top of page

A (No) Nonsense Reading at Hebrews 12:15

The Nestle-Aland Novum Testamentum Graece 28th edition (NA28) leverages a fantastic assortment of signs and abbreviations to capture a great deal of textual variants in a minimal critical apparatus. One of those is the symbol “(!),” for “sic!” which “indicates an accurate transcription of an apparently meaningless reading” (1). By my count, it is used 19 times in the NA28 apparatus.

Hebrews 9:3 presents one of the least debatable examples of a “meaningless” reading, occurring in the papyrus P46. The edited text of the end of the verse reads ἡ λεγομένη  ⸂Ἅγια Ἁγίων⸃ (“called the most holy place” ESV). In P46, a single word, ἀνά, is substituted in for Ἅγια Ἁγίων. Ἀνά is a preposition that can be glossed “in the midst of” or “between,” among other things. It would change the translation (presented without much effort) to “called between.” This just doesn’t fit, even if one were to draw on the text from the next verse. No amount of grammatical mind-bending can account for the substitution of ἀνά here in P46, yet there it is.

Another example, also from P46, likewise appears in Hebrews. Yet this one may not be as absurd as it at first seems.

In Hebrews 12:15, a portion of the text reads μή τις ῥίζα πικρίας ἄνω φύουσα ἐνοχλῇ (“that no ‘root of bitterness’ springs up and causes trouble” ESV). NA28 marks that final word ἐνοχλῇ with a “sic!” notation, as having one of those “meaningless but accurate” variants in P46: ενχ[.]λη (where the fourth letter is uncertain due to manuscript damage). There are good reasons to think the reading in P46 is actually quite meaningful.

First, as even NA28 indicates in a marginal reference, this portion of Hebrews 12:15 is drawing from Deteronomy 29:17 in the Septuagint (LXX, the Greek translation of the Hebrew Old Testament; 29:18 in modern English translations). Here is a comparison of the two texts:

Dt 29:17 (LXX): μή τίς ἐστιν ἐν ὑμῖν ῥίζα ἄνω φύουσα ἐν χολῇ καὶ πικρίᾳ

Heb 12:15 (NA28): μή τις ῥίζα πικρίας ἄνω φύουσα ἐνοχλῇ

There are several differences between the source and what appears in Hebrews, but the one in focus here is between ἐν χολῇ (LXX) and ἐνοχλῇ (NA28). The former is a prepositional phrase (“in bitterness”) while the latter is a finite verb in the subjunctive mood (“it may trouble,” from ἐνοχλέω). The importance is that the sequence of letters in Deuteronomy (ἐν χολῇ) matches that of the variant in papyrus P46 (ενχ[.]λη), save for the indecipherable fourth character. If that fourth character is an omicron (ο), it would match the LXX text of Deuteronomy 29:17 perfectly.

Second, as it turns out, that fourth character may not be so impossibly lost as may be implied, as Figure 1 demonstrates.

Figure 1: ενχ[.]λη from papyrus P46. (Image credits: CC BY-NC. Chester Beatty, Dublin)

The ink is not completely absent in that fourth character position. While it is not possible to know for certain, given the ink that is present, and knowing that this is a quote from Deuteronomy 29:17, positing an omicron there would not be a hazardous guess. Indeed, comparing the extant ink with an unblemished omicron on the line below (part of πολλοι, see Figure 2), it seems even probable.

Figure 2: The remaining ink of the unknown fourth character in ενχ[.]λη of Heb 12:15 (left) compared with an unblemished omicron on the same page of P46. The remaining ink could be the right side of the omicron; note the apparent arcing of what remains of the pen stroke. (Image credits: CC BY-NC. Chester Beatty, Dublin) 

Here, though, one must reckon with James Royse, for he identifies this variant in Hebrews 12:15 as one of the many “Nonsense Singulars” he finds in P46 (2). Royse maintains that an omicron cannot fit into the space between the extant chi and lambda—”If the letter were Ο or Η it is crowded a bit to the left”—and so takes it to be an iota (ενχιλη), nonsense indeed (3).

Comparing occurrences of the chi-omicron letter pair elsewhere in P46 (Table 1), however, demonstrates that the scribe’s positioning of the omicron in relation to the chi is somewhat variable. By reconstructing the chi-omicron in εν χ[.]λη in Heb 12:15, it becomes evident that an original omicron there is not at all far-fetched when compared with the other occurrences.

[εκδε]χομενος, Heb 10:13

υπαρχοντων, Heb 10:34

προσερχομενον, Heb 11:6

εχουσαν, Heb 11:10

εν χολη (reconstructed), Heb 12:15

εχομεν, Heb 12:28

εχουσιν, Heb 13:10

Table 1: Comparing instances of the chi-omicron letter pair in papyrus P46. The image for the reconstructed εν χολη in Heb 12:15 is a composite of εν χ[.]λη and an omicron taken from πολλοι, also in v. 15. (Image credits: CC BY-NC. Chester Beatty, Dublin)

Further, the rounded arc of the extant omicron seems to overlay perfectly with the remnant ink of the unknown character in εν χ[.]λη (Figure 3). 

Figure 3: The remnant ink in εν χ[.]λη matches almost perfectly with the superimposed second omicron in πολλοι (Heb 12:15). Both elements at same scale.(Image credits: CC BY-NC. Chester Beatty, Dublin)

Thus, with the benefits of more modern, digital tools, we can demonstrate that εν χολη is a reading well within the bounds of the physical evidence.

Finally, ἐν χολῇ would fit better with the syntactical structure of Hebrews 12:14–16. Below is a diagram of those three verses (NA28 text, punctuation removed, with line numbers added for ease of reference and explanation), with the finite verbs colored red and the adverbial participles colored blue.


14 Εἰρήνην διώκετε μετὰ πάντων καὶ τὸν ἁγιασμόν


οὗ χωρὶς οὐδεὶς ὄψεται τὸν κύριον


15 ἐπισκοποῦντες 


μή τις ὑστερῶν ἀπὸ τῆς χάριτος τοῦ θεοῦ


μή τις ῥίζα πικρίας ἄνω φύουσα ἐνοχλῇ 


καὶ δι’ αὐτῆς μιανθῶσιν πολλοί


16 μή τις πόρνος ἢ βέβηλος ὡς Ἠσαῦ


ὃς ἀντὶ βρώσεως μιᾶς ἀπέδετο τὰ πρωτοτόκια ἑαυτοῦ

There is one finite verb governing this whole sentence, διώκετε, in line (1), a second person plural imperative meaning “you (pl.) pursue!” Line (2) is a relative clause subordinate to Εἰρήνην . . . καὶ τὸν ἁγιασμόν (“peace and holiness”) in line (1). Line (3) supplies the way in which the audience is to “pursue peace with everyone,” by “taking care” or “observing” or (as the ESV has it) “See[ing] to it.” Lines (4), (5), and (7) all begin similarly, with μή τις (“no one”), revealing the deliberate structure. Line (4) has it’s own adverbial participle, ὑστερῶν (“falling short of”). So far, the audience is to:

Pursue peace with everyone, and the holiness

without which no one will see the Lord

taking care that

no one is falling short of the grace of God

Line (5) introduces a more difficult structure with ἐνοχλῇ and the participle φύουσα. This third person subjunctive, being in an independent clause, may be translated as hortatory, with an imperative sense: let no bitter root, rising up, cause trouble. Line (6) is another relative clause, referring back to the “bitter root” of line (5). Line (7) may be thought to contain an implied ὤν participle: no one being sexually immoral or unholy as Esau. Finally, line (8) is another relative clause describing Esau further for the purposes of this exhortation.

The variant in question, again, occurs in line (5). While this line functions just fine syntactically as it appears in the NA28 edited text, it does break from the pattern of μή τις + participle, a pattern that would be restored were it ἐν χολῇ. Further, the P46 reading is more faithful to the LXX text of Deuteronomy 29:17. Here is the comparison shown before, this time using the text of P46 at that line:

Dt 29:17 (LXX): μή τίς ἐστιν ἐν ὑμῖν ῥίζα ἄνω φύουσα ἐν χολῇ καὶ πικρίᾳ

Heb 12:15 (P46): μη τις ριζα πικριας ανω φυουσα εν χ[ο]λη

It follows that the translation of Hebrews 12:15 in P46 is likewise more faithful to Deuteronomy 29:17: “no root of animosity growing up in bitterness.”

This short post does not go all the way to suggest that the edited text of NA28 be upset, and that the reading of P46 be adopted instead; the papyrus is after all only a single voice. Rather, the goal is that NA28 would drop its label of the P46 reading as “nonsense” or “absurd.” It clearly harmonizes with the text of Deuteronomy in LXX and thus makes perfect sense.


(1) Eberhard Nestle, Erwin Nestle, Barbara Aland, Kurt Aland, Iōan. D. Karavidopoulos, Carlo Maria Martini, Bruce M. Metzger, and Holger Strutwolf, Novum Testamentum Graece, 28th ed. (Stuttgart: Deutsche Bibelgesellschaft, 2012), 881.

(2) James R. Royse, Scribal Habits in Early Greek New Testament Papyri (Leiden: Koninklijke Brill, 2008), 254ff.

(3) Royse, Scribal Habits, 259n332.

94 views3 comments

Recent Posts

See All


I think you're correct here. I went with an omicron in my thesis as well but kept it as one word (oops!) (Peterson, "GA 1739: A Monk, His Manuscript and the Text of Paul's Letters", 492). I went to the CBL to inspect the papyrus and my notes for this reading were: "This is almost certainly an omicron. The character definitely exhibits curvature, eliminating η or ι. Plus, it is smaller and if it is an omicron it would be centered. If it were an η, it’s width would result in weird letter spacing. The decision now is whether it is ενχολη or εν χολη." This reading in P46 and its relation to Deut 29.17 is also discussed in Johannes de Vries,…

Replying to

Welcome! I'm pretty sure I got the 517 reference from Royse as well (he's usually a pretty good guess as a source anyway) Let's hope your suggested correction makes it into the NA29 apparatus!

bottom of page