; 1:15: 2; 1:16: 0; 1:17: 2; 1:18: 0; 1:19: 0. The chapter about the lights and the treatise about the lights share no more than two consecutive words in Geʻez (upon the earth, heaven that); 1:14, 1:15, 1:17 each reach two, tied with 7, 7, 7 of the 154 verses here, and three of Genesis’s six verses share no two consecutive forms with any of them.

Only preserved letters are printed or counted. Of the Aramaic placed against these chapters:

• 76:3–10, 4Q210 frg. 1 ii — the winds of the gates: בתרעא קדמיא נפקא רוח קדים, through the first gate goes out the east wind.
• 76:14–77:3, 4Q210 frg. 1 ii and 4Q209 frg. 23 — תרי עשר תרעי ארבע רוחי שמ, twelve gates of the four quarters of hea[ven]; the quarters named (below); and the three parts of the earth, חד מנהון למדבר בה בני אנשא, one of them for the sons of men to dwell in.
• 78:6–8, 4Q210 frg. 1 iii — the moon filling to the fourteenth and fifteenth day: עד יום ארבעת עשר … חמשת עשר ומשלמין בה כל נהורה, until day fourteen … fifteen, and all her light is completed in her.
• 78:17–79:2 (placed loosely), 4Q209 frg. 26 — כדמות אנש, like a man (78:17: at night she appears like a man), and וכען מחוה אנה לך ברי, and now I show you, my son (79:1).
• 82:9–13, 4Q209 frg. 28 — מעדיהון לחדשיהון, their festivals, their months, and לן שמהת, [and th]ese are the names (82:13).

The quarters. Charles in 1912, arguing that these chapters came “From a Hebrew original”, wrote of 77:1 that “In this verse there is a play on the four quarters of the earth”, that it could be recovered in Hebrew or Aramaic for the east and the north but not for the south and the west, and of the south — because the Most High will descend there — “This is possible only in Hebrew.” The Aramaic copy, published by Milik in 1976, forty-five years after Charles’s death, gives both those quarters a wording the Ethiopic does not have, worked in Aramaic words: the south, דרום, because there the Great One dwells (דאר רבא) — the Ethiopic reads descends; the west, מערבא, because there the stars ערבין, set — the Ethiopic calls the west diminished, because the lights wane and go down: the setting is in both, the name is not. It also explains the north, בדי בה צפנין, because in it they hide, and a second name for the east, מזרח בדי… זרחין, because … they rise — neither of which the Ethiopic has: it names the north and goes straight on to the earth’s three parts (which the Aramaic also has, in line 8), and it has one name for the east.

None of this settles the book’s first language, and the page does not try. It settles that Charles’s “only in Hebrew” was said of a wording the Aramaic does not have. The word for south, דרום, is also Hebrew: the Masoretic text uses it 17 times, 13 of them in Ezekiel, and once, beside the north and the wind, in the passage on the sun’s round — Ecclesiastes 1:6, הוֹלֵךְ֙ אֶל־דָּר֔וֹם וְסוֹבֵ֖ב אֶל־צָפ֑וֹן, “The wind blows to the south and goes around to the north; around and around goes the wind, and on its circuits the wind returns.” (ESV; the CATSS Greek begins its 1:6 one clause earlier, ἀνατέλλων αὐτὸς ἐκεῖ πορεύεται πρὸς νότον).

The only Greek of these chapters is Oxyrhynchus Papyrus 2069, identified by Milik, in 4 scraps. 77:7 in full: ... τὴν] ἐρυθρὰν θ[άλασσαν .... All four hold 13 whole preserved words. Against the 7,927 verses of the Greek New Testament their deepest run is 2 — into the (band 283), and in (band 138), whole of the (band 2: 1Jn 2:2, Ac 10:22). Against the 30,348 Greek Old Testament units screened, also 2, four times: into the (band 759), and in (band 1,017), whole of the (band 2: 1/3Kgs 6:10, 2/4Kgs 9:8) and the Red [Sea] (band 2: Ps 105:7, Ps 105:9). Those references are CATSS labels: the two Psalm verses are Psalm 106:7 and 106:9 in the Hebrew and English Psalter, and 1/3Kgs and 2/4Kgs are CATSS’s names for 1 and 2 Kings (3 and 4 Kingdoms). Two common words are not contact, and the page reads nothing into them. Normalisation: the shelf’s grk_seq_cmp; unigram fallback on, so a 0 would mean no word at all.

• Galatians 4:10; Colossians 2:16–17 — days … and months and seasons and years; a feast or a new moon or sabbaths, which are a shadow of the things to come (the page’s literal glosses of the SBLGNT). Paul names calendars without naming whose. The page draws no line from 1 Enoch to the people Paul answers.
• Luke 21:25; Mark 13:24–25 — signs in sun, moon and stars before the end; beside 1 Enoch 80:4–7, where the moon alters her order and the stars leave theirs.
• Jude 13 — wandering stars (literal gloss). Charles’s 1912 notes set Jude 13 beside the stars of 1 Enoch 18, not beside 80:6; it is named here only because 80:6 also has stars that transgress their order.
• Jude 14–15 — Jude names Enoch, the seventh from Adam (literal gloss), and quotes words that 1 Enoch 1:9 also carries (the first page of this series). That is what Jude’s text does; it does not make 1 Enoch Scripture.

• Ten numerals of the Geʻez digitisation differ from Dillmann’s printed page (Stage 4 §2, all listed). Five of them are the sign for sixty: in the year at 74:12, 75:2 and 82:6, and twice in 74:14. Any tool that parses the digitised numerals of 1 Enoch 72–82 will report a year of 310 or 312 days in three places where the book says 364. The slips are in HaCohen’s own pages and in the Beta maṣāḥəft copy, not introduced by this shelf.
• Two letter slips noticed in passing: 72:26 is digitised ወመዓት where the page reads day in full, and 72:32 ወረብዐ where the page has the vowel u. Neither changes a number.
• The English file drops Charles’s daggers at 77:5. At 77:5 both scans on disk — the 1912 Oxford printing and the SPCK printing — print †west†; the transcription reads plain west. This page does not quote 77:5.

• Every other numeral. 24 numerals were read on the printed page; the rest of the 210 numeral-bearing word forms were not, so ten slips is a floor.
• The manuscripts. Dillmann’s apparatus, Charles’s Ethiopic edition and Knibb’s are not held. Where this page reports what “Charles’s manuscripts” read (82:11, 81:5, 72:28), it reports his 1912 notes, read from an OCR scan.
• The Aramaic editions. The Qumran text is SQE’s transcription; Milik’s 1976 edition and the DJD volume were not consulted, the preserved/restored marking is taken as given, and the placement of fragments against Charles’s verses is the shelf’s acquisition record’s reading, not an editor’s.
• Any Greek or Latin of 72–82 beyond 13 whole words of papyrus. The Greek New Testament screen cannot run in any useful sense.
• The Ethiopian Orthodox printed edition. Its rights holder is not named, so it is measured and never printed; its readings are reported only as numbers.
• Every claim about date and first language. Charles’s arguments are reported as his; nothing on this page dates the text.

• Its headline finding is partly about a digitisation, not about the book. The book’s own claim — 364 — stands in every printed witness this page saw. What is new here is that the machine-readable copy breaks it, and that matters for anyone who measures 1 Enoch by computer, including this shelf.
• The page images were read by eye and by a simple glyph comparison, not by an Ethiopist. The four sixty-signs score clearly; the sixes and sevens of chapter 72 rest on eye-comparison with Dillmann’s own margin numerals.
• The English is quotation, not a measured parallel, and the Geʻez it is set beside is not the text Charles translated — he worked from his own manuscripts. Where the page found them parting it says so: 72:10 and 72:28 in Stage 4, 81:5 in Stage 3, 82:11 in the codex. Only those places and the 24 numerals were compared; the rest of the Geʻez was not read against Charles verse by verse.
• The arithmetic checks were chosen by the page: the statements of day-lengths, months, years and multi-year totals in the Charles text of 72–82. The moon’s sevenths and fourteenths (73:3–8, 74:3, 78:6–8), her courses of seven and eight days (74:6–8), her twenty days (78:17) and the once twenty-eight days of 78:9 were not checked. A differently drawn list would give different totals; the list is printed in full.

• R. H. Charles, The Book of Enoch, or 1 Enoch (Oxford: Clarendon Press, 1912; SPCK reprint 1917).
• A. Dillmann, Liber Henoch aethiopice, ad quinque codicum fidem editus (Leipzig: F. C. G. Vogel, 1851) — text, and the printed pages read here from the archive.org scan.
• Scripta Qumranica Electronica, 4Q208–4Q211 — transcription by Reinhard G. Kratz, Ingo Kottsieper and Annette Steudel after the Qumran-Wörterbuch (Göttingen) and Martin Abegg, CC BY-SA 4.0.
• Online Critical Pseudepigrapha, 1 Enoch (I. W. Scott et al.), CC BY 4.0 — P.Oxy. 2069 after A. S. Hunt (1927).
• J. T. Milik, The Books of Enoch (1976); M. A. Knibb, The Ethiopic Book of Enoch (1978) — named, not consulted.

• Check the number before you quote it. The book writes 364 six times. The free digitisation this shelf reads has it three times. The difference is one numeral sign each time, and this page found it by opening the 1851 page.
• A careful text can still be wrong about God. These chapters count with real care, and the sums this page checked close. That does not make their calendar binding, and it does not make the book Scripture.
• The church’s view, labelled as such: Pleasant Springs reads Colossians 2:16–17 as settling that no calendar — 364 days, 354, or any other — is a test of faithfulness; festivals and new moons are a shadow of the things to come, and the body is Christ’s (Colossians 2:17, literal gloss).
• The sky is still a sign. Scripture’s own sentence is short: “He made the moon to mark the seasons; the sun knows its time for setting.” (Psalm 104:19). This week, go out once at sunset or moonrise and thank God for the order of the lights — without needing to count it.