Interestingly, but not surprisingly, the NIV has smoothed over a difficulty in Genesis 2:2 with a “possible” reading.
The NIV reads: “By the seventh day God had finished the work he had been doing; so on the seventh day he rested from all his work.”
The ESV reads: “And on the seventh day God finished his work that he had done, and he rested on the seventh day from all his work that he had done.
The Hebrew reads: וַיְכַל אֱלֹהִים בַּיּוֹם הַשְּׁבִיעִי, מְלַאכְתּוֹ אֲשֶׁר עָשָׂה; וַיִּשְׁבֹּת בַּיּוֹם הַשְּׁבִיעִי, מִכָּל-מְלַאכְתּוֹ אֲשֶׁר עָשָׂה.
The problem is that the ESV is more accurate. To translate the bet-preposition as “by” is a stretch. To translate wayekal as a pluperfect (“he had finished”) is also a stretch. Even more of a stretch is to translate bayom as “by…day” in the first instance and then translate the exact same phrase only 5 words later “on…day.” Why does the NIV do this?
Well, because God is supposed to be resting on the seventh day, not finishing up his work. The same language is used in Exodus 20:10 and even explictly says that God made heaven and earth in six days. So what do we do?
Well, rather than trusting in all of these stretches Ronald Hendel in his apologetic for a critical edition of the Hebrew Bible called The Text of Genesis 1-11: Textual Studies and Critical Edition suggests taking the textual variant found in the Samaritan Pentateuch, Septuagint, Syriac Peshitta, and Jubilees 2:16 and replacing “seventh” with “sixth.”
Now, typically I am skeptical of taking textual variants but Hendel makes a good case for it.
1. “…to posit that scribes or translators changed the text independently in three (or four) textual traditions is extremely unlikely, given our cognizance of the numerous shared readings in G, S, and Syr.” Also, G of Genesis is known for conserving the Vorlage so that reading “sixth” for a Proto-G is warranted. So then, it is better to argue for a common root than independent traditions.
2. So the question must now be settled on text-critical grounds. While typically the harder reading is to be accepted, in this case there is another plausible motive for why “sixth” could have given rise to “seventh.” Verse 2 can be split up in this way:
בַּיּוֹם הַשְּׁבִיעִי מְלַאכְתּוֹ אֲשֶׁר עָשָׂה
בַּיּוֹם הַשְּׁבִיעִי מִכָּל-מְלַאכְתּוֹ אֲשֶׁר עָשָׂה.
“With the exception of the stylistic variation of mkl in v 2b, the two sequences are identical but for the variation of [“sixth”] in the place of [“seventh”]. It is entirely possible that a scribe could have miswritten [“seventh”] in lace of [“sixth”] in the first clause, triggered by anticipation of the parallel in the second clause. This would be an accidental assimilation by anticipation” (33).
So the difficult reading is chalked up to scribal error. I like it, mostly because that’s the best explanation I’ve heard so far.