Abraham and the Daylight Stars

A close exegetical study of Genesis 15:5 and the hour at which the star-promise was given

The hour hidden in the sequence

Genesis 15 is one of the load-bearing chapters of the Bible. It contains the verse Paul quotes to prove that a man is justified by faith, the covenant rite in which God alone passes between the divided animals, and the promise that made Abram the father of a countless people. It is read in synagogues and at Mass, preached on, argued over, and set to music. And it contains a detail whose reception is easy to flatten: on the face of the chapter, Abram is told to count stars before the chapter narrates its sunset.

The Douay–Rheims prints the promise at Genesis 15:5:

And he brought him forth abroad, and said to him: Look up to heaven and number the stars if thou canst. And he said to him: So shall thy seed be.¹

Seven verses later the same chapter says this:

And when the sun was setting, a deep sleep fell upon Abram, and a great and darksome horror seized upon him.²

And five verses after that:

And when the sun was set, there arose a dark mist, and there appeared a smoking furnace, and a lamp of fire passing between those divisions.³

The order is not in doubt. The star-promise stands at verse 5. The sun is still in the sky, going down, at verse 12. Full dark arrives at verse 17. Read the chapter forward, in the order it gives its own events, and the command to number the stars falls in the daytime.

That is the claim this study exists to test. It is not a claim that the chapter is incoherent, and it is not a claim about what Abram’s eyes physically registered on some afternoon in the second millennium before Christ. It is a claim about the text: that Genesis 15, taken in its own sequence, places a command to look at stars before its own sunset, and that every reading which removes the difficulty has to supply something the chapter does not say.

Why the difficulty is easy to miss

Three habits conceal it.

The first is that we read the verse for its content and not its staging. “Number the stars if thou canst” functions in Christian and Jewish memory as a figure of multitude, interchangeable with the sand of the sea and the dust of the earth. Once the verse has become a figure, the scene around it stops being consulted. Nobody asks what time it was when a metaphor was uttered.

The second is chapter-level reading. Verses 5, 12, and 17 belong to different paragraphs in every modern printing; verse 5 sits under the heading of the promise and verses 12 and 17 under the heading of the covenant rite. A reader who takes the chapter as two episodes will not put the two clocks side by side, and if he does, he will assume without noticing that a night has passed somewhere in between. Whether that assumption is warranted is precisely the question.

The third is versional. Swete’s identified Greek printing has no thick darkness at verse 17 and prints “there” rather than “in that day” at verse 18. That particular witness makes the closing clock less insistent. Augustine’s contrary Old Latin at verse 18 prevents projecting Swete’s wording onto Greek transmission generally or onto the texts known to the Fathers.

What a claim like this needs before it can be believed

An observation about a biblical text that seems novel is owed a demanding reception search. The responsible first move is not to defend it but to look for readers who confronted the same sequence and determine what they said.

That search is the body of this study. It has four results, and it is worth stating them at the outset so that the reader can watch them being established rather than waiting to be told:

1. The difficulty was seen. It is stated, in almost exactly the terms above, as the seventh of twenty-one prefatory questions on this chapter by Don Isaac Abarbanel, writing at the end of the fifteenth century. He puts the dilemma as a dilemma: it cannot have been day, because stars are not visible by day; and it cannot have been night, because of “and the sun was about to set” at verse 12.

2. Abarbanel’s own resolution goes against the claim, and so does the resolution of every other interpreter who addresses the hour. The tradition divides into at least five distinct answers, and they contradict one another; what unites them is that all five refuse the daylight.

3. Christian reception is not silent. Augustine does not ask the exact question at the checked loci, but Ambrose reads the going-out as visionary, while Chrysostom and Ephrem supply night–day–evening chronologies. Theodoret gives only a local non-treatment, and Jerome remains unchecked.

4. The reading that “he brought him forth abroad” means brought out from under the government of the stars, rather than merely out of doors, is attested — and attested far earlier than the medieval commentaries usually credited with it. Philo has it in the first century, attached to a different verse of the same chapter.

The judgment this study reaches, with its reasons, its strongest counterargument, and the finding that would overturn it, is stated in a labelled block at the end. The reader who wants the conclusion before the evidence will find it there; the sections in between are the reason it is worth anything.

A word about what is being quoted

Scripture in English is quoted only from the registered public-domain Douay–Rheims in Bishop Challoner’s revision, and always in the Vulgate chapter-and-verse numbering that witness uses. Where the Douay numbering differs from a reader’s modern Bible — most importantly, the book this study cites as 2 Esdras is Nehemiah in a Protestant or postconciliar Catholic Bible — the difference is flagged at the point of use.

Hebrew, Aramaic, Greek, and Latin are quoted in their own words where the argument turns on them. Hebrew and Greek are given in a plain romanization without diacritical points, and any English rendering of a Hebrew or Greek word offered by this study is labelled a gloss: it is a working equivalence made to show what a word is doing in an argument, and it is never offered as a translation of Scripture. Where an existing published translation is quoted, its translator, edition, and rights status are identified, and it is that translator’s wording and not the project’s.

The chapter’s own clock

Genesis 15 marks time four times. Once implicitly, at verse 5, by commanding an act that presupposes a condition of the sky; and three times explicitly, at verses 12, 17, and 18. Everything in the argument depends on getting those four markers right in each of the witnesses that a Catholic or Jewish reader might reasonably consult, so this section sets them out before any interpretation is attempted.

The four markers

Two things about this table matter more than they look.

The first is that verse 5 is the only place in Genesis 15 — and, as the later section on the recurrences will show, the only place in the whole biblical star-promise tradition — where somebody is told to look. Everywhere else the stars are a figure inside a speech. Here they are the object of a command to perform an act of vision, and an act of vision has an hour.

The second is verse 18. Bayyom hahu does not merely conclude; it retroactively binds. It says that what has just been narrated happened on a day, in the singular. If the Hebrew of verse 18 is doing what it appears to do, then the chapter itself asserts that the star-promise and the passage of the fire between the pieces fall inside one twenty-four hour span — which is exactly the assertion that makes the daylight problem sharp, because a single day containing verse 12’s approaching sunset cannot also contain a separate, earlier, unmarked night for verse 5 without saying so.

The markers across the witnesses

The four witnesses relevant to a Catholic reader are the Masoretic Hebrew, the Septuagint, the Clementine Vulgate, and the Douay–Rheims made from it. They do not agree. The disagreements are small in extent and large in consequence, and they run in a consistent direction.

Read the Septuagint column downward and a different chapter appears. Its sun never finishes setting: at verse 12 it is “about the setting”, and at verse 17 it is still egineto pros dusmais, becoming toward its setting.⁴ It has no thick darkness at all. And at verse 18 it says not “in that day” but “there” — a marker of place where the Hebrew and Latin have a marker of time.

Swete’s printed Greek therefore makes the close less temporally insistent: it omits the Hebrew darkness phrase and prints a spatial marker at verse 18. That is a finding about this identified base text, not about every Septuagint witness and not an explanation of Christian reception.

A necessary qualification belongs here rather than in an appendix. The ekei at verse 18 is the reading of the codex Swete prints, and it is not universal: Augustine, reading Genesis 15 in an Old Latin version made from Greek, quotes verse 18 as In die illa disposuit Dominus Deus testamentum ad Abram — “in that day” — which means the Greek exemplar behind his Latin had the temporal phrase.⁵ Augustine’s Latin agrees with Swete’s Greek at verses 11 and 17 and disagrees at 18. The honest statement is therefore narrow: Swete’s identified printing omits the Hebrew darkness phrase at verse 17 and prints “there” at verse 18, while Augustine’s Old Latin preserves “in that day.” This does not identify Augustine’s Greek exemplar or establish a history of Greek transmission.

What the Latin does

Jerome’s rendering, as printed in the Clementine recension, tightens the sequence rather than loosening it. Three details deserve notice.

Eduxitque eum foras. Foras is an adverb of motion meaning out of doors — the word one uses for going outside a building. Where the Greek has the colourless exo, “outside”, the Latin specifies leaving an interior. The Vulgate thus commits itself, more than the Greek does, to the plain domestic reading: Abram was under a roof and went out from under it. That is a commitment about space, and it makes the question about time harder rather than easier, because a man who steps outdoors to look at the sky is a man for whom the state of the sky is the point.

Suspice caelum. Suspicere is to look up at. So is the Septuagint’s anablepson. Both ancient versions therefore render the Hebrew habbet as an upward look. This will matter in the section on “outside”, because one influential rabbinic reading of the same Hebrew verb insists that habbatah is looking downward, from above — and neither ancient translator saw it that way.

Cumque sol occumberet … Cum ergo occubuisset sol. The Latin distinguishes the two sunsets by tense with a precision the English inevitably blurs: an imperfect subjunctive at verse 12 for a setting in progress, a pluperfect subjunctive at verse 17 for a setting completed. And the ergo at verse 17 — “therefore”, “so then” — is a continuative particle. It presents verse 17 as the resumption of a narrative already in motion, not the opening of a new scene.⁶

What the English witness adds

The registered Douay–Rheims follows the Latin exactly at all four markers, and its chapter argument — the one-line summary Challoner prints above the chapter — reads: God promiseth seed to Abram. His faith, sacrifice and vision. The word vision in that summary is not decoration. It comes from per visionem at verse 1, and Challoner’s decision to carry it into the chapter heading tells a reader that the eighteenth-century English Catholic tradition regarded the whole chapter, and not merely its opening sentence, as falling under a visionary frame. Whether that frame governs the star-promise, and what follows if it does, is the subject of the section after next.

Neither Challoner nor the 1609 Douay Old Testament from which his revision descends prints any note at verse 5.⁷ Neither does Haydock’s nineteenth-century English Catholic commentary, which is otherwise generous with notes on this chapter. That silence is itself evidence, and it is picked up again in the section on the reception.

What the Hebrew actually does

The whole weight of the claim rests on one question of Hebrew narrative syntax: does the chain of verbs running from verse 1 to verse 18 assert that its events happened in the order it gives them? If it does, the star-promise precedes the sunset and the daylight reading follows. If it does not — if Hebrew narrative routinely permits an unmarked interval, a resumption, or a flashback — then the chapter is compatible with a night at verse 5 and the claim collapses into one possible reading among several.

This section takes that question seriously enough to answer it in both directions.

The verbs, in order

The forms are these. Romanization is plain and unpointed; the glosses are glosses.

Three observations follow directly from the forms and require no interpretation.

First, the chain is unbroken. From verse 5 through verse 18 the narrative advances by consecutive preterites — the form that is the ordinary backbone of Hebrew narration and whose default force is sequence. There is no disjunctive clause, no fronted subject, no circumstantial construction, and above all no temporal expression of the kind Hebrew uses freely when it wants to signal a gap: no “on the morrow”, no “and it was morning”, no “after these things”. The chapter has such an expression exactly once, and it is at the head, at verse 1: ahar hadevarim ha’elleh, “after these things”, separating chapter 15 from the war of the kings in chapter 14. Rashi comments on that phrase specifically to say that ahar without the final yod means immediately after, as against aharei, which means at a distance.⁸ Whatever one makes of the derivation, the point stands that the one temporal joint in the chapter is the one at its head, and that the tradition read even that one as close rather than distant.

Second, verse 12 and verse 17 use the same opening formula, wayhi plus a temporal expression, and are distinguished from one another only by the form of the verb of setting. That is a strong indication that they are two moments of one evening rather than two evenings. If the author had wanted to mark a new day at verse 12, the resources to do so were available and were not used.

Third, verse 5 contains two speech verbs: wayyomer, “and he said”, introducing the command to look, and wayyomer lo, “and he said to him”, introducing “so shall thy seed be”. Doubled speech verbs inside a single verse are a genuine feature of the chapter — verses 2 and 3 have the same doubling for Abram — and, as the section on Abarbanel will show, one interpreter builds a whole reconstruction of the episode on them. The Septuagint, incidentally, lacks the pronoun in the second: it has simply kai eipen, while the Vulgate has Et dixit ei with the Hebrew.

Does the consecutive preterite require sequence?

Here the honest answer is: strongly, but not absolutely.

The consecutive preterite is a sequential form. It is what Hebrew narrative uses to say “and then”. But Hebrew prose also uses it in ways that are not strictly sequential: to resume after a digression, to summarize what has already been narrated in detail, and — most relevantly here — after a wayhi-plus-temporal-clause that opens a new scene at an interval the narrator has not troubled to measure. A reader who wants a night before verse 5, or a day between verse 6 and verse 9, is not proposing something ungrammatical. He is proposing an unmarked scene change, and unmarked scene changes exist.

What that reader is doing, however, is supplying the interval rather than reading it. The distinction matters, because the difference between the daylight reading and its rivals is not a difference about what Hebrew can do. It is a difference about who bears the burden. The daylight reading takes the chapter as it stands and accepts the consequence. Its rivals take the consequence as intolerable and insert a break to avoid it. Both are legitimate moves in exegesis. Only one of them adds material.

The rabbinic rule about order, and its limit

There exists in rabbinic hermeneutics a principle that would, if it applied, dissolve the difficulty at a stroke: ein muqdam ume’uhar batorah, “there is no earlier and later in the Torah” — that is, the Torah does not narrate in strict chronological order. It is a real and frequently invoked principle, and it is exactly the sort of thing a reader reaching for a harmonization would reach for.

It will not do the work, and the reason is stated in the same passage of the Talmud that transmits the principle. At Pesachim 6b, Rav Menashiya bar Tahlifa reports the rule in the name of Rav; and immediately, Rav Pappa limits it:

Amar Rav Pappa: la amaran ella bitre inyane, aval behad inyana, mai demuqdam — muqdam, umai dime’uhar — me’uhar.⁹

Gloss of the Aramaic: we said this only in two matters; but in one matter, what is earlier is earlier and what is later is later.

Genesis 15 is one matter on any reasonable construal. It is a single continuous encounter between God and Abram, opened by one announcement of the divine word, closed by one covenant formula, and interrupted by nothing. If Rav Pappa’s limitation is applied to it, the sequence within the chapter holds, and with the sequence, the daylight.

Two qualifications must travel with this, and they are not small. The first is that Rav Pappa’s native context is the derivation of law from the arrangement of verses, not the reconstruction of narrative chronology; extending his rule from one to the other is a move this study makes and labels as its own. The second is that “one matter” is itself interpretable, and an interpreter determined to find a night before verse 5 could argue that verses 1–6 and verses 7–21 are two matters — which is, in substance, precisely what Abarbanel’s third reading and Ibn Ezra’s reading of verse 12 both propose. The rule does not settle the question by itself. What it does is show that the appeal to non-chronological narration is not freely available: the tradition that owns that principle has itself fenced it off from cases like this one.

What the syntax establishes and what it does not

Result of the syntactic argument

The Hebrew of Genesis 15 marks no interval between the star-promise and the sunset, uses its ordinary sequential narrative form throughout, distinguishes its two sunsets by verbal form in a way that reads naturally as two moments of one evening, and closes by gathering the whole into “that day”. The syntax therefore supports the daylight reading and does not compel it. It is not possible to derive from the grammar alone a proof that no unmarked night intervened, because Hebrew narrative permits unmarked scene changes. It is possible to say, and this study does say, that anyone who posits such a night is adding to the text, and that the burden of that addition falls on him.

The vision at the head of the chapter

Genesis 15 begins by telling the reader what kind of event he is about to witness. “The word of the Lord came to Abram bammahazeh” — in the vision. The Greek and the Latin agree: en horamati, per visionem. The word appears once and is not repeated.

Everything turns on how far it reaches. If bammahazeh governs only verse 1, then verses 2 onward narrate waking events and the star-promise is an episode in the open air with a real sky above it. If it governs the whole chapter, then the bringing-out, the looking, the animals, the birds, the deep sleep, and the fire between the pieces are all things Abram saw in prophecy, and the question of what hour it was becomes a question about the hour of the vision rather than the hour of the stars.

Both readings are held by first-rank interpreters, and they contradict each other flatly. This section sets them against one another.

The maximal reading: the whole chapter is vision

The most uncompromising statement belongs to Maimonides, in the forty-sixth chapter of the second part of the Guide of the Perplexed. The chapter’s whole purpose is to establish a rule of reading: that when Scripture narrates a prophet going somewhere, doing something, or being spoken to, and the account is introduced as a vision, then the going and the doing belong to the vision and not to the world. His examples are Ezekiel digging through the wall, Jeremiah hiding the girdle by the Euphrates, Hosea taking a wife, Isaiah going naked — and Abraham. Friedländer’s public-domain translation gives it thus:

It is analogous to the description of the vision of Abraham which begins, “The word of the Lord came to Abram in a vision, saying” (Gen. 15:1); and contains at the same time the passage, “He brought him forth abroad, and said, Look now to the heaven and count the stars”… It is evident that it was in a vision that Abraham saw himself brought forth from his place looking towards the heavens and being told to count the stars.¹⁰

And then, at the end of the chapter, a sentence that is a direct answer to any argument from narrative sequence:

Whatever is said in the account of a vision, that the prophet heard, went forth, came out, said, was told, stood, sat, went up, went down, journeyed, asked, or was asked, all is part of the prophetic vision; even when there is a lengthened account, the details of which are well connected as regards the time, the persons referred to, and the place.¹¹

Maimonides is saying, in advance, that the argument of the previous section proves less than it looks. A vision narrative can be internally consistent about time — can have its own afternoon and its own sunset — without those being hours of any real day. That is a serious objection and it is answered later, not here.

Radak makes the same move at the verse itself, and adds the mechanism:

[He] had received the prophetic vision and words of G’d while in the tent. As part of this prophetic insight, Avram felt that G’d took him outside the tent, showing him the stars. This was the mahazeh, “the vision.” We have a parallel situation in Ezekiel 8,3… “He brought me to Jerusalem in visions of G’d.”¹²

The Ezekiel parallel is the strongest single argument for this reading, and it is worth stating why. Ezekiel 8:3 says that a spirit lifted the prophet up and brought him to Jerusalem in visions of God. Every reader accepts that Ezekiel did not travel; he was in Babylon throughout, with the elders of Judah sitting in front of him. If “he brought me to Jerusalem” can be a visionary transport, then “he brought him forth abroad” can be one too, and the burden shifts to whoever wants it otherwise.

Chizkuni states the same reading in five Hebrew words: kakh nir’ah lo bahalom ke’illu hotsi oto hutsah — gloss: so it appeared to him in the dream, as though He had brought him outside.¹³

The minimal reading: the vision was by day, and Abram was awake

Against all of this stands a tradition just as respectable, and it makes two distinct claims that must be kept apart: that the vision was a waking vision, and that it was a daytime one.

Nahmanides asserts the second in his opening words on the chapter, and asserts it as a promotion in Abraham’s prophetic rank:

Abraham now merited that the word of G-d should come to him in a daytime vision, for at first his prophecy came to him in nocturnal visions.¹⁴

This is the single most important datum in this study after Abarbanel’s question, and its importance is easy to miss. Nahmanides is not defending the daylight reading; he is not thinking about stars at all at this point. He is making a technical claim about grades of prophecy: mahazeh is the higher, daytime mode, and Abraham has now been raised to it from the night visions of chapters 12 and 13. But the consequence is unavoidable. If the frame declared at verse 1 is a daytime vision, then the vision reading does not remove the daylight. It relocates it. Abram is shown the stars, in the hour of daylight, in a prophetic seeing.

The waking claim is the second and separate one. Cornelius a Lapide, whose commentary on the Pentateuch was for three centuries the standard Catholic resource for exactly this kind of question, glosses per visionem at verse 1:

Per visionem: non insomnio, sed in visione qua Abram vigil vidit angelum, Deum repraesentantem, in corpore assumpto; vel oculis corporeis; vel potius oculis mentis, cum eoque foedus iniit. Ita Tostatus, Pererius, Oleaster.¹⁵

Gloss: by a vision: not in a dream, but in a vision in which Abram, awake, saw an angel representing God in an assumed body; either with bodily eyes, or rather with the eyes of the mind, and with him he entered into the covenant.

Haydock, the standard English Catholic annotator of the nineteenth century, says the same thing in one line, and — this is the point — argues it from verse 5: “Abram was not asleep, but saw a vision of exterior objects. v. 5.”¹⁶ That bare cross-reference is an argument in compressed form: he is reasoning that because verse 5 has Abram looking at external objects, the whole must have been a waking perception. Which is to say that the Catholic annotator takes the star-promise as proof of waking sight — the exact inverse of the use Maimonides makes of it.

The middle reading: two prophecies with a waking interval

There is a third position, and it deserves its own heading because it is the one that does the most work. Abarbanel offers it as the third of three ways of construing “he brought him outside”:

… bari’shonah hayah devar H’ el Avram bammahazeh … ve’ahar kakh nifseqah nevu’ato veheqits Avraham, uvihyoto biyqitsah hotsi’o misham huts le’oholo asher hayah sham hammahazeh bari’shonah, ve’az ba’at’hu nevu’ah aheret vehi vayyomer habbet na hashamayim.¹⁷

Gloss: at first the word of the Lord came to Abram in the vision … and afterward his prophecy was interrupted and Abraham awoke, and while he was awake He brought him out from there, outside his tent where the vision had first been, and then another prophecy came upon him, namely, “And he said, Look now toward the heavens.”

On this reading the chapter contains two prophetic acts with a waking gap between them, and the doubled speech verb at verse 5 marks the seam. It is a genuinely elegant proposal: it accounts for a detail of the Hebrew, it explains why Scripture bothers to mention a change of location inside a vision, and it supplies exactly the interval that the daylight problem requires. Its cost is that the interval it supplies is inferred from the doubled verb alone.

Where this leaves the frame

Result of the vision argument

The declared frame at verse 1 does not settle the hour, because the interpreters who invoke it do not agree about what it covers. Maimonides, Radak, and Chizkuni take it to cover the bringing-out, so that the stars are seen in prophecy; Nahmanides takes it to be a vision by day; a Lapide and Haydock take it to be a waking perception of external objects, with Haydock reasoning to that conclusion from verse 5 itself; Abarbanel divides the chapter into two prophecies around a waking interval. Crucially, the vision reading does not by itself dissolve the daylight claim. On Nahmanides’ construction it entails it: if the frame is a daytime vision, then whatever Abram was shown, he was shown it in the hours of daylight.

The Greek and the Latin

Textual questions about Genesis 15 are usually raised about verse 2, where the Hebrew is famously obscure, or about the four hundred years of verse 13. The verses that bear on the hour have attracted almost no attention, and they turn out to carry three divergences of real consequence. This section sets them out, states what each does to the timing, and marks the boundary of what has actually been checked.

The bird at verse 11

The Hebrew of verse 11 says that the birds of prey came down on the carcasses and Abram wayyashev them — drove them away. The Septuagint says kai sunekathisen autois Abram: and Abram sat down together with them. The difference is one of vocalization: the same consonants can be read as a causative “he drove away” or as “he sat”. Cornelius a Lapide identifies the mechanism exactly, and, characteristically, refuses to choose:

Verum Septuag. aliis punctis legentes … contrarie vertunt, sedit cum eis Abram, quod tamen verum quoque est: sedebat enim eminus Abram cum avibus a se abactis.¹⁸

Gloss: but the Seventy, reading with other points, translate the opposite — “Abram sat with them” — which is nevertheless also true, for Abram was sitting at a distance from the birds he had driven off.

Either way, verse 11 describes duration. Whether Abram is repeatedly shooing carrion birds off a row of split animals or sitting on watch beside them, the verse asks the reader to imagine a stretch of time passing. It is the only verse in the chapter that does. This will matter twice: once in support of the intervening-day reading, which needs the interval to be long, and once against it, since a vigil of some hours is not the same as a vigil of a day.

The trance at verse 12

The Hebrew tardemah is the word used of the sleep God causes to fall on Adam at Genesis 2:21. The Septuagint renders it ekstasis in both places. The Vulgate renders it sopor in both places. The Greek choice is not neutral: ekstasis is a displacement of the mind, and a reader whose Genesis says that an ecstasy fell on Abram at sunset has been told, in the text itself, that what follows is visionary.

Philo builds an entire section of his treatise on this chapter out of the word, distinguishing four kinds of ekstasis and assigning Abram’s to the fourth — prophetic possession. He then does something that determines the whole ancient reception of the chapter’s clock. He allegorizes the sunset:

And under the symbol of the sun he intimates our mind… As long therefore as our mind still shines around and hovers around, pouring as it were a noontide light into the whole soul, we, being masters of ourselves, are not possessed by any extraneous influence; but when it approaches its setting, then, as is natural, a trance, which proceeds from inspiration, takes violent hold of us… for when the divine light shines the human light sets.¹⁹

Once the sunset of verse 12 has been read as the setting of the human intellect, it has stopped being a time of day. The question “was it dark yet?” cannot even be formulated. Augustine will do the same thing with different content, reading the horror of great darkness as the tribulation at the end of the world and the fire between the pieces as the day of judgment. That is the mechanism by which the difficulty stayed invisible for a millennium: the chapter’s clock was not read as a clock.

The darkness that is not there at verse 17

This is the substantial divergence. The Hebrew of verse 17 has two clauses: the sun had gone in, and there was thick darkness (wa-alatah hayah). The Greek of the codex Swete prints has: epei de egineto ho helios pros dusmais, phlox egeneto — when the sun was going toward its setting, a flame arose. There is no equivalent of alatah, and the verb of setting is an imperfect, not a completed action.

Two consequences follow.

The first is that the Greek Genesis 15 never reaches night. Its two temporal markers are “about the sun’s setting” and “when the sun was going toward its setting”. Both describe the same late-afternoon condition. A reader of the Greek has no dark to set against the stars of verse 5, because the chapter he is reading does not get that far.

The second is that this makes the daylight difficulty worse in Greek, not better — or would, if anyone had been counting. In the Hebrew, verse 5 is before a sunset. In the Greek, verse 5 is before a sunset that never completes, inside a scene entirely bounded by daylight. The Greek chapter is, on its face, a chapter that happens in the afternoon and contains a command to count stars.

Augustine’s Old Latin confirms the reading from the other side. He quotes verse 17 as Cum autem iam sol erat ad occasum, flamma facta est, et ecce fornax fumabunda et lampades ignis — “and when the sun was now toward setting, a flame arose, and behold a smoking furnace and lamps of fire”. Flame, no darkness, sun still going down. His Vorlage had what Swete prints.

“There” or “in that day” at verse 18

The Hebrew and the Vulgate open verse 18 with a temporal formula: bayyom hahu, in illo die. Swete’s Greek opens it with ekei — there. A marker of place stands where the other witnesses have a marker of time.

If ekei were the Septuagint reading simply, it would be a third de-temporalization and the pattern would be complete. It is not. Augustine’s Old Latin reads In die illa disposuit Dominus Deus testamentum ad Abram, which shows that the Greek behind his version had the temporal phrase. The correct statement is therefore that the Greek transmission is divided here, and that the codex Swete prints belongs to the branch that removes the day.

This is a real limit on the argument and it is stated rather than smoothed. What survives it is still substantial: in every Greek witness reached by this study, verse 17 lacks the darkness; and in the principal Greek codex, verse 18 lacks the day. The de-temporalizing tendency is real even where its extent is uncertain.

Where the Latin stands

Jerome’s Latin restores the Hebrew at both points — caligo tenebrosa at verse 17, in illo die at verse 18 — and thereby restores the difficulty in full force to the Western church from the end of the fourth century onward. From Jerome forward, a Latin reader has a chapter whose stars are commanded at verse 5, whose sun is going down at verse 12, whose darkness falls at verse 17, and which then says the whole thing happened in one day.

It is worth pausing on how completely that opportunity was declined. The Vulgate’s own scholarly apparatus, in the Hetzenauer printing this study read, carries a foot-note reference at 15:5 — and it points to Romans 4:18. The verse is cross-referenced for its use by Paul. Nobody, in that apparatus or in the marginal summaries beside it, has anything to say about the sky.

Two divergences that do not bear

For completeness, and to keep the argument from claiming more than it should: the Septuagint at verse 2 differs materially from the Hebrew (it has ho huios Masek tes oikogenous mou, the son of Masek my home-born servant, where the Hebrew has the obscure ben-mesheq beti), and at verse 13 it adds a clause about dwelling in Egypt and in Canaan which Haydock’s annotator explicitly rejects.²⁰ Neither touches the hour, and neither is used here.

The count that cannot be completed

Before the hour can be argued about, it is worth being clear what kind of act is being commanded, because the interpreters who deny the daylight and the interpreters who assert it agree about very little else and disagree least about this.

The command at Genesis 15:5 has three parts, and only the first is a command to look. Habbet-na hashamaymah — look heavenward. Usfor hakokhavim — and count the stars. Im tukhal lispor otam — if you are able to count them. The third part is a real conditional, and it is a conditional whose failure is presupposed by the promise that follows. Abram is not being shown a large number. He is being set a task, and the task is designed not to be completable.

Genesis 13:16 has already used the same device with a different material: “if any man be able to number the dust of the earth, he shall be able to number thy seed also.” The two verses are a pair. The second escalates the first, and the escalation runs from a material too fine to count to a material too distant. What the tradition has to say about the second half of that escalation turns out to bear on the whole question of the hour.

Whether the stars are numerable at all

Philo, at the point where he quotes the verse, treats the impossibility of the count as its meaning, and then moves the impossibility inward: “if it is impossible to find out the number of the stars which are perceptible by the outward senses, how much more so must it be to count those which are discernible only by the intellect?” The visible stars stand for a larger invisible set, and the failure of the count is the argument.

Augustine goes further and turns the verse into a weapon. In The City of God XVI.23 he denies that all the stars can be seen at all — nec omnes eas uideri posse credendum est — gives an empirical reason (a sharper eye sees more of them), adds that some rise and set in a part of the world remote from us, and then names two men who had claimed to have finished the count:

Postremo quicumque uniuersum stellarum numerum conprehendisse et conscripsisse iactantur, sicut Aratus uel Eudoxus uel si qui alii sunt, eos libri huius contemnit auctoritas.

Gloss: Finally, whoever boast that they have comprehended and written down the whole number of the stars — Aratus, say, or Eudoxus, or any others there may be — the authority of this book despises them. Aratus of Soli wrote the Phaenomena; Eudoxus of Cnidus was the astronomer his poem versified. Augustine is using Genesis 15:5 as a text against Hellenistic star catalogues.

Cornelius a Lapide, fifteen centuries later, has the number. Some, following Ptolemy and the astrologers, count 1,022 stars — but, he says, they are counting only the ones conspicuous to sight, the bright and more notable ones. And then, in a sentence that would have surprised Augustine:

Dioptra vero plures stellas ostendit, quae solis oculis cerni nequeunt.²¹

Gloss: but the dioptra shows more stars, which cannot be discerned by the eyes alone.

Malbim, in the nineteenth century, closes the sequence: where the ancients counted 1,022, “the later ones, who invented polished mirrors” have established that the number is beyond reckoning, and God gave into Abraham’s power of sight that he should see the whole multitude “floating in the expanses of distances that have no end”.

Why this matters to the hour

Three things follow, and the third is the one that bears.

The tradition has always understood the command as one that fails. From Philo to Malbim, nobody reads “count the stars” as an invitation to a successful enumeration. The conditional is read as concessive: try, and you will not manage it. That is agreed.

The tradition has always understood that the visible stars are not all the stars. Augustine says the keenest eye is defeated; a Lapide says the naked eye reaches about a thousand and the telescope more; Malbim says the true number is endless and Abraham’s sight had to be enlarged to reach it. Every one of them concedes that the sign as given exceeds what an unaided man standing outdoors could take in.

And therefore the concession that a daylight reading requires is a concession the tradition has already made in another form. The daylight reading says: Abram was told to count stars he could not see, because they were not visible at that hour. Malbim says: Abram was told to count stars he could not see, because they were beyond the reach of his eye, and his eye had to be supernaturally extended. Augustine and a Lapide say: whatever Abram saw, it was not all of them.

This does not prove the daylight reading. It removes one objection to it. The objection would be that a command to look at what cannot be seen is absurd and therefore cannot be what the verse means. But a command to count what cannot be counted is already what the verse means on every reading, and the tradition has never found that absurd. Whether the uncountability is a matter of number or of hour changes the physics and not the logic.

An aside on Abram the astronomer

There is a related tradition, old and remarkably widespread, which holds that Abraham was himself an expert in the science of the stars. It is worth recording because it is the background against which both the astrological reading and the counting command take their sense.

Josephus has it in the Antiquities: Abram, in Egypt, “communicated to them Arithmetick; and delivered to them the science of Astronomy. For before Abram came into Egypt they were unacquainted with those parts of learning: for that science came from the Chaldeans into Egypt; and from thence to the Greeks also.”²² A Lapide reports it in a stronger form, noting that God bids Abram number the stars “both because he was an astronomer, and because he was accustomed to look up at them”, and cites verses attributed to Orpheus in the fifth book of Clement’s Stromateis in which Abraham is called an astrologer.²³ And the Talmudic passage at Shabbat 156a presupposes the same thing: Abram’s objection is that he has consulted his own astrological chart.

The tradition is unanimous, then, that the man being told to count the stars was a man who knew what a star catalogue was — and who, on the rabbinic reading, had been using one to work out his own childlessness. This is the thing the promise is answering. It does not settle the hour. But it makes the second half of the command — count them — a good deal sharper than a figure of speech, because it is addressed to somebody whose profession, in the received picture, was counting them.

The problem stated in 1500: Abarbanel’s seventh question

Don Isaac Abarbanel — statesman to Portugal, Castile, and Naples; exiled with the Jews of Spain in 1492; commentator on the whole of Scripture — prefaces each unit of his Torah commentary with a numbered list of the difficulties he proposes to solve. On Genesis 15 he lists twenty-one. He introduces them with a sentence that has the ring of a man who has been at this a long time: ve’ani be’iyuni bah yamim rabbim ra’iti lish’ol bah she’elot — gloss: and I, examining it for many days, saw fit to ask questions of it; the number, twenty-one, stands in the printed text as a Hebrew numeral.

The seventh is this.

The text

Abarbanel, Perush al ha-Torah, on Genesis 15, question 7

Ha-she’elah ha-zayin: omro “vayyotse oto hahutsah”. Ve’im hayah kol zeh bemar’eh hannevu’ah keda’at hamefareshim, meh hayah hatsorekh lomar shennir’ah lo shehotsi’o meha’ohel lahuts lomar lo “habbet na hashamayim”, af ki lo zakhar hakkatuv shehayah ba’ohel, velo hayah me’inyan hammar’eh lehodia sheyyatsa mimmaqom ehad el maqom aher, vehayah dai sheyyomar “vayyomer lo habbet na hashamaymah”, ki ulai basadeh ba’ah lo hannevu’ah velo ba’ohel. Ve’ulai shemippenei zeh amru zal (Shabbat 157) hotsi’o me’itstagninut shello.

Ve’im amarnu shehayah zeh behaqits, mi yitten veyada’ti im hayah bayyom o vallaylah; ki i-efshar sheyyihyeh bayyom kemo shennemar “usfor hakokhavim”, ki einam nir’im bayyom; ve’im hayah ballaylah, yiqsheh omro “vayhi hashemesh lavo”.²⁴

Gloss of the second paragraph, which is the part that matters:

And if we say that this was in the waking state, would that I knew whether it was by day or by night; for it is impossible that it was by day, as it says “and count the stars”, since they are not visible by day; and if it was by night, then His saying “and the sun was about to set” is difficult.

What has just happened

The maintainer’s claim, which prompted this study, was that Genesis 15:5 was spoken in daylight, on the ground that the chapter’s own sequence puts the star-promise before its own sunset. That claim is not a modern misreading of a chapter nobody has looked at closely. It is the second horn of a dilemma posed, in writing, five hundred and some years ago, by one of the most systematic readers the chapter has ever had — and posed in the same two moves: verse 5 requires visible stars, verse 12 forbids the night.

Three features of Abarbanel’s formulation deserve attention because they show how carefully he has seen it.

He conditions the dilemma on the waking reading. Ve’im amarnu shehayah zeh behaqits — if we say this was while awake. He is perfectly aware that the whole difficulty is generated by treating the episode as an event in the world rather than an event in prophecy, and he says so. This is the correct analysis and it is the analysis this study has reached independently in the two preceding sections.

He treats the daylight as impossible, not implausible. I-efshar is a strong word. His reason is astronomical and is stated plainly: ki einam nir’im bayyom, because they are not visible by day. It is important to be exact about what this shows. It shows that Abarbanel grants the maintainer’s textual observation completely — that on a plain sequential waking reading verse 5 lands in daylight — and rejects the conclusion on a premise that is not textual at all. The premise is about the sky, not about the sentence.

He does not resolve it by appeal to non-chronological narration. The rabbinic principle ein muqdam ume’uhar was available to him and he does not reach for it here. He resolves the difficulty structurally instead, by dividing the chapter — which is the third of the three constructions quoted in the preceding section, with the waking interval between two prophecies.

The sixteenth question, and the Ran

Abarbanel’s sixteenth question on the chapter is the natural companion of the seventh, and it asks something nobody else in this study asks: why is the hour mentioned at all?

Be’omro “vayhi hashemesh lavo vetardemah naflah” … vehinneh lo matsanu bekhol nevu’ot Avraham sheyyizkor hakkatuv hazzeman veha’et mehayyom asher hal alav hashefa, velo hayah bahagbalah hazzot tsorekh kelal.²⁵

Gloss: And behold, we have not found in any of Abraham’s prophecies that Scripture records the time and hour of the day on which the influx came upon him, and there was no need for this delimitation at all.

This is a sharp observation and it cuts in the claim’s favour. Genesis is generally indifferent to the hour of a divine communication; it does not tell us when the word came to Abram in chapter 12, or 13, or 17, or 18. Genesis 15 tells us twice, in twelve verses, with a change of tense between them, and then adds “in that day” for good measure. Whatever else is going on, the chapter is insisting on its clock. An interpreter who then supplies an unmarked night in the middle of it is working against the grain of the one chapter that cares.

One thing Abarbanel does not consider

For all its acuteness, the seventh question has a gap, and naming it is part of testing the claim rather than defending it.

Abarbanel poses a two-horned dilemma — day or night — and treats the horns as exhaustive for the waking case. But there is a third possibility he does not raise: that it was day, that the stars were not visible, and that this is the point. On that reading the impossibility he identifies is not a reason to reject the daylight but the content of the sign. Abram is told to number what he cannot see, and told that his seed will be so, and the next verse says he believed.

Abarbanel cannot take that road, and it is worth being clear about why, because the reason is not carelessness. His whole method is to find a construction on which the text says something coherent and complete; a command whose execution is impossible is, for him, a defect to be removed rather than a datum to be interpreted. Whether he was right about that is a question this study returns to in its synthesis, where the reading is offered as project synthesis and labelled as such.

Five answers, kept apart

The Jewish exegetical tradition has more than one answer to what happens at Genesis 15:5, and the answers are not variants of one another. They rest on different premises, they were proposed for different reasons, and at least two pairs of them are incompatible. Harmonizing them would destroy the most interesting fact about the reception, which is that a difficulty this small produced disagreement this wide.

What follows sets out five, each at its own locus.

The five

The vision answer

Treated at length in the preceding section on the frame. Its representative formulation is Radak’s — the prophecy came in the tent and Abram felt himself brought outside — with the Ezekiel 8:3 parallel as its warrant. Its strength is that it is a general principle of prophetic narrative rather than an expedient invented for this passage. Its weakness, for present purposes, is that it does not by itself answer the question asked: a vision has an hour too, and Nahmanides says that this one was by day.

The intervening-day answer

This is the most economical answer available and it deserves its own analysis, because it is the strongest thing standing against the claim.

Ibn Ezra states it as an inference from verse 12 rather than as a solution to anything. His comment there, in the public-domain Piotrkow Hebrew, reads: zeh lanu le’ot ki “vayyiqqah lo et kol elleh” beyom … shehiqits mimmahazeh hannevu’ah.²⁶ Gloss: this is a sign to us that “and he took him all these” [verse 10] was by day, when he had awoken from the vision of prophecy.

So Ibn Ezra’s Genesis 15 runs: a prophetic vision, containing the star-promise; Abram wakes; a day of work fetching, killing, dividing, and guarding the animals; sunset at verse 12; the second prophecy; darkness at verse 17. He confirms the shape at verse 17, where he notes that the sun had not yet set at verse 12 and now has.

Cornelius a Lapide reaches the same structure from the Latin side and, so far as this study could determine, without depending on Ibn Ezra. His note on verse 12:

Sopor hic Abrae partim fuit naturalis, ex nimio labore diurno, occidendi, dividendi, sacrificandi victimas, abigendique ab eis volucres; partim fuit Abrae a Deo immissus, sicut Adae soporem immisit Genes. 2, 21.²⁷

Gloss: This deep sleep of Abram’s was partly natural, from excessive daytime labour — killing, dividing, sacrificing the victims and driving the birds off them; and partly sent upon Abram by God, as He sent a deep sleep upon Adam.

Nimio labore diurno: excessive daytime labour. A Lapide has, in one clause, supplied the entire missing day. And — this is the striking part — he has done it while asserting, at verse 5, that the star-promise took place at night, without ever noticing that he has thereby committed himself to a two-day chronology for a chapter that says “in that day”.

The astrological answer

Treated in full in the next section, because the maintainer’s question about it was specific and the answer requires its own evidence.

The above-the-firmament answer

Genesis Rabbah 44:12 preserves three readings of “he brought him outside” side by side. The second is R. Yehudah’s, in the name of R. Yohanan:

He took him up above the dome of the heavens. That is what He said to him: “Look [habet] now to the heavens” — [the expression] habatah refers only to looking from above to below.²⁸

Rashi carries this into his own commentary as a third reading, after the plain sense and the astrological one, and gives the same philological argument.²⁹ Rabbeinu Bahya makes the argument explicit as an argument: ein leshon habbatah ella millema’lah lemattah — and therefore, he reasons, if Abram had been standing on the earth looking up at the sky, habbet would have been the wrong word.³⁰

It deserves saying that this answer, alone among the five, would actually dissolve the daylight problem in the terms in which it arises: above the vault of the sky there is no atmosphere to scatter sunlight, and the stars are there to be seen at any hour. Whether that consequence occurred to R. Yehudah is not knowable from the text and is not claimed. What can be said is that the reading is not offered as an answer to the hour, and none of the sources that transmit it connects it to verse 12.

Against it stands a fact from the versions. Both ancient translators read habbet as an upward look: the Septuagint’s anablepson and the Vulgate’s suspice are both unambiguously “look up”. Whatever the philological merits of ein habbatah ella millema’lah lemattah, the two oldest translations of the verse did not see it.

The enlarged-sight answer

Malbim, writing in the nineteenth century, supplies the most physical answer of the five. After giving the astrological reading he continues that God “gave into the power of Abraham’s sight that he should look upon all the multitude of stars floating in the expanses of distances that have no end, and said to him, ‘if you are able to count them’ — as I am showing them to you now.” He notes, in passing, that where the ancients counted 1,022 stars, “the later ones, who invented polished mirrors”, established that the number is vastly greater.³¹

The answer is candid about what it concedes: ordinary human sight would not have sufficed for the command as given. Malbim’s reason for saying so is the number of the stars, not the hour — but the concession is the same concession, and it sits oddly beside the confident assertions elsewhere in the tradition that Abram simply stepped outside and looked.

What the disagreement shows

Result of the reception survey

No consensus exists. The five answers are mutually inconsistent at several points: the vision answer and the waking-interval answer cannot both be right about the same verse; the above-the-firmament answer contradicts both ancient versions on the meaning of habbet; the enlarged-sight answer concedes what the plain-sense answer denies. What every one of them shares is a refusal to accept that Abram was told to look at an empty sky. That refusal, and not any reading of the chapter’s sequence, is what unites the tradition against the daylight.

“He brought him forth abroad”: out of what?

There is a reading of Genesis 15:5, well known in Jewish circles and often repeated in Christian ones without a source, according to which “he brought him outside” does not mean that Abram left his tent but that God brought him out from under the government of the stars — out of astrology, out of the determinism of the horoscope. It is a striking reading, and it has the further interest that if it is right, it explains why the promise is sealed by stars in particular: the sign is taken from the very thing whose power over Abram is being cancelled.

The question put to this study was whether that reading is attested, and by whom. It is. It is attested earlier and more widely than the usual attributions suggest, and the evidence is worth laying out carefully, because the popular form of the claim gets the provenance wrong.

The Talmudic locus

The classical statement is at Shabbat 156a, in a sugya about whether the constellations govern Israel. Rabbi Yohanan holds that they do not, and derives it from Jeremiah 10:2. Then Rav is brought in with a second derivation:

Amar Rav Yehudah amar Rav: minnayin she’ein mazzal leyisrael? Shenne’emar “vayyotse oto hahutsah” … Amar lefanav: ribbono shel olam, nistakkalti be’itstagninut shelli, ve’eini ra’ui lehold ben. Amar lo: tse me’itstagninut shellekha, she’ein mazzal leyisrael.³²

Gloss: Rav Yehudah said in the name of Rav: whence that Israel has no constellation? Because it is said, “And He brought him outside”… He said before Him: Master of the world, I have looked in my astrology and I am not fit to beget a son. He said to him: go out from your astrology, for Israel has no constellation.

The pun is on hahutsah. Abram is brought out — and the midrash supplies the object: out of the discipline he had been consulting.

The midrashic loci

Genesis Rabbah 44:12 preserves the reading in a different form and beside two competitors. The passage opens by registering that the word “outside” is odd:

“He took him outside” — Rabbi Yehoshua of Sikhnin in the name of Rabbi Levi: Did He take him outside the world, that the verse should say: “He took him outside”? Rather, he showed him the streets of the heavens… The Rabbis say: [God told him:] ‘You are a prophet, not an astrologer’… your forefather Abraham sought to adopt that mindset, but I did not allow him to do so.³³

Two sections earlier, at Genesis Rabbah 44:10, the same idea is attached to Abram’s complaint rather than to God’s answer: Rav Shmuel bar Rav Yitzhak has Abram say hammazzal dohaqeni ve’omer li: Avram, ein attah molid — gloss: my constellation presses me and says to me: Abram, you will not beget.

Rashi, and the shape the reading took

Rashi’s note on the verse is the channel through which the reading reached almost every later reader, Jewish and Christian. He gives three senses in eighteen words of Hebrew, and their order is itself a statement:

And he brought him forth outside — Its real meaning is: He brought him outside his tent so that he could look at the stars. Its Midrashic explanation is: Go forth from (give up) your astrological speculations — that you have seen by the planets that you will not raise a son; Abram indeed may have no son but Abraham will have a son… I will give you other names, and your destiny will be changed. Another explanation: He brought him forth from the terrestrial sphere, elevating him above the stars.³⁴

The plain sense comes first, and the plain sense is the domestic one: out of the tent, to see the stars. Rashi does not replace the tent with the horoscope. He lays the astrological reading beside it as midrasho, its midrashic sense. That distinction is regularly lost in the popular retelling, where the astrological reading is presented as what the verse means rather than as one of three things it has been taken to mean.

Nahmanides quotes Rashi’s midrashic sense in full and then presses on it: but Abram did father Ishmael, so what can it mean that “Abram will have no son”? His answer is that the astrological reading must be about an heir, not a child.³⁵ That is a real objection, and it shows the reading being handled critically rather than repeated.

Radak transmits it with an attribution to Genesis Rabbah. Rabbeinu Bahya combines it with the above-the-firmament reading and draws the moral: ume’khan hokhihu z”l ein mazzal leyisrael ki hem lema’lah mehem — gloss: and from here the sages proved that Israel has no constellation, for they are above them. Malbim, six centuries later, states it in the language of law and government: God revealed to Abram that he and his seed are not placed under mishterei hakokhavim vehammazzalot, the regimes of the stars and constellations.

The reading is older than the rabbis: Philo

Here the usual account of the reading’s provenance needs correction.

Philo of Alexandria wrote an extended treatise on this very chapter, Quis rerum divinarum heres sit. He comments on “he led him out” at verse 5 — and reads it, in the first century, as a bringing-out from something other than a tent:

Very correctly, therefore, it is said, he led him out (exegagen exo) of the prison according to the body, of the caves existing in the external senses, of the sophistries displayed in deceitful speech; and beyond all this, out of himself and out of the idea that by his own self-exerted, self-implanted, and independent power he was able to conceive and comprehend.³⁶

That is not yet astrology. But two chapters later, commenting on verse 7 — “I am the Lord who brought thee out from Ur of the Chaldees” — Philo makes the astrological identification explicitly, and attaches it to the other bringing-out in the same chapter:

[T]he good which was previously bestowed upon him was the departure from the Chaldaean philosophy, which was occupied about the things of the air, which taught men to suppose that the world was not the work of God, but was God himself; and that good and evil is caused in the case of all existing things, by the motions and fixed periodical revolutions of the stars… the soul departs from astronomy and learns to apply itself to natural philosophy… to quit the creature for the Creator.³⁷

This is a genuinely important result for the question as it was put. The reading that Abram’s “bringing out” is a liberation from astral determinism is not a medieval rabbinic invention. It is present in Greek-speaking Judaism in the first century, where it is fastened to Genesis 15:7 and the departure from Ur. What the Talmud does at Shabbat 156a is transfer it two verses backward, from the departure out of Chaldea to the going out of the tent, and thereby weld it to the stars that are the subject of the promise. The transfer is what makes the reading beautiful; but the idea it transfers was already there.

What can and cannot be said about it

Result on the “out from under the stars” reading

Attested: yes, and in three distinct layers. In Philo (first century) as the soul’s departure from Chaldean star-fatalism, attached to Genesis 15:7. In the Babylonian Talmud (Shabbat 156a) and Genesis Rabbah 44:10 and 44:12, attached to hahutsah at 15:5. In the medieval commentaries — Rashi, Ramban, Radak, Rabbeinu Bahya, Malbim — as a transmitted midrashic sense.

Not established: that it is the plain sense of the verse. Rashi, who is the reason most readers have heard of it, ranks it second, behind the tent; so does Nahmanides, in the same words. Both ancient versions render the going-out spatially (exo, foras) and the looking as an upward look (anablepson, suspice), which is what one says of a man outdoors, not of a man delivered from a discipline.

Bearing on the hour: indirect. The astrological reading neither requires nor forbids daylight, because it does not treat the going-out as a change of place at all. But it does supply the missing motive for the sign. If the promise is being sealed against the horoscope, then the stars are named because they are the opposing authority, not because they were conveniently overhead — and the question of whether they were visible stops being the question the verse is about.

What the checked Christian witnesses do

The earlier edition grouped several Christian witnesses under a broad negative. Exact-locus inspection overturns that result. The witnesses do not form one answer, and their differences must be retained.

Augustine: a bounded non-treatment

Augustine discusses the impossibility of numbering the stars and allegorizes the sunset and darkness in the covenant vision. At City of God XVI.23–24 he does not ask whether the stars could be seen at the hour of the command. That is a negative result at two identified chapters, not a statement about Augustine’s whole corpus or Christian tradition.

Ambrose: visionary movement

Ambrose, De Abraham II.8.48, treats Abram’s being led outside as a visionary movement beyond bodily confines. He does not settle a waking hour. His reading instead supplies a concrete ancient instance of the vision answer and prevents the study from treating Christian reception as silent.

Chrysostom and Ephrem: an interval

John Chrysostom, Homilies on Genesis 36.13 and 37.3–6, places the initial divine appearance by night and the sacrifice through the following day to sunset. Ephrem’s Commentary on Genesis XII.2–5 is more explicit: the words come at night, the commanded work occupies the day from dawn to evening, and the later manifestation occurs that evening. Both witnesses resolve the sequence by supplying elapsed time that Genesis does not narrate. Their readings answer the difficulty without becoming a single patristic rule.

Theodoret and the remaining limit

Theodoret’s Questions on Genesis 66–67 does not address the hour. That is only a local negative. Jerome’s authenticated Hebrew Questions on Genesis remains unchecked, as do wider portions of the authors named here. No author-wide or tradition-wide silence follows from this bounded corpus.

Philo dissolves the clock

Philo’s treatise on this chapter reads the going-out as the soul’s exit from the body and the senses, the setting sun as the setting of the human intellect, and the trance as prophetic possession. Immediately after the command to count the stars he considers their uncountability, not the hour at which they could be seen.³⁸ His allegorical treatment remains useful evidence for one reception of the temporal markers. It no longer supplies a cause for an alleged patristic silence.

Josephus removes the evidence

Josephus’ identified retelling retains the posterity-like-stars comparison and the commanded sacrifice while omitting the command to go outside and look and all the temporal markers that generate the question.³⁹ This is a finding about that retelling’s architecture, not evidence that the question was absent from Christian tradition.

The Targums keep the sequence

Targum Onkelos renders the three controlling markers closely. Targum Pseudo-Jonathan expands the dread and darkness at verses 12 and 17 but inserts no night before verse 5, morning after verse 6, or explicit interval between the promise and the rite.⁴⁰ The negative is bounded to those inspected loci. It does not prove that every targumic witness preserves the same sequence.

Later Catholic commentators

Cornelius a Lapide explicitly infers night from the command at verse 5: erat ergo tunc nox. At verse 12 he explains Abram’s deep sleep by daytime labour. Read together, the notes imply a night, a following day of work, and a later sunset, although he does not spell out that chronology or reconcile it with “in that day.”⁴¹

The 1609 Douay apparatus has no note at verse 5. Haydock likewise supplies no note there after using the verse at 15:1 as evidence that Abram was awake. These are local publication-apparatus negatives. They remain worth reporting without making them stand for Catholic reception as a whole.

What the versional evidence cannot explain

Swete’s identified Greek printing weakens the chapter’s closing temporal chain, but Augustine’s Old Latin retains “in that day.” Chrysostom supplies one identified Greek Christian reception, while Ambrose and Ephrem supply different Latin and Syriac receptions. These witnesses do not permit the claim that one Greek form explains patristic silence. They permit only the more modest finding that several Christian readers resolved or reframed the sequence in different ways.

Where the promise recurs

The comparison of Abraham’s seed to the stars is one of the most durable formulas in the Hebrew Bible. It returns at least eight times after Genesis 15, in the Pentateuch, the historical books, and the New Testament. If the daylight problem were a feature of the promise rather than of one scene, it should leave traces in the recurrences. This section checks all of them.

The result is clean and slightly surprising: not one recurrence reproduces the problem, and the reason is the same in every case.

The recurrences

All are quoted from the registered Douay–Rheims and cited in its Vulgate numbering. The reader should note in particular that the book called 2 Esdras here is Nehemiah in a modern Bible.

What the table shows

The right-hand column is the whole finding. Genesis 15:5 is the only place in the entire tradition where a person is commanded to perform an act of sight. Everywhere else the stars appear inside a speech, as a term of comparison, and a term of comparison has no hour. “I will multiply thy seed as the stars of heaven” is as true at noon as at midnight, and nobody hearing it is being asked to check.

That is why the daylight problem neither survives nor is dissolved by the recurrences: it is not the sort of thing that could travel with them. It is a property of a scene, not of a formula, and the scene occurs once.

Three further observations follow from the table and are worth keeping.

The formula begins with dust, not stars. Genesis 13:16 has already promised a seed as the dust of the earth, with a conditional of exactly the same shape — “if any man be able to number the dust”. The move from dust to stars at 15:5 is therefore an escalation within a series, and both members of the series work by proposing a count that cannot be completed. Radak notices the progression and takes both as hyperbole; the Tur reads the two as pointing to Esau and Jacob respectively, the one trodden underfoot like dust and the other untouchable like a star.⁴² Neither construal has anything to do with the hour.

The one recurrence that is a demonstration is a fulfilment, not a promise. Deuteronomy 1:10 says “you are this day as the stars of heaven, for multitude” — and Moses is pointing at a crowd standing in front of him, in daylight, in the wilderness. The comparison has become checkable, and what is checked is the people, not the sky. The direction of the sign has reversed.

The combination with sand is later than Genesis 15. Genesis 22:17 pairs stars with the sand of the seashore for the first time, and it is that combined formula, not the bare star-promise, that Hebrews 11:12 quotes. The letter to the Hebrews is therefore drawing on the language of the Aqedah rather than on the scene in chapter 15, which is a small point against reading Hebrews 11:12 as a commentary on Genesis 15:5.

One recurrence that repays a second look

Exodus 32:13 is the only recurrence spoken by a man to God, and Moses uses it as a legal instrument: at the golden calf, he reminds God of the oath sworn “by thy own self” and asks that the sentence be revoked. The Vulgate’s own apparatus in the Hetzenauer printing gathers the covenant of Genesis 15:18 to a chain of later passages — Genesis 12:7, 13:14, 26:4, Deuteronomy 34:4, 3 Kings 4:21, and 2 Paralipomenon 9:26 — but not to Exodus 32:13.⁴³ The list is a list of land promises. It is a small sign of how the chapter was remembered: Genesis 15 entered the Latin cross-reference tradition as the land covenant, with its star-promise carried under verse 5 and pointed not at any other star text but at Romans 4:18.

Paul, and whether the hour bears on him

Genesis 15:6 is one of the two or three most consequential sentences in the Christian doctrine of justification. Paul quotes it at Romans 4:3 and Galatians 3:6, and quotes the oracle of 15:5 at Romans 4:18. The question put to this study was whether the timing of the star-promise bears on Paul’s use of it.

The answer is that it does not, and this section says so plainly and then says exactly how far the “not” extends, because there is one respect in which the question is not quite closed.

What Paul actually quotes

Two facts follow immediately from the middle column.

The first is that Paul quotes the oracle of verse 5 — “so shall thy seed be” — and never the command that precedes it. He does not quote “look up to heaven”, he does not quote “number the stars”, and he does not mention the bringing-out. The sky is not in his argument at all.

The second is that verse 6, which carries all the doctrinal weight in both epistles, contains no temporal or visual element whatsoever. It says that Abraham believed and that it was reckoned to him. Whether the sun was up when he believed is not a variable in any inference Paul draws.

The Latin tradition confirms this by its own habits. Hetzenauer’s apparatus at Genesis 15 attaches Romans 4:18 to verse 5 and Romans 4:3, Galatians 3:6, and James 2:23 to verse 6 — and nothing else. What the Latin church took from the passage was the faith and the reckoning, not the scene.

The one place where the question is not quite closed

Romans 4:18–21 does not merely report that Abraham believed. It characterizes the believing, and it characterizes it by contrast with appearances:

Who against hope believed in hope; that he might be made the father of many nations, according to that which was said to him: So shall thy seed be. And he was not weak in faith. Neither did he consider his own body, now dead (whereas he was almost an hundred years old), nor the dead womb of Sara. In the promise also of God he staggered not by distrust: but was strengthened in faith, giving glory to God.⁴⁴

The logic of the passage is that Abraham’s faith was exercised against the evidence of his senses: a body as good as dead, a womb as good as dead. Paul does not add a third item to that list. But it is at least legible that a scene in which the promise is sealed by a sign that cannot be inspected would have the same shape as the faith Paul describes — and that a scene in which the sign is plainly overhead would have a different one.

This study declines to argue from that. Paul does not make the argument, the Fathers who read Paul on this text do not make it, and constructing it would be reading a modern textual observation back into an apostle. It is recorded here only to mark the boundary of the negative result: the timing has no bearing on Paul’s inference, and this study makes no claim that it has any bearing on his rhetoric either.

The one thing the timing does bear on, and it is not Paul

There is a Catholic doctrinal question in the neighbourhood which the timing does touch, and it is worth naming so that a reader does not go looking for it in the wrong place. Augustine’s use of Genesis 15:6 in The City of God XVI.23 is directed against a specific opponent: “lest the circumcision should glory, and be unwilling to receive the uncircumcised nations to the faith of Christ. For at the time when he believed, and his faith was counted to him for righteousness, Abraham had not yet been circumcised.” The argument is chronological — it turns entirely on when verse 6 stands relative to Genesis 17 — and Augustine takes it straight from Romans 4:10.

So the tradition is perfectly capable of arguing from the chronology of Genesis 15. It argues from the chapter’s position relative to the rest of Abraham’s life, because Paul taught it to. It does not argue from the chapter’s position relative to its own sunset, because nobody taught it to, and because, as the section on the versions showed, the sunset in the text Augustine read was not a sunset at all.

Result on Paul

The hour of Genesis 15:5 has no bearing on Paul’s use of Genesis 15:6 in Romans 4 or Galatians 3. Paul quotes the oracle and the reckoning and never the command to look. The one chronological argument the New Testament does make from this chapter — that Abraham was justified before he was circumcised — concerns the chapter’s place in Abraham’s life, not the place of verse 5 within the chapter. Any attempt to make the daylight bear on justification would be this study’s construction and not the apostle’s, and it is not offered.

Project synthesis: what hour it was

Everything to this point has been evidence. What follows is the judgment this study reaches, and it is labelled as project synthesis throughout: it is an editorial conclusion drawn from checked sources, not the teaching of any of them. Where a claim in this section belongs to a source, the source is named; where it does not, it is this study’s.

The judgment

Project synthesis: the judgment reached

The claim survives, in a form narrower than it was put and stronger than it might have been.

Genesis 15, read as one continuous scene in the order it narrates, places the command to number the stars before its own sunset. That is not an inference; it is the arrangement of the chapter, and every witness consulted here — Hebrew, Greek, Latin, Aramaic, and English — preserves it. No version, no Targum, and no manuscript tradition reached by this study inserts a night before verse 5, a morning after verse 6, or any interval between the promise and the rite.

The claim survives as a claim about the text and not about a recoverable hour in the world. This study asserts that Genesis 15 places its star-promise in daylight. It does not assert that on some afternoon in the second millennium before Christ a man stood outside a tent under a bright sky. That second claim would require the chapter to be a chronicle, which is precisely what the strongest strand of the tradition — Philo, Augustine, Maimonides, Radak — denies.

And the claim survives only under one reading of the narrative syntax: the reading on which the consecutive chain from verse 1 to verse 18 is taken to report one continuous episode. That reading is the natural one and is supported by four independent considerations, but it is not compelled by grammar, and the honest statement of the result includes the condition.

The four reasons

One. The chapter marks no interval, and it marks other things. Genesis 15 is unusual in Genesis precisely for its interest in time: it fixes the hour twice in twelve verses, distinguishes the two fixings by verbal form, and closes by gathering the whole into “that day”. Abarbanel’s sixteenth question makes the point that this is unparalleled among Abraham’s prophecies. A chapter that takes that much trouble over its clock, and then omits to mention that a whole day has passed in the middle of it, is not omitting it by accident. It is not there.

Two. The tradition’s own rule against chronological reading has been fenced off from cases like this. Ein muqdam ume’uhar batorah is the obvious escape, and Rav Pappa closes it at Pesachim 6b: the principle applies between two matters, not within one. Genesis 15 is one matter. The extension of Rav Pappa’s rule from halakhic derivation to narrative chronology is this study’s move and is labelled as such — but the fact that the tradition itself felt the need to limit the principle is not this study’s move, and it tells against a free appeal to disordered narration.

Three. Every removal of the daylight adds material to the text. The five answers surveyed above are not five readings of the chapter. They are five supplements to it: a vision that swallows the topography, an unmarked day of labour, a metaphorical exit that leaves the location unstated, a transport above the firmament, an enlargement of the eye. Each is intelligible; none is in the chapter. By contrast the daylight reading adds nothing at all. It reads the verses in the order printed and accepts the result.

Four. The reason the tradition refuses the daylight is astronomical, not exegetical, and Abarbanel says so. This is the most important of the four and it is worth stating exactly. Abarbanel does not argue that the sequence permits a night. He argues that the daylight is i-efshar, impossible, ki einam nir’im bayyom — because the stars are not visible by day. The premise is about the sky. Once it is granted, the sequence must give way, and the whole apparatus of vision frames and intervening days is generated to make it give way. Grant the premise and the tradition’s position is unanswerable. The question is whether the premise should be granted, and that is a question about what kind of sign is being given.

What the daylight reading makes of the sign

This subsection is project synthesis in the strongest sense — an original theological reading, offered as a proposal and attributed to no source.

If the stars were plainly overhead, the sign at Genesis 15:5 is an illustration. God shows Abram a large number and says: like that. It is a good illustration and Genesis 13:16 has already used the same device with dust.

If the stars were not visible, the sign is not an illustration but a demand. Abram is told to perform an act he cannot perform, on an object he cannot see, and to take the impossibility itself as the measure of the promise. The verb is not “behold” but “count”: the command is not to admire the sky but to attempt an enumeration, and the conditional — “if thou canst” — concedes in advance that the attempt will fail. On the daylight reading the failure is doubled. He cannot count them because they are too many, and he cannot count them because they are not there to be counted.

And then verse 6. Abram believed God, and it was reputed to him unto justice. The sentence that the whole Christian doctrine of justification hangs on stands immediately after a command that could not be obeyed.

That is a reading, not a proof. It is offered because it is the only reading found in the course of this study on which the difficulty is a feature of the passage rather than a defect in it, and because the alternative — that the chapter’s arrangement is an accident that four traditions of interpreters have had to repair — is a heavier hypothesis than it looks. But it is this study’s proposal. Nobody quoted in these pages holds it, and the reader should weigh it as an editorial suggestion and nothing more.

The counterargument, at its strongest

The strongest case against everything above is Ibn Ezra’s, sharpened by Cornelius a Lapide, and it deserves to be stated without softening.

The counterargument at full strength

The intervening day is not supplied by the interpreter. It is entailed by the chapter.

Verses 9 to 11 describe a substantial physical undertaking: procuring a three-year-old heifer, a three-year-old she-goat, a three-year-old ram, a turtle and a pigeon; killing them; dividing three of them lengthwise; arranging the halves in facing rows; and then standing guard over the carcasses against birds of prey for long enough that the guarding is worth narrating. That is not an hour’s work. A man who is told at verse 5 to look at the stars, and who at verse 12 is so exhausted that a deep sleep falls on him, has been working through a day. A Lapide says so in as many words — ex nimio labore diurno — and he is not harmonizing anything; he is explaining the sleep.

So the sequence is: a night vision, in which the star-promise is given under a sky that has stars in it; waking; a day of preparation; the sun going down at verse 12; the second prophecy in the dark. Nothing is inserted. The interval is read off the amount of work the text describes, and the “vision” declared at verse 1 covers the first part, which is exactly what Ibn Ezra says at verse 12: the taking of the animals was by day, when he had awoken from the vision of prophecy.

Against “in that day” at verse 18 the counterargument has a ready answer: bayyom hahu in Hebrew idiom regularly means “at that time” with no implication of a twenty-four hour span, and in any case it most naturally refers to the day of the covenant rite — the day just narrated at verses 12 to 17 — and not retroactively to everything from verse 1.

And against the argument that the versions insert no interval, it answers that translators translate what is there; the absence of an interval in the Targums shows only that the Hebrew has none to render, not that no interval occurred.

This study’s reply is threefold and is offered as a reply, not a refutation. First, the labour argument establishes duration, not a calendar day: a late afternoon of work is compatible with the text and is what verse 12’s approaching sunset actually describes. Second, if the vision covers verses 1–6 and the waking begins at verse 9, the transition is unmarked at the very point where the chapter is otherwise most careful about time. Third — and this is where the counterargument is genuinely strong — nothing in the reply is decisive. The intervening-day reading is a good reading. What it is not is the reading of the text as printed.

What would change this judgment

Findings that would overturn or materially weaken the judgment

Would overturn it. A marked temporal transition between Genesis 15:6 and 15:9 in any ancient witness. If a Targum, a Greek recension, an Old Latin manuscript, the Samaritan Pentateuch, or a Qumran copy read “and on the morrow” or “and it was morning” at verse 9, the intervening day would be textual rather than inferred and the daylight reading would be finished. Targum Neofiti, the surviving Genesis Apocryphon material, and the Samaritan Pentateuch remain unreached. Jubilees 14 was inspected as rewritten Scripture: it separates the promise from the later sacrifice but does not fix the hour of the look.

Would materially weaken it. A pre-medieval witness — patristic, targumic, Qumranic, or Hellenistic-Jewish — that treats Genesis 15:5 as a distinct night episode. Chrysostom and Ephrem now supply ancient night–day–evening sequences. They materially establish that the interval reading is ancient, while the interval remains a supplement to the biblical narration rather than an explicit temporal marker in it.

Would weaken one supporting argument without touching the judgment. A critical apparatus showing that ekei at Genesis 15:18 is secondary in the Greek. The Göttingen Genesis and Rahlfs’ text were not collated for this study; Augustine’s Old Latin already witnesses against ekei, and a critical judgment against it would remove one of the two Greek de-temporalizations this study reports. The absence of the darkness at verse 17 would stand.

Would not change it. Further attestations of the astrological reading of “outside”; further bounded Christian receptions; further recurrences of the star-formula. None of these bears on the sequence, which is where the judgment lives.

A last word on what this is and is not

The question this study set out to answer was a small one, and the reader is entitled to a plain statement of how much has been settled.

What has been settled: that Genesis 15 puts the star-promise before its own sunset; that this was noticed, and stated as a dilemma, by Abarbanel; that no checked ancient version or Targum supplies an explicit interval; that Ambrose, Chrysostom, and Ephrem preserve distinct Christian responses while no tradition-wide silence claim survives; and that the reading of “outside” as a deliverance from astral fate is attested from the first century onward.

What has not been settled, and cannot be by this method: what happened. Genesis 15 is a text about a prophetic encounter, and the four traditions that have read it most closely disagree about whether its topography is topography at all. A study of the chapter’s clock can establish what the chapter says about its own hours. It cannot establish that those hours were hours, and this one does not claim to.

Witness register

Every witness on which a claim in this study rests, with the locus at which it was read, the state of the evidence, and what it contributes. Witnesses reached only through another author’s citation are marked as such; witnesses not reached at all are listed in the following appendix among the limitations.

Scriptural text and versions

Ancient Jewish and Christian witnesses

Medieval and later Jewish commentators

Later Catholic commentators

Reported only, not read

The following are named in this study solely as another author reports them, and no claim rests on their own wording: Rabbi Nissim of Gerona, the Ran (reported by Abarbanel); Alonso Tostado, Benedictus Pererius, and Oleaster (reported by a Lapide); the verses attributed to Orpheus in Clement’s Stromateis (reported by a Lapide); Nicholas Abram and Tournemine (reported by Haydock).

Scope, Method, and Qualifications

Reader, question, and claim boundary

This study is written for a reader who wants a single verse worked to the bottom: someone who can follow an argument about verbal form without knowing Hebrew, and who wants to see the evidence rather than be told the result. No prior acquaintance with rabbinic literature or with the Septuagint is assumed; every technical term is glossed where it first appears.

The governing question is whether Genesis 15:5 — “Look up to heaven and number the stars if thou canst” — was spoken in daylight, on the internal evidence of a chapter whose sun is going down at verse 12 and set at verse 17. The claim under test was supplied by the maintainer and is not this study’s; the study’s task was to test it, not to establish it.

The boundary of the claim is textual. This study reaches a conclusion about what Genesis 15 asserts about its own sequence of hours. It reaches no conclusion about the historicity of the episode, about whether the chapter derives from one source or several, about the date of its composition, or about what a man standing in Canaan would have seen. It does not adjudicate between the Jewish and Christian readings of Genesis 15:6, does not treat the doctrine of justification, and does not treat the theology of covenant.

Terms fixed for this study

Daylight reading means the reading on which Genesis 15:5 falls before the chapter’s own sunset and therefore in the hours of light. It is the claim under test, not a conclusion assumed.

Intervening-day reading means the reading on which the star-promise precedes a night that the text does not mention, with the covenant rite falling on the following day. It is the principal rival.

Vision frame means the declaration bammahazeh at 15:1 and the question of its extent, not any particular theory of prophecy.

Gloss means a working English equivalence made by this study to show what a Hebrew, Aramaic, Greek, or Latin word is doing in an argument. A gloss is never a translation of Scripture and is never offered for recitation or quotation as such. Where a published translation is used, its translator is named at the point of use.

Project synthesis marks an editorial conclusion or proposal of this study. The sign-as-demand reading in the synthesis section, the extension of Rav Pappa’s limitation from halakhic derivation to narrative chronology, the observation that the above-the-firmament reading would incidentally dissolve the daylight problem, and the classification of the tradition into five answers are all project synthesis and are attributed to no source.

Method and source hierarchy

Each witness was read at its own locus, in its own language where the argument turned on the wording, and never through a quotation in a later author. Where that was not possible, the witness is listed in the “reported only” section of the witness register and no claim rests on it.

The hierarchy applied was: the biblical text in its identified editions first; then ancient versions and Targums; then ancient Jewish and Christian commentary; then medieval and later commentary; then modern editorial apparatus. A commentator’s report of an earlier authority was treated as evidence about the commentator, not about the authority.

Three collations were performed. The Hebrew forms at Genesis 15:5, 12, and 17–18 were verified against the identified pages of the Kittel–Beer 1909 printing. The Latin of Genesis 15 was read from a page image of the Hetzenauer printing rather than from a transcription. Abarbanel’s seventh question, on which the central historical finding of this study depends, was read in two independent digital transcriptions of the Warsaw 1862 printing and found to agree word for word.

Limitations

Text-critical. The Septuagint claims in this study are claims about the text Swete prints and the witnesses in his apparatus. Rahlfs’ edition, the Göttingen Genesis, and Wevers’ notes on the Greek text were not consulted. The reading ekei at 15:18 is reported as the reading of that edition and is explicitly qualified by Augustine’s contrary Old Latin witness. The Kittel–Beer pages verify the forms printed at the controlling loci, but that printing was not collated against a manuscript facsimile or a modern critical Hebrew edition.

Not reached. The Genesis Apocryphon and any other Qumran witness to Genesis 15; Targum Neofiti and the Fragment Targums; the Samaritan Pentateuch; Jerome’s Hebrew Questions on Genesis; the Glossa ordinaria; and the sixteenth-century Catholic commentators a Lapide names. Any of the first three could contain a temporal marker between 15:6 and 15:9 that would settle the question against the claim, and the synthesis says so.

Newly bounded witnesses. A dated Syriac web delivery was inspected, but its textual base is unidentified; it supports only a claim about that delivery, not the Peshitta tradition. Jubilees 14 was inspected as rewritten Scripture rather than as a textual version. Ambrose, Chrysostom, and Ephrem supply exact-locus positive reception evidence; Theodoret supplies only a local negative. Their deliveries still await reusable normalization.

Optical character recognition. The texts of Cornelius a Lapide, of the 1609 Douay Old Testament, of Haydock, of Swete, of Philo in Yonge’s translation, and of Josephus in Whiston’s were read in digital text layers of scanned printings. Long-s orthography and ligatures in the a Lapide and 1609 Douay texts were normalized when quoting; no word was altered. Passages quoted from these witnesses were read in enough surrounding context to establish that the sense is continuous, but were not collated against page images. This is a real limitation and it bears most heavily on the a Lapide quotations, which are short and decisive.

Language. Hebrew, Aramaic, Greek, and Latin are romanized or quoted without vowel points and without diacritical marks throughout, for reasons of typesetting rather than scholarship. Readers who need pointed or accented text should consult the editions named in the references.

Negative results. Every negative result in this study is bounded by an enumerated list of witnesses actually read. None of them is a claim that a difficulty was never noticed by anyone; each is a claim that it was not noticed in the named places.

Rights and quotation boundary

Scripture in English is quoted only from the registered public-domain Douay–Rheims in Challoner’s revision, and in the Vulgate numbering that witness uses. This study composes no translation of Scripture and offers none.

Third-party translations are quoted briefly, always with the translator named, and they retain their own rights status: Friedländer on Maimonides, Dods on Augustine, Yonge on Philo, Whiston on Josephus, Etheridge on Pseudo-Jonathan, and Rosenbaum and Silbermann on Rashi are all in the public domain. Chavel on Nahmanides, Munk on Radak, Chizkuni, Rabbeinu Bahya, and the Tur, and the Sefaria Midrash Rabbah are published under Creative Commons Attribution licences and are quoted briefly with attribution. The William Davidson Talmud and Strickman and Silver on Ibn Ezra are published under Creative Commons Attribution-NonCommercial licences; they were consulted only as controls on this study’s own glosses of the Aramaic and Hebrew, and their wording is deliberately not reproduced anywhere in this study. No third-party translation is offered under this project’s licence.

Hebrew, Aramaic, Greek, and Latin source wording is quoted from editions identified at the point of use. The Latin page image of the Hetzenauer printing and the checked transcription made from it are retained under the rights basis recorded with those artifacts. Project-created glosses, tables, and synthesis are project content.

Review state

Completed for this revision: profile routing to the scriptural-studies section of the discursive-articles profile; claim-driven witness selection; direct reading of every witness listed in the register at the locus recorded there; independent collation of the Abarbanel passage against a second transcription; a check that no verse quoted in English falls among the recorded divergences between the vendored Douay–Rheims and the 1899 American edition; source-library validation; multipass build with a clean log; and raster review of every rendered page.

Outstanding: independent review by a Hebraist, a Septuagint specialist, a rabbinics scholar, a patristics scholar, and a Catholic exegete. The text-critical limitations above have not been closed. No imprimatur, nihil obstat, or other ecclesiastical approval is claimed, and none of the conclusions here has any magisterial standing. This is a study aid.

References

Editions are identified as they were used. Where a work was read in a digital transcription or text layer rather than in the printed edition itself, that is stated. Loci given in the body are not repeated here.

Scripture and versions

• The Holy Bible, Douay–Rheims, Bishop Richard Challoner’s revision (1749–1752), in the electronic form registered in this repository. All English scriptural quotations; Vulgate chapter and verse numbering throughout. Public domain.

• The Holie Bible faithfully translated into English, out of the authentical Latin, volume 1 (Douay, 1609). Read in the archive.org text layer of the Princeton Theological Seminary copy. Public domain.

• Biblia Sacra Vulgatae Editionis, ed. Michael Hetzenauer, O.F.M. Cap. (Ratisbon and Rome: Pustet, 1914). Genesis 15 read from a 200 dpi image of numbered scan page 0012. Public domain.

• Henry Barclay Swete, ed., The Old Testament in Greek according to the Septuagint, volume I, 4th ed. (Cambridge University Press, 1909; reprinted 1925). Genesis 15 text and apparatus read in the archive.org text layer. Public domain.

• Targum Onkelos and Targum Pseudo-Jonathan on Genesis 15. Aramaic read at Sefaria; Pseudo-Jonathan also in J. W. Etheridge, The Targums of Onkelos and Jonathan Ben Uzziel on the Pentateuch (London, 1862). Public domain.

• Rudolf Kittel and Paul Kahle, eds., Biblia Hebraica, Genesis 15, Kittel–Beer 1909 printing. The controlling loci were checked against the identified page images. Public domain.

• Syriac Peshitta Genesis 15, dated peshitta.eu delivery. The textual base is unidentified; used only as delivery-specific comparison evidence.

Ancient witnesses

• Philo of Alexandria, Quis rerum divinarum heres sit, in The Works of Philo Judaeus, trans. C. D. Yonge, volume II (London: Bohn, 1854). Read in the archive.org text layer. Public domain.

• Flavius Josephus, Antiquitates Judaicae I.10.3, trans. William Whiston, read in the University of Chicago (Penelope) transcription. Public domain.

• Babylonian Talmud, tractates Shabbat and Pesachim. Vocalized Aramaic and facing English of the William Davidson edition, at Sefaria. The translation is licensed CC BY-NC and is not offered under this project’s licence.

• Bereshit Rabbah 44:10 and 44:12. English of The Sefaria Midrash Rabbah (2022), licensed CC BY; Hebrew of the Torat Emet text.

• Augustine, De civitate Dei XVI.23–24. Latin in Emanuel Hoffmann, ed., CSEL 40 (Vienna, 1899–1900); English in Marcus Dods, trans., The City of God (Edinburgh, 1871). Both registered in this repository; both public domain.

• Ambrose, De Abraham II.8.48, Latin in Patrologia Latina 14. Exact locus inspected; public domain.

• John Chrysostom, Homilies on Genesis 36.13 and 37.3–6. Copyrighted English consulted for verification only; no wording reproduced.

• Ephrem the Syrian, Commentary on Genesis XII.2–5. Copyrighted English consulted for verification only; no wording reproduced.

• Theodoret of Cyrus, Questions on Genesis 66–67. Consulted only for the bounded local negative reported in the text.

• Jubilees 14, in R. H. Charles’s 1913 public-domain English; treated as rewritten Scripture, not as a version of Genesis.

Jewish commentary

• Rashi on Genesis. Pentateuch with Rashi’s Commentary, trans. M. Rosenbaum and A. M. Silbermann (1929–1934). Public domain.

• Abraham Ibn Ezra on Genesis. Hebrew of the Piotrkow edition (1907–1911), public domain; English of H. Norman Strickman and Arthur M. Silver (Menorah, 1988–2004), licensed CC BY-NC and not offered under this project’s licence.

• Moses Nahmanides (Ramban) on Genesis. Commentary on the Torah, trans. Charles B. Chavel (New York: Shilo, 1971–1976), licensed CC BY.

• David Kimhi (Radak) on Genesis, in Eliyahu Munk, HaChut Hameshulash, licensed CC BY.

• Hezekiah ben Manoah (Chizkuni) on Genesis. Hebrew public domain; English of Eliyahu Munk, licensed CC BY.

• Bahya ben Asher, Midrash Rabbeinu Bachya (Warsaw, 1878). Hebrew public domain; English of Eliyahu Munk, licensed CC BY.

• Jacob ben Asher, Tur HaArokh on Genesis (Hanover, 1838). Hebrew public domain.

• Meir Leibush ben Yehiel Michel (Malbim) on Genesis, Hebrew Wikisource text. Public domain.

• Naftali Zvi Yehuda Berlin (Netziv), Haamek Davar, in Sefer Torat Elohim (Vilna, 1879). Public domain.

• Isaac Abarbanel, Perush al ha-Torah, on Genesis (Warsaw, 1862). Read in two independent digital transcriptions, at Sefaria and at Hebrew Wikisource, and collated between them for the passage quoted. Public domain.

• Moses Maimonides, The Guide for the Perplexed, trans. M. Friedländer, 2nd ed. (London, 1903). Public domain.

Catholic commentary

• Cornelius a Lapide, S.J., Commentaria in Pentateuchum Mosis, on Genesis. Read in the archive.org text layer of a 1700 printing. Public domain.

• George Leo Haydock, notes to The Douay–Rheims Edition of the Holy Catholic Bible with a Comprehensive Catholic Commentary. Read in the archive.org text layer. Public domain.

On the digital transcriptions used

Several of the Jewish sources above were read through Sefaria’s digital library, which supplies both the Hebrew or Aramaic base text and, where one exists, a facing translation. Sefaria is a delivery route, not an edition: in every case the edition it serves is named in the list above and in the witness register, and the rights status recorded is that of the named edition and translation, not of the platform. Where the platform’s own base text has no named printed exemplar, that is stated as a limitation in the preceding appendix rather than concealed behind the platform’s name.

Last revised (UTC): 2026-07-26T03:46:52Z

Reuse and rights. To the extent Triptych holds the rights, project-created content and design are licensed under CC BY 4.0. Scripture, liturgical or official texts, received prayers or hymns, quotations, fonts, and other third-party material retain their own status; public-domain material remains public domain. Identify changes. Attribution implies neither Triptych nor ecclesiastical approval. See LICENSE and THIRD_PARTY.md in the source.