Bible Translation Confidence


While scholars agree on the translation of most passages in scripture, some chapters and verses are difficult to translate and there are varying levels of consensus on the meaning of the text. While reading The New International Commentary of the Old Testament - The Book of Job by John E. Hartley, I realised that my application of scripture would be helped if I understood the degree of confidence that scholars have in the translation that I’m reading. This is a proof-of-concept of an inline indicator of translation confidence. It shows footnotes from the commentary and the purpose of the footnote, either a clarification with no impact to meaning, some disagreement as to the correctness of the translation or a significant disagreement as to the correctness of the translation.

This proof-of-concept shows part of chapter 30 from the book of Job, with text and notes from The New International Commentary of the Old Testament - The Book of Job by John E. Hartley.

Components of Confidence

  • Text: How agreed the scholars are on what was actually written in the original text. Lack of consensus may be because of a word that seems out of place, or looks like it may have transcription errors based on content.
  • Translation: Lack of consensus may be because the word is unfamiliar, uncommon in this context
  • Meaning: Often because of problems with translation, but may also happen because of turns of phrase that are unfamiliar or not understood


Bundling all the components of confidence gives the following legend:

  • Consensus, with elaboration in footnotes: There is consensus on the translation and the footnotes provide additional information
  • Some consensus: There is some consensus on the translation and the footnotes explain diverging opinion
  • Little/no consensus: There is little consensus on the translation and the footnotes explain diverging opinion

Open the footnote references to show the translation notes.

16Now my soul is emptied from me;
days of affliction seize me.
17At night God1 painfully pierces my bonds,
and my veins2 do not sleep.
18 With great force he3seizes4 my clothing
and girds me about like the collar of5 my tunic
19He6has cast me7 into the mire;
I have become as dust and ashes.
20I cry to you, but you do not answer me;
I stand, but you do not8 heed me.
21You have grown cruel toward me;
with your strong hand you act hatefully against me.
22You lift me up and mount me on the wind;
you toss me about with a tempest.10
23I know that you will bring me to death,
to the meeting house for all the living.”
24Yet God does not stretch out his hand in destruction,
if one cries to him for help in his disaster.11
25Did I not weep for him whose day was hard?
Was not my soul grieved12 for the poor?
26But13 when I looked for good, evil came;
when I hoped for light, darkness came.
27My bowels boil unceasingly;
days of affliction meet me.
28I go about blackened, but not by the sun;14
I stand in the assembly and cry for help.15
29I am a brother of jackals
and a companion of ostriches.
30My skin blackens and peels;
my body16 burns with fever.
31My lyre is tuned to wailing,
and my flute to the voice of mourners.”17
  1. In this verse God, the cause of all Job’s afflication, is assumed to be the undefined subject of “pierce” (niqqar). Two other possibilities of treating MT niqqar are to take laylâ, “night”, as its subject or to revocalize it as Pual, nuqqar, with ‘ªṣāmay, “my bones,” as its subject: “At night my bones are pierced with pains.” ↩

  2. Of the many suggestions for MT ‘ōrᵉqay, two are more likely. One is “my veins,” gained from the Arab. ᵉirq, “veins and sinews” (Tur-Sinai, Gordis). This view has the advantage of having a part of the body parallel to “bones.” The other is “my gnawing pains.” based on a gonate root in Syriac and Arabic. A major problem with this position is that since the Syriac version translates the word here “my body,” the Syriac translators did not equate ‘ōrᵉqay with the same root in their language. ↩

  3. The subject of the very yiṯḥappēś, “he seizes,” is either indefinite, alluding to the disease, or God (Pope, NIV). Either way God is the final cause and thus is accepted as the subject. ↩

  4. For MT yiṯḥappēś, “disguise oneself,” Dhorme, Fohrer, and Pope read yiṯpōś, “he seizes,” based on LXX. But Gordis thinks that Heb. ḥāpaś may be a phonetic variant of ḥāḇaš, “bind up.” He argues that the Hithpael arose as a conflation of second and third person forms by reason of the shift from third person (vv. 17, 19) to second person (vv. 20-23). ↩

  5. Some (e.g., Pope and Gordis) think that for kᵉpî, “like the collar of,” it is better to read bᵉpî, “by the collar of.” Since it is possible to make sense out of MT, it is better not to emend the text. ↩

  6. Since the first line is unusually short, Pope may be correct in supplying “God” as the subject. ↩

  7. Many (e.g. Duhm, Hölscher, Fohrer) believe the verb hōrānî, “throw me,” often use of shooting arrows, is an error of haplography for hōriḏanî, “he has sent me down.” This view is possible, but not compelling. ↩

  8. It seems best to read lō’, “not,” or wᵉlō’, “and not.” before tiṯbōnen, with Vulg. (so Dhorme, Rowley) ↩

  9. The Qere tušîyâ, “success,” is less preferable than the Ketib tᵉšuwwâ as a variant spelling of tᵉšu’â, “noise, roar” (of a storm) (cf. 36:29, 39:7). Dhorme takes it as subject, but it is possible to interpret it as the accusative of place (Gordis). ↩

  10. Verse 24 is very difficult and has received numerous reconstructions. For MT ‛î, “heap,” Pope and Dhorme read ‛ānî, “needy, afflicted”: “One does not turn his hand against the needy, when in distress he cries fo help.” Hakam takes bᵉ‛î as equivalent to a common Aramaic word for “prayer, request” (cf. Dan. 6:8, 14 [Eng. 7, 13]), with the preposition bᵉ omitted for assonance. This word parallels šûaᵉ in the second line, which is elliptical for ᵉāraḵ šûaᵉ, “cry for help.” As for the antecedent of lāhen, “to them,” Hakam identifies them as “death” and Sheol (“meeting house”) in v. 23 and “hand” in v. 24a, and he reasons that lāhen is a feminine plural, for šᵉ’ôl and yāḏ are feminine: “Surely God does not bring death by petitiion, If in his distress one cries for them (i.e., death, its instrument and its abode).” Although this interpretation is somewhat strained, it has the advantage of reading the existing text.

    Gaining a clue from T.B. Aboda Zara 4a, Grabbe (Comparative Philology, pp. 101-103), believes that ‛î could mean “destruction”: “Indeed let him not send (his) hand with (complete) destruction/ruin if in his calamity there is accordingly a cry (for help).” This suggestion, too, has the advantage of no textual change plus some rabbinic support. This view of ‛î is accepted tentatively, but it seems best to find lōh yᵉšawwēaᵉ, “one cries to him,” in lāhen šûaᵉ (see BHS). Such a view takes the n as a scribal error rising from the coalescing of an h and a y. ↩

  11. The meaning of ᵉāgam, “be grieved,” is clear from its use in Mishnaic Hebrew. Perhaps this word occurs in the Ugraritic text Krt 26-27: yᵉrb bḥdrh ybky bṯn []gmm wydmᵉ, “He entered his chamber to weep; while repeating his grief, he shed tears” (cf. A. Ceresko, Job 29-31, pp. 91-92). ↩

  12. Gordis suggests that the gives reason for v. 27. ↩

  13. The expresssion bᵉlō’ ḥammâ, “without the sun,” may be elliptical for “without the light of the sun” (Delitzsch), or it may mean that one’s face is covered so that he does not see the sun (Hakam). Duhm, Budde, and Fohrer read bᵉlō’ neḥāmâ, “where there was no comfort.” Although this change makes excellent sense, both LXX (despite its own textual problem) and Syr. read the same consonants as MT, which is thus retained. ↩

  14. The use of the imperfect ’ªšawwēaᵉ, “I cry for help,” after the perfect qamtî, “I stand,” comes close to having the sense of a final clause (Driver-Gray; GKC, § 120c; Driver, Hebrew Tenses, § 163). ↩

  15. Dhorme shows that ‛aṣmî (lit. “my bone[s]”) in the singular denotes one’s frame; cf. Ps 102:6 (Eng. 5): “My frame [‛aṣmî] cleaves to my flesh.” ↩

  16. Gordis suggests that bōḵîm, “those who weep,” may be an abstract noun, “weeping,” parallel to ’ēḇel, “mourning.” ↩

Source code is available in my Github repo.

Please let me know if you are aware of any translations with any sort of confidence indicator.