Abraham ibn Ezraאַבְרָהָם אִבְּן עֶזְרָא

English

Our sages have said that the [letter] bet was added, like the bet in barishonah (“in the beginning”) Bereshit 13:4, for you find “at the beginning (rishonah) they travelled” Bamidbar 2:9. But were that the meaning, the bet would have been vocalized with a kamatz gadol. And there are those who say that [the word] bereshit always [appears in] attached [form], where the meaning is “at the beginning of the evening, or the night, or the darkness”. But behold, they forgot “he saw a beginning reishit for himself” Devarim 33:21. And there are those who say that the bet is a subject without meaning. Their reason is that it is unthinkable that there is no beginning to the heavens and the earth. Therefore, it said “bereshit. But according to my opinion, it is indeed an attached form, like “At the beginning (bereshit) of the rule of Yehoyakim” Yirmiyahu 26:1. Now don’t wonder how can it be attached to a past tense verb [instead of a noun], for behold, thus: “At the beginning (techilat) spoke Hashem with Hoshea, and Hashem said to Hoshea” Hoshea 1:2, “The city (kiryat) where David camped” Yishayahu 29:1. The meaning [of this usage] will be explained for you in the second verse.

Sefaria Community Translation · cc0

Hebrew / Aramaic

Ibn Ezra on Gen 1:1, s.v. בראשית · pd

Notes

The crux conducted purely as grammar, with the theology left out. Ibn Ezra sets down three opinions and refutes each from a counter-example: that the bet is otiose (then it would carry a long qamats); that bereshit is always construct and means 'at the beginning of the evening, or the night' (they forgot Deut 33:21, 'he saw a reshit for himself', where it is absolute); that the bet carries no sense at all and was written only so that no one should think heaven and earth had no beginning. His own view is Rashi's — a construct, like 'in the reshit of the reign of Jehoiakim' — and he answers in advance the objection that a construct cannot govern a finite verb by citing Hos 1:2 and Isa 29:1. Then he stops: 'the meaning will be explained for you at the second verse.' The Latin bench never argues at this level about in principio, because in Latin there is nothing to argue about.

Threads

• → parallel · Rashi
  The same conclusion from the same evidence with no acknowledgement: bereshit is a construct, like 'in the reshit of the reign of Jehoiakim' (Jer 26:1), and the objection that a construct cannot govern a finite verb is answered from Hos 1:2 by both. Ramban, reading them side by side, says Ibn Ezra 'explained it in an identical way'.
• ← cites · Moses ben Naḥman (Naḥmanides)
  Ramban names R. Abraham and reports the refinement that is Ibn Ezra's own and not Rashi's — that the vav of ve-ha'aretz serves as 'when' — and draws its consequence, that on Ibn Ezra's reading only light was created on the first day.
• ← parallel · Chalcidius
  Aquila's Greek, in a fourth-century Latin dress, is the grammarian's reading eight hundred years early. Caput rerum condidit Deus coelum et terram takes reshit as a construct governing what follows — the head of things — which is exactly Ibn Ezra's argument at Lucca, that bereshit is a construct and must be read with what comes after it. Chalcidius prints it as one of four possible renderings and draws nothing from it; Ibn Ezra makes it the plain sense of the verse; and the reading reaches modern Bibles from the second bench and not the first.