When the New Testament Misreads the Hebrew Bible

In my previous discussion of James White’s reading of Psalm 82 I pointed out that James appeals fallaciously to the notion that Jesus’ reading of the psalm (John 10:34-35) must govern a believer’s interpretation. This is the principle of univocality, or the notion that the Bible represents a single, unified worldview, from beginning to end. This post will draw upon two places where the New Testament demonstrably misreads the Hebrew Bible, whether on its own or through the mediation of a mistranslated Septuagint, in order to show that assumptions like univocality are precluded by an informed reading of the biblical texts.

Heb 2:7–10 quotes LXX Ps 8:5–7, but the reading provided in the former is vastly different from the meaning of the latter. Here’s a brief look at Ps 8:5–7 as found in the Hebrew:

מָֽה־אֱנֹ֥ושׁ כִּֽי־תִזְכְּרֶ֑נּוּ וּבֶן־אָ֝דָ֗ם כִּ֣י תִפְקְדֶֽנּוּ׃
וַתְּחַסְּרֵ֣הוּ מְּ֭עַט מֵאֱלֹהִ֑ים וְכָבֹ֖וד וְהָדָ֣ר תְּעַטְּרֵֽהוּ׃
תַּ֭מְשִׁילֵהוּ בְּמַעֲשֵׂ֣י יָדֶ֑יךָ כֹּ֝ל שַׁ֣תָּה תַֽחַת־רַגְלָֽיו

The first line is clearly referring to humanity collectively. Both singular references to humans are indefinite and generic. The second line is grammatically contrastive (lowered // crowned), but semantically synonymous. The human is given a place of honor within the hierarchy of being, namely just under the gods (or “God,” although less likely). The dominion mentioned in the last line should not be understood as dominion over all God’s creation, terrestrial and celestial. Humanity obviously has no dominion over astral bodies. The following two lines provide proper contextualization: “sheep and oxen, and also cattle of the field; birds of the sky and fish of the sea, that which passes along the courses of the waters.” The author likely has Gen 1:28 in view: “. . . have dominion over the fish of the sea and over the birds of the air and over every living thing that moves on the earth.”

The Septuagint translation is little different, although “gods” is rendered “angels.” The spatial “a little less than” of the Hebrew is also translated with βραχυς, which can be read spatially or temporally. This is the text quoted by the author in Hebrews 2, although the meaning there is altered. To begin, the author applies Psalm 8 exclusively to Jesus. The seemingly generic singular is used throughout, as with the Hebrew, but the referent is identified as Jesus when the author finishes the quotation thus (NSV):

Now in putting everything in subjection to him, he left nothing outside his control. As it is, we do not yet see everything in subjection to him. But we see Jesus, who for a little while was made lower than the angels, crowned with glory and honor because of the suffering of death, so that by the grace of God he might taste death for every one.

“All things” is understood by the author of Hebrews to signify all created things, not just those put under humanity’s dominion. Heb 2:7 also omits the first half of Ps 8:6 (according to the best manuscripts), which states (in the Greek), “And you placed him over the works of your hands.” Other witnesses have this section, but the critical editions omit it. In the original composition this could not remain, since it was Jesus who was thought to be creator of “all things.”

The RSV translates βραχυς temporally here. Other translations offer a spatial rendering, but the author seems to be contrasting Jesus’ temporary subordination to the angels with his crowning with glory and honor. In other words, his death elevated him above that subordination. The author is likely reading βραχυς temporally. (The NRSV, by the way, changes the generic singular to plural and attempts to salvage the quotation as an accurate reading of the text as referring to humanity in general.)

What we see here is an example of a text being read according to contemporary ideologies and expediencies which differed greatly from those of its original author and community. Psalm 8 does not refer to Jesus, to the incarnation, or to his glorification. It refers to God’s grace in giving humanity a place of honor, which it does not merit, within God’s glorious created order. The author of the psalm and the author of Hebrews thus present two conflicting readings, undermining the notion that any principle of univocality governs the literature of the Bible. This does not mean the reading in Hebrews is useless, though. It renegotiated Christianity’s relationship with its sacral past, injecting new relevance into the text for Christians and strengthening their connection to Judaism’s sacred literature.

My next case study involves the application of a mistaken translation to a question of doctrine. In Acts 15:13–17, James appeals to Amos 9:11–12 in an effort to support through scripture the taking of the gospel directly to the Gentiles. It even seems James’ quotation settles the debate. The critical portion of Amos 9 reads,

I will rebuild the tent of David, which has fallen, and from its ruins I will rebuild it and set it up, so that the remnant of the people might seek the Lord, and all the nations which call upon my name.

This reading comes from LXX Amos, although there is a bit of movement. For instance, “the Lord” is an addition. The LXX actually omits the object, reading, “so that the remnant of the people might seek, and all the nations . . .” There is also a clause missing from Acts’ quotation (“as the days of old”). The important observation, however, is the Greek translation’s relationship to the Hebrew. The crucial section reads in the Hebrew, “that they may possess the remnant of Edom,” but is translated, “so that the remnant of the people might seek,” in the Greek. The confusion arises likely because of the lack of the mater lectionis which we find in MT in the word אדום. Without it, the word looks an awful lot like אדם, “man,” or “humanity.” The verb “to possess” (יירשׁו), was also misunderstood as “to seek” (ידרשׁו).

It is unlikely that MT is secondary. First, there’s no object for the transitive verb εκζητησωσιν, “that they might seek.” Second, the reading in MT makes more sense within the context. David’s fallen house would be restored so that it might reassert its authority, specifically in overtaking the remnant of Edom (see Amos 1:11–12) and “all the nations,” for which Edom functions as a synecdoche (Edom commonly acts as a symbol for all of Israel’s enemies [Ps 137:7; Isa 34:5–15; 63:1–6; Lam 4:21]). The notion that the restoration of the Davidic kingdom would cause the remnant of the people (why are they only a remnant?) and all the nations to seek the Lord is also a bit of a disconnect within Amos.

This quotation shows not only that the early church relied on the Septuagint, but that it rested significant doctrinal decisions on the Greek translation, even when it represented a misreading of the underlying Hebrew.

The notion of univocality within the Bible as a whole is irreconcilable with these data. The axiom that scripture should interpret scripture is wholly undermined by these two examples (and many others could be pointed out). There is not only one voice in the Bible, and I do not think it prudent to approach any single chapter or verse within the Bible assuming that it contains only one voice. There are numerous voices throughout the biblical texts saying numerous different things for numerous different reasons. New Testament exegesis of an Old Testament text is no more authoritative a reading than that of any other exegete.


15 responses to “When the New Testament Misreads the Hebrew Bible

  • Chris

    This post will draw upon two places where the New Testament demonstrably misreads the Hebrew Bible, whether on its own or through the mediation of a mistranslated Septuagint

    Doesn’t this presuppose the idea that there was a “fixed” version of scripture the ancient congregation was using? Couldn’t the quote come from a perfectly valid manuscript (or tradition) that they used for teaching?

    The writer of each book (when quoting OT) was probably using scriptures that they believed to be authoritative but not a collection of “approved” texts, correct?

    • Daniel O. McClellan

      Thanks for commenting Chris. I don’t believe there was a canon at this time, or particularly standardized manuscript traditions for each biblical text, but that only serves to further my point that the Bible is a cacophony of voices, and not univocal. The two texts I shared are rather securely identified with the LXX translation, and come from Vorlagen which didn’t vary significantly from MT.

      In the end, modern conservatives assert total harmony between a specific version of the Hebrew Bible and a handful of versions of the New Testament when it’s clear that the authors of the New Testament both (1) saw a different version of the HB as authoritative, and (2) read their contemporary (and distinct) worldviews into those texts. All readers of the Bible do the exact same thing today to some degree or another, and I’m not condemning early Christians for that. My fundamental point is that we have to recognize that biblical univocality is utterly precluded by the nature of the texts.

  • Jason A. Staples

    I’m not sure I agree that Hebrews is only referring to Jesus in its interpretation. Rather, it appears to me that the Psalm is read as referencing humanity, but it is wrestling with the problem that at this point creation is not subject to human beings. Only then does it turn to Jesus, explaining that he (first pointing out his humanity) is presently crowned, serving as the firstfruits of this promise. Thus human beings will indeed rule over all things. So Jesus is read in light of the Psalm, but I’m not so sure the meaning of the Psalm is passed over quite so quickly. It’s a fine distinction, but one worth making.

  • Daniel O. McClellan

    Thanks for the follow-up, Chris. I’m glad we agree on “many voices.” My post is largely in response to comments that I’ve heard over the past week or so about reading Hebrew Bible texts in light of New Testament readings, as if New Testament exegesis (or Christ’s exegesis) is the only secure way to know what the texts originally meant. As I stated in the post originally, this was James White’s view in his critique of the consensus reading of Psalm 82.

    • Chris

      I think my issue was the title heading and part how you started this post out:

      “When the New Testament Misreads the Hebrew Bible”

      If there was no “bible” at the time Hebrews was written, but there were [some] collections of written scriptures and probably lots of oral traditions being passed around, it seems a bit misleading to say the NT misreads the HB.

      I think that is where my disconnect is coming from. That’s all.

      I talk to a lot of people who only interpret scripture based on Luke 24:44-45 – that all of scripture points to Christ and must be interpreted that way.

      • Daniel O. McClellan

        I see where you’re coming from, Chris. My title was more rhetorical than anything else, although I don’t think it misses the mark significantly. I agree there was no concept of a canon, but there was a somewhat fluid collection of authoritative Jewish scripture. No doubt different people had different ideas about what did and did not constitute scripture, but neither of the texts I’ve discussed really belong to areas of contention, or to oral tradition. They were cited as prophetic literature. Acts 15:15, for instance, states, “this agrees with the words of the prophets, as it is written . . .” Now, granted, Acts isn’t misreading the text they had before them, but in the context of this discussion of the Bible’s univocality, the current NT is misreading the OT that is asserted to be the authoritative one.

  • John Meade

    Daniel,

    Thanks for bringing this issue to our attention. As one who believes the canon and the text form of the OT was pretty well fixed by the NT era, I thought I would offer a bit of a reply to your latest post on Psalm 8 and Heb. 2 by commenting first on the relation of the LXX to HT, and then the relation of the text to the NT.

    The Greek Psalter uses αγγελος 13 times to render three Hebrew words: אביר אלוהים, and מלאך. Of these 13 times, it uses αγγελοι to render “gods” 3 times (8:5, 96(97):7, and 137(138):1). In 96:7, αγγελοι is a very fitting translation of אלוהים since it clearly has “gods” or “divine beings” in view and not the God of Israel, YHWH, while the other two references are debatable. There is clearly not a different Vorlage underlying the LXX, since it is an established equivalent in the Psalter (see also Job 1:6 et al.). So there is no textual fluidity here, but rather indication that the translator is attempting a somewhat close rendering of a text similar to M. When the Hebrew Psalter refers to false gods/judges(?) or elohim in 81(82):1 and 85(86):8, the translator uses θεοι. Thus it would seem that the LXX has detected three meanings of elohim: God, angels or divine beings, and false gods/judges, and it has a different equivalent for each, clearly distinguishing the true God of Israel from the false gods by employing the singular and plural of θεος respectively. All of this seems like good translating on the part of the LXX, not a different Vorlage.

    Now, the NT use of Psalm 8 is interesting. I will say that your strongest point is actually not the Author’s use of the LXX Version with its translation of αγγελοι or βραχυς (even you admit this is a perfectly acceptable translation of Hebrew מעט), but the seeming different text form than the OG since one of the lines is missing (Ps 8:6 [sic 8:7] LXX; Heb 2:7). You also note that there is a textual variant in the NT witnesses, but you dismiss the longer text without much demonstration. You may be right, but I would like to see some demonstration for this, since the longer reading is still in Sinaiticus, Alexandrinus, and D original among many others. True P46 and B has the shorter text, but I would view the external evidence as split, and there needs to be an explanation of the internal evidence. Perhaps the scribes of those witnesses made a theological adjustment. Of course it is possible that the Author omitted the line, but does this theological reading of the text actually indicate textual fluidity?

    Setting aside the issue of text form, does Hebrews play fast and loose with the text? Perhaps you are aware of the debate over the referent of the “Son of Man” in this text. Some understand the text to refer to humanity, even in this text, while others understand it to refer to Christ. One can actually argue for both here. The Christological interpretation is not as far off as you imagine, if the Psalm is originally composed by David, the anointed one, who was promised to be King over all of humanity and creation (2 Sam. 7:13-19). If the OT is actually presenting David as a new Adamic figure, a king, who will reign over all things, then it is not a stretch to apply this text to David’s greater son, Jesus Christ.

    In summary, 1) the LXX translation is faithful to a Hebrew text similar to M, and thus not a witness to textual fluidity. 2) the NT text may actually attest to the exact wording of the LXX, but an argument will need to be made for one or other of the variants. Even if the author omits the line, does this indicate textual fluidity? 3) There is a complex exegetical issue in Heb. 2: does the Author refer to Christ or to Humanity? Again, either option will have to be argued. So, rather than textual fluidity or bizarre exegesis, the author seems to be making a fair argument for Christ’s supremacy on the basis of a fairly exact Greek translation.

    I apologize for the length of this comment, but I figured that your post deserved it :).

  • Daniel O. McClellan

    Hi John. I appreciate all the replies my posts gets, and especially the ones which approach the issues I bring up intelligently.

    Your first paragraph raises an area of research I think is fascinating, namely the Hellenistic Period’s perception of deity. I presented a paper at SBL on LXX Deuteronomy’s view of deity and spent much of the time discussing the Septuagint’s proclivity for rendering אלהים and בני אלהים with αγγελοι. I agree that the translator thought without much concern that “angels” was a perfectly legitimate equivalent for אלהים, but I tried to show in my SBL paper that that position was a new one and it was one that translator’s seemed to think needed enforcement. Heb 1:6 also cites a Septuagint translation that changes “gods” to “angels,” although in that verse the translator altered the text quite a bit, turning “gods” into “sons of God” and then adding an additional colon that read “angels of God.” I have argued that the intention was to identify the sons of God with angels, which was a conflation I argue was developing at the time. I would argue that “angels” is not what the author of the psalm intended. As far as I can tell, when they meant angels they were pretty specific about writing “angels.” In that sense, I believe the Septuagint translator has misread his Vorlage.

    Regarding your second paragraph, I would first appeal to lectio brevior. The shorter text, all other things being equal, is preferred. The omission is also easily explained, since it irons out problems with the fact that the object of Psalm 8 doesn’t seem to be the one who created all things, and Christ was certainly thought to have created all things. This seems to me to make the most sense as the work of the author. The full quotation in the other witnesses is also easily explained as later harmonization with the LXX version. I’m open to other ideas, though.

    Regarding your third paragraph, I haven’t read the scholarship on that debate, but I’d be interested to see if any of it addresses the concerns I brought up in the post. I think it is interesting that the RSV seems to side with me, but the NRSV pulls a complete 180, going so far as the change the singular to plurals. I also don’t see what kind of argument could be made for dating Psalm 8 that far back.
    Additionally, I think it would be difficult to argue “son of man” is a messianic designation in the Hebrew Bible. It just refers to humans generically (it’s used numerous times in Ezekiel that way). Isa 56:2, in fact, has the exact same parallelism as Psalm 8 (אנוש and בנ־אדם), and there it’s unambiguously generic. The messianic reading is an innovation of the Greco-Roman period.

    Regarding your summary, I’m not necessarily arguing that LXX Psalm 8 was textually fluid. The translator rendered a fairly isomorphic translation. I think there the only significant divergence, really, is in the interpretation of the author of Hebrews, but I don’t think his exegesis is bizarre by any means. I don’t think he had access to the original contexts, and I think he definitely had a rhetorical agenda, but he was certainly a capable exegete for his time. The reason I brought up that text was because it provides an example of understanding a text in a much different way than when it was originally composed. The Acts/Amos example is an example of the LXX translator missing the mark more than the NT author, but I brought up that one to point out the problem with insisting our current NT is ideologically linked with our current OT, based on MT.

    Thanks again for your comments, and I’m always interested in your LXX thoughts.

  • Does the NT Misread the OT? « LXX Studies

    […] having a bit of an exchange on his blog regarding the whole issue of the NT’s use of the OT here. Daniel posits that Hebrews 2:5ff has misread Psalm 8 via the mistranslation of the LXX. In […]

  • Howard Mazzaferro

    Hi Daniel,

    How do we know that the Hebrew wasn’t wrong and that the LXX was right? Below is a quote from, “Old Testament Quotations in the New Testament by Gleason L. Archer and Gregory Chirichigno” where he seems to think the Hebrew was wrong. (The Hebrew does not appear to be showing up, I hope its readable.)

    But Acts 15:17 introduces ton kurion as the object of the “seeking.” This strongly suggests the following emendation to the MT: (a) read MT’s Wvr.yyI as Wvrd.yI, quite possibly relying on the LXX Vorlage that read (in the Qumran period, perhaps) wXrdy rather than WvryyI. “In order that the rest of mankind might seek him” fits in much better with the context than a promise of taking possession of Edom! (b) read -ta, before tyriaev. as Atao or ytiao. In the course of scribal transmission it would be quite possible for a final waw or yod to drop out because of a worn spot in the Vorlage of the Sopherim text. As amended then, Amos 9:12 should read as follows: ŒŒwgw ~yIAGh;-lk’w> ~d’a’ tyriaev. Atao Wvrd.yI ![;m;l.. Thus the rendering of the LXX (= NT) would be completely accurate, and we may feel grateful that in this verse we have access to the earlier and more authentic reading: “In order that the remainder of mankind might seek him/me (ytiao ?), and all the Gentiles (upon whom my name is called).” -(Old Testament Quotations in the New Testament by Gleason L. Archer and Gregory Chirichigno)

  • The Biblical Studies Carnival LVIII « כל־האדם

    […] he explores the question, “What is Deity in LXX Deuteronomy?” He also writes about when the New Testament misreads the Hebrew Bible, to which John Meade (LXX Studies) responds with the question, “Does the New Testament […]

  • Should Scripture Interpret Scripture? | timothymichaellaw

    […] him; should I brag again that I once supervised his award-winning Master’s thesis?) called ‘When the New Testament Misreads the Hebrew Bible’. It’s a good post, although I’d disavow the dependence of the NT authors on the Hebrew […]

  • Should Scripture Interpret Scripture? « Thoughts in the Dark

    […] Here is the link to the post that the above post is referring to! Advertisement […]

  • Review: Journal for Trinitarian Studies and Apologetics (2) | Daniel O. McClellan

    […] in v. 16 and then adapts the Septuagint’s misreading of Amos 9:12 in v. 17 (see my discussion here).[9] The resulting text has absolutely nothing whatsoever to do with the actual words of Amos 9, […]

Leave a comment