When did God divide light from darkness?

What follows is one scholar’s way of reading this disagreement — a critical-historical view that often concludes the texts genuinely disagree. It draws on the peer-reviewed sources listed below and stands as one position among the answers above, presented for transparency, not as GEBible’s position.

This is an unusual case because both verses sit within a single chapter and pivot on a single Hebrew root, b-d-l ('to separate, divide'), in the Hiphil stem. Day one reports a completed act: vayyavdel elohim bein ha-or u-vein ha-choshek, 'and God separated the light from the darkness' (Gen 1:4), and God then names the light Day and the darkness Night. Day four assigns the same function to the newly made luminaries: they are set in the firmament 'to give light upon the earth, and to rule over the day and over the night, and to separate (u-le-havdil) the light from the darkness' (Gen 1:18). The puzzle is plain: if the light/darkness division was accomplished and named on day one, before any celestial body existed, what is left for the sun and moon to divide on day four? The same verb governs both, once as a finite wayyiqtol (vayyavdel) and once as an infinitive of purpose (le-havdil), which makes the repetition unmistakable rather than incidental.

The traditional harmonization, run by Archer and many conservative commentators, posits two phases: day one introduces a primordial, non-solar light — sometimes identified with the divine glory or an undifferentiated cosmic luminescence — and day four institutes the regular, mechanically governed alternation of day and night through the sun and moon. This reading has ancient pedigree. The notion of a primordial light distinct from the sun appears in rabbinic tradition (the or ganuz, the 'hidden light' that b. Chagigah 12a says was created on day one and then stored away for the righteous in the world to come), and the difficulty of pre-solar light was felt across Second Temple literature (4 Ezra 6:38-41 dwells on the brightness commanded forth at the beginning of creation). As a piece of theology the primordial-light reading is coherent and old. As an account of what the text describes, it strains, because the verb and object are identical in both verses; the text does not say day four creates a new kind of light-cycle but that the luminaries perform the same dividing already reported in 1:4.

The more economical explanation, and the mainstream critical one, is literary-structural. Mark Smith, Gordon Wenham, and Claus Westermann all observe that Genesis 1 is organized as a deliberate two-panel framework: days one through three create the realms or domains (light/day, sky and sea, dry land with vegetation), and days four through six populate or govern those domains (luminaries to rule day and night, birds and fish, land animals and humans). Day four corresponds to day one precisely because the luminaries are the 'rulers' (the verb m-sh-l, 'to rule,' Gen 1:16, 18) installed in the light-domain established on day one. The repetition of the separating vocabulary is therefore not a claim about two distinct physical events on a single timeline but a literary signal of the day-1/day-4 correspondence. Wenham presents this framework as the key to the chapter's architecture; Westermann reads the luminaries' 'separating' as their ongoing governing function rather than a redundant re-creation of the day-one division. The polemical thrust reinforces the reading: by delaying the sun and moon to day four and assigning them a merely functional, subordinate role, P strips the great astral deities of the surrounding cultures of their divinity — refusing even to name them, calling them only the 'greater' and 'lesser light' (Gen 1:16) — while still acknowledging the obvious fact that it is the sun that now governs the day.

It is worth adding that the framework reading does not require denying that P thought light somehow existed before the sun; the ancient cosmology that produced this text plainly did distinguish light as a created reality (1:3) from the luminaries as later furniture (1:14-19), and that distinction is itself a marker of the chapter's pre-scientific astronomy rather than a covert two-phase apologetic. The day-one light is the cosmic condition that makes day and night possible; the day-four luminaries are the bodies installed to administer that already-instituted cycle. On the author's own terms the repetition is functional, not contradictory.

So the assessment depends on the question being asked. If one reads Genesis 1 as the carefully patterned liturgical cosmology its author composed — with its sevenfold rhythm, its refrains ('and God saw that it was good'), and its panelled symmetry — the repeated b-d-l language is intelligible and not self-contradictory; the structure is the point, and the author very likely was not narrating sequential natural history at all. If one insists on reading it as a strict chronological report, then the text does separate light from darkness twice, once with no luminaries and once with them, and that is incoherent. The contradiction is thus real under a flat-literal reading and dissolves under the literary reading the text's own design invites.

Where the critical reading lands: harmonizable as deliberate framework structure; a genuine contradiction only under a chronological-literalist construal the author probably did not intend.