What Is Jude Instructing in Verse 22?
(avg. read time: 8–17 mins.)
I have not written much about Jude on this Substack. The most extensive post I have about Jude is on its relation to 2 Peter, since that is a matter that is frequently raised and, I contend, addressed in a wrongheaded fashion. Today, I want to illustrate the difficulties that come with dealing with the text of Jude by focusing on one portion of text where there is some contention about what the text is even instructing.
I aim to focus on Jude 22, although this necessitates looking at the immediate context as well. Most English translations take it as a command to have mercy on those who doubt or are wavering (ASV, AMP, CSB, CEB, CEV, DLNT, ERV, EASY, EHV, ESV, EXB, GW, GNT, HCSB, ICB, ISV, LSB, LEB, TLB, MSG, Mounce, NOG, NABRE, NASB, NCB, NCV, NET, NIRV, NIV, NJB, NLV, NLT, NRSV, NTFE, OJB, TLV). Others think the key participle refers to “disputing” while taking the verb as “rebuke” or “convince” (CJB, JB, RSV). Still others have an entirely different rendering as “And on some have compassion, making a difference” (KJ21, BRG, Darby, KJV, NKJV, RGT, WEB; cf. GNV, Phillips). Yet other more peculiar translations call to “reprove” some who are “being judged” (DRA) or to “be kind” while “judging thoroughly” (YLT).
Some of the aforementioned diversity concerns difficulties for the translators, while other aspects stem from the difficulties concerning the text itself. Jude appears to have had a peculiar transmission history, and some parts take on a great variety of forms in manuscript traditions. But of all the textual difficulties in Jude, the text of vv. 22–23 may present the most difficult problem. Tommy Wasserman has done a thorough collation of the Greek manuscripts of Jude and has noted the following readings for v. 22:1
και ους μεν ελεατε διακρινομενους (supported by fourteen witnesses)
και ους μεν ελεειτε διακρινομενους (supported by six witnesses)
και ους μεν ελεειτε διακρινομενοι (supported by 475 witnesses)
και ους μεν ελεγχετε διακρινομενους (supported by forty-four witnesses)
και ους μεν ελεγχετε διακρινομενοι (supported by five witnesses)
και ους μεν ελεγετε διακρινομενους (supported by one witness)
και ους δε ελεειτε διακρινομενοι (supported by one witness)
και ους ελεειτε διακρινομενοι (supported by two witnesses)
Two other witnesses are difficult to categorize precisely, but they favor the nominative plural form of the participle
The earliest extant manuscript (p72) presents what looks like a merged form of vv. 22–23 in other texts: ους μεν εκ πυρος αρπασατε διακρινομενους δε ελεειτε εν φοβω
The first reading is one favored by Wasserman, the Editio Critica Maior, and the Nestle-Aland and United Bible Society texts that provide the bases for most translations these days. It has the most distinct reading of the main verb, while multiple other lines of evidence favor the rendering of the participle as the accusative plural and thus as the direct object of the verb. The second reading supports the participle of the first reading, but it follows the most common rendering of the verb. But for all of that, it is favored by slightly over one percent of the witnesses, as it tends to be recognized as something of a composite reading. The third reading is by far the most common, since most Byzantine texts favor it, but the diversity of its witnesses beyond the vast majority of Byzantine texts is limited. It is the basis of the KJV and similar translations, where the participle refers to the subject of the action—the addressees themselves—rather than the object of the action. The fourth reading is often considered today to be a later adjustment of this text, but it has considerable witness support, and it has the support of Samuel Tregelles’s text, which is followed by the Tyndale House Greek New Testament (as well as La Bible de Jérusalem, the Wycliffe Bible, CJB, DRA, JB, RSV, Tischendorf’s text, and several others among older textual critics and commentators).2 Of course, the verb is completely different in sense. The first and last syllables are the same as the different forms of the most common verb, but the middle two syllables are remarkably different. The other readings are so sparsely attested as to have little to recommend them. Still, because one reading is represented by the earliest extant manuscript, some scholars ultimately support it or something like it (in terms of contents and two-clause structure with v. 23) as the original.
Before we get into more arguments for the readings in this verse, we need to focus on the participle before we address which verb best fits the context. Beyond the question of nominal case, this is more of a translation issue than a textual issue, but we have already noted that most translations render the sense of the participle as “doubters/those who doubt” or “those who waver.” But this is derived from what is commonly thought to be a special NT sense of the middle form of διακρίνω that is the extension of the idea of “disputing/contending with(in) oneself.” Along with Jude 22, this special NT sense is claimed, including by standard lexica like BDF and LSJ as well as many commentators, to be represented in the texts of Matt 21:21 // Mark 11:23; Acts 10:20; Rom 4:20; 14:23; and Jas 1:6.
It is intuitively improbable that such a special sense would be restricted to this one corpus. That does not mean it is impossible, but there should be a strong evidentiary basis from specific texts to demonstrate such a peculiar sense. The problem is that the strong evidentiary basis for this use is lacking in the NT. Its senses outside of the NT, and in cases within the NT, include reference to judging, discerning, (making) distinction, separating, disputing, and contending. Most significantly, Jude 9 uses one of these typical senses of the middle form as well, and there is no particularly strong reason to think the sense has shifted not long thereafter. Alternatively, if one takes this form as a passive, it could signify being subject to such judgment activity as conveyed by this verb.
In other words, if the accusative plural is original, as multiple streams of textual transmission suggest and as would fit with the relative pronoun, the referent is those who dispute and cause division among the community by steering them away from the truth. This would be another way for Jude to refer to those he has been criticizing throughout the letter, as opposed to writing about a new group altogether and one time only at that.3
And now let us look at the readings for v. 23, where the greatest variation occurs prior to “hating.” Wasserman records thirty-four different readings (besides the peculiar situation of p72), which mostly vary by spelling differences, arrangement, or repetition/omission of clauses or phrases. We will focus here only on those with ten or more witnesses:4
ους δε σωζετε εκ πυρος αρπαζοντες ους δε ελεατε εν φοβω (supported by thirty witnesses)
ους δε σωζετε εκ πυρος αρπαζοντες ους δε ελεειτε εν φοβω (supported by ten witnesses)
ους δε σωζετε εκ πυρος αρπαζοντες ους δε ελεγχετε εν φοβω (supported by eighteen witnesses)
ους δε εν φοβω σωζετε εκ πυρος αρπαζοντες ους δε ελεγχετε εν φοβω (supported by twenty-four witnesses)
ους δε εν φοβω σωζετε εκ πυρος αρπαζοντες (supported by 353 witnesses)
ους δε εν φοβω σωζετε εκ του πυρος αρπαζοντες (supported by sixty-seven witnesses)
ους δε σωζετε εκ πυρος αρπαζοντες (supported by ten witnesses)
The majority is once again favored by the KJV and the others noted for supporting the majority reading of v. 22. The other translations tend to maintain the same basic sense in their translations as represented by the first two readings (where the last verb is simply spelled in different ways with the same sense, as the verb could be spelled as either ἐλεάω or ἐλεέω). I have yet to find a translation that supports a text with ἐλέγχετε in the position of the third verb.
We have seen already how p72 presents what looks like a merger of vv. 22–23 (ους μεν εκ πυρος αρπασατε διακρινομενους δε ελεειτε εν φοβω). That text broadly fits what we find in Clement of Alexandria (και ους μεν εκ πυρος αρπαζετε διακρινομενους δε ελεατε; Strom. 6.8.65.4). Both are our earliest witnesses to the text of Jude, and they attest to a two-clause form, as opposed to the three-clause form in most translations today. This is also the overall most widely attested structure, as supported by the vast majority of the Byzantine texts. However, the earliest codices, including Sinaiticus and Vaticanus, feature the three-clause structure. Scholars have carried on in arguing for two-clause and three-clause forms.5 I am personally convinced by the arguments that the three-clause form is original, not least because it is easier to explain how the two-clause structure arose on the assumption of the originality of the three-clause structure than the other way around. The three-clause structure is also befitting of what we can see of Jude’s style (limited though the extent of text is), and the two-clause form arose from difficulties in distinguishing three statements and who they were concerning (including in dealing with various transmission issues involving repetitious verbs, repetitious sounds of verbs, and repetitious structure that sometimes became overly elaborate).6 The simplification in structure thus became normalized.
But something still must be said about the verbs. Nearly all witnesses (except for the earliest extant ones with their merged readings) agree on the first verb of v. 23, which is the second verb of the sentence of vv. 22–23. But there is significant variation with the last verb. Some drop it altogether, as with the two-clause constructions. Much of the variation among three-clause texts is due to the spelling, which is not a major issue. But still others vary the same way here as they do with the verb from v. 22. Dozens of texts feature ἐλέγχετε (which has senses of rebuking, refuting, reproving, and so on) as either the first verb or the third verb. Interestingly, however, with the possible exception of a lectionary from the fifteenth century (l 1281),7 not a single Greek text of the dozens to feature the verb in some capacity repeats the verb in both cases. It appears in Alexandrinus and in two other uncials (C/04* and 0316, apparently), as well as many minuscules (including but not limited to 6, 33, 93, 326, 1611, 1739, 1836, 1837, 1875, and 1881). Additionally, the first appearance of the verb has support from the Vulgate and Bohairic texts (according to Tregelles, it is also supported by Armenian and Ethiopic texts). I have not been able to find many references to this part of Jude in early and medieval texts, but of what I could confirm, this reading is attested in Ephrem Graecus, Cassiodorus, Bede, Oecumenius, and Theophylact.8 In any case, it appears to have had a significant ancient basis, even if that is not as well reflected by the sheer number of extant witnesses in Greek.
Similarly, among extant three-clause texts, I have not found many that repeat verbs in the first and third clauses. The spelling of ἐλεᾶτε appears repeated in both cases in only three manuscripts, but two of them are Sinaiticus and Vaticanus (the other is 044), and those manuscripts tend to weigh heavily for many text critics, meaning that critical texts tend to print the text as repeating this term in this spelling in both cases. If it was not these two manuscripts specifically that featured this repetition, I doubt so much credence would be given to this repetition surrounding the more ubiquitously common second verb in the three-clause structure. Only 385 and 639 (dated to the fifteenth and eleventh centuries respectively) repeat the spelling of ελεειτε in a curiously elongated reading that also includes ἐλέγχετε (giving them four verbs instead of two or three).
Otherwise, texts with the three-clause format have a proper threefold imperative structure. Most of this subset of texts (nearly a hundred) attests to ἐλέγχετε in either the first or third clause. The confusion on placement probably has a common root cause as the dittography of ἐλεᾶτε in the aforementioned manuscripts and the confusion in most three-clause manuscripts of where to place the call for mercy. That is, from v. 21 to v. 23 there are three words that begin the same way, two of which also end the same way, and which occupy similar positions in their respective clauses. Verse 21 refers to the ἔλεος (“mercy”) of our Lord Jesus Christ in the fourth place of the subordinate clause beginning with the participle προδεχόμενοι. Verse 22 then features a verb that is a four-syllable word in nearly all Greek witnesses written ἐλε…τε, which occupies the fourth place in its clause. While the precise number of words by which this is separated from the previous word depends on the manuscript, there are at least eleven words consisting of forty-eight letters (fewer with nomina sacra) in between these two. Verse 23 then features a verb that is a four-syllable word in nearly all Greek witnesses attesting a three-clause structure that is written ἐλε…τε, which occupies the third place in its clause. In manuscripts representing the three-clause tradition, there are at least nine words consisting of around forty-seven letters in between these two. Similarities like this have caused numerous textual variations of many kinds, including haplography (as I think happened at the roots of the two-clause manuscripts early in the transmission process) and dittography (as in the few extant manuscripts that repeat verbs), and so it would be unsurprising for such problems to emerge here as well. Moreover, since these words were written in continuous script and in all capitals, this is what they would look like in nearby lines with the two previous words (and more precise handwriting) in an exemplar if my claim about the original text is correct:
ΟΥΣΜΕΝΕΛΕΓΧΕΤΕ...
ΟΥΣΔΕΕΛΕΑΤΕ… or
ΟΥΣΔΕΕΛΕΕΙΤΕ…
Since the verb has attestation in multiple locations, why prefer the first position as opposed to the third position in this three-clause structure? While there are many manuscripts that attest to the third position, none of them are from the earlier manuscripts, and many others of these witnesses can be shown to more faithfully transmit ancient texts among the minuscules. The position as the third verb appears to be a secondary development of the aforementioned mix-up and the need to include the verb somewhere to maintain the threefold structure.
With this, we can begin getting a clearer view of Jude’s instruction in v. 22. It does not concern doubters who have been influenced by the teachers among those Jude has been criticizing. Nor is the first clause a call to show them mercy. Rather, he calls for them to rebuke, refute, or reprove (ἐλέγχετε) the disputers, those who cause division.
This fits Jude’s purpose in the letter to exhort his audience “to contend earnestly for the faith that was once and for all handed down to the saints” (3). Amidst comparisons of the situation and the opponents who have infiltrated them (4) to various scriptural stories, he notes how Michael responded to the devil he was disputing, when he dared not pronounce a reviling judgment (in contrast to the opponents, per vv. 8 and 10), leaving the matter to God, but said, “May the Lord rebuke you” (9; not the same term as the one we are focusing on). For “these ones” (7, 8, 10, 12, 14, 16, 19), though, there is the promise of finding themselves on the wrong side of God’s promise of judgment, since the Lord promises that he will come to execute judgment and to “convict/rebuke” (ἐλέγχω) every life concerning their deeds (15). Despite all their works in causing division among the community of faith, Jude reminds them, “But you, beloved, remember the words that were spoken beforehand by the apostles of our Lord Jesus Christ that they would say to you, ‘in the last times there will be mockers going according to their own ungodly desires. These are ones who cause divisions, soul-driven [or “given to mortal life”], because they do not have the Spirit’” (17–19). Together with the promise of judgment, there is also the promise of mercy, as Jude reminds his audience in the subsequent text: “But you, beloved, by building yourselves upon your most holy faith, praying in the Holy Spirit, keep yourselves in the love of God, while waiting eagerly for the mercy of our Lord Jesus Christ for everlasting life” (20–21).
It is in this context that Jude then uses the relative pronoun to point back to “these ones” who were other than those he is addressing directly. That relative pronoun in v. 22 is a match in gender, case, and number (masculine accusative plural) for the key participle in v. 22. In the first clause of the three-clause structure extending to the end of v. 23, Jude says here, “And rebuke them who dispute.” We can thus read v. 23, using the same relative pronoun combined with conjunction markers that signify subdivision (though in a way similar to v. 8, where the second and third members of the structure are practically synonymous), to say, “but save others [of them] by snatching them out of the fire, and on others [of them] have mercy in/with fear, while hating even the tunic that has been polluted by the flesh.” The saints are thus called upon to imitate God’s action in ways that are appropriate for them while awaiting the fulfillment of their eschatological hopes that their own actions anticipate and (by extension) participate in, including in rebuking/convicting, saving from a path of destruction, and showing mercy with due reverence for God. This is how they show love to the person while showing hate for the stains of sin.
This part of the text is often noted for its allusions to Zech 3:1–5. The allusions further uphold the framing of imitating divine action. But there are other texts that should be noted as well. The verb of v. 22 is likewise used in reference to rebuking one’s neighbor in the LXX of Lev 19:17, which also warns about incurring sin because of the neighbor, and that instruction is then followed up with the call to love one’s neighbor as oneself (19:18). Similarly, the LXX of Prov 10:10 has an additional clause compared to the MT, wherein we are told that the one who rebukes with boldness makes peace.
Among NT texts, the key word appears in instructions about rebuking brothers and sisters who have sinned in Matt 18:15, the Spirit’s work in convicting the world concerning sin, righteousness/justice, and judgment in John 16:8, exposing/rebuking the deeds of darkness rather than taking part in them in Eph 5:11 and 13, rebuking those who persist in sin so that others may have proper fear in 1 Tim 5:20. This is also part of the instruction to Titus on multiple occasions (1:9, 13: 2:15). And we are also reminded on multiple occasions that God rebukes and disciplines those he loves (Heb 12:5; Rev 3:19; cf. Prov 3:11–12). As we have seen from J. R. R. Tolkien in words I have cited on multiple occasions in commentary on his works, “A divine ‘punishment’ is also a divine ‘gift’, if accepted, since its object is ultimate blessing, and the supreme inventiveness of the Creator will make ‘punishments’ (that is changes of design) produce a good not otherwise to be attained” (Letter #212).
The instruction to rebuke those who dispute and cause division against the faith is thus of one piece with a broader biblical theme about rebuke. What Jude instructed his audience to do still applies to us in terms of rebuking those who would lead astray from the faith and fracture the community of faith in the manner that these teachers did. Such a practice of rebuke is part of a larger aim to save others from destruction as if by snatching them from the fire, whether it is those rebuked or others who witness the rebuke (as especially in the references from the Pastoral letters). And as the Lord has mercy, so we too are to show mercy so as to encourage repentance without excusing, justifying, or accepting sin. By such means we become participants in divine action by imitating how God acts. This is one of many ways by which we can fulfill our created purpose to be bearers of God’s image and likeness.
Tommy Wasserman, The Epistle of Jude: Its Text and Transmission, ConBNT 43 (Stockholm: Almqvist & Wiksell International, 2006), 196–97.
Charles Bigg, A Critical and Exegetical Commentary on the Epistles of St. Peter and St. Jude, 2nd ed., ICC (Edinburgh: T&T Clark; New York: Scribner’s Sons, 1902), 341.
On the translation issue, see Peter Spitaler, “Doubt or Dispute (Jude 9 and 22-23): Rereading a Special New Testament Meaning Through the Lense of Internal Evidence,” Bib 87 (2006): 201–22; Spitaler, “Διακρίνεσθαι in Mt. 21:21, Mk. 11:23, Acts 10:20, Rom. 4:20, 14:23, Jas. 1:6, and Jude 22—the ‘Semantic Shift’ That Went Unnoticed by Patristic Authors,” NovT 49 (2007): 1–39.
Wasserman, Jude, 198–99.
For more on this subject, see Sakae Kubo, “Jude 22-23: Two-division Form or Three?” in New Testament Textual Criticism: Its Significance for Exegesis, Essays in Honour of Bruce M. Metzger, ed. Eldon J. Epp and Gordon D. Fee (Oxford: Oxford University Press, 1981), 239–53.
See also Wasserman, Jude, 320–25.
Wasserman (ibid., 198–99) indicates that the readings appear in different locations in the same lectionary.
It should be noted that Theophylact copied Oecumenius on this front and others concerning Acts and the Catholic Epistles. For Ephrem Graecus, you can access a copy of the Greek text here with the relevant text being located on p. 221 (derived from vol. 5 of K. G. Phrantzolas’s collection of these works of Ephrem’s). I tend to use as a baseline for patristic references the database of BiblIndex. However, I do not know how to get it to work for single-chapter books like Jude. Nor could I easily find any surveys on the patristic use of Jude that would help me track down citations. As it is, I saw several older commentaries reference a possible allusion to this reading in Can. ap. 6.4, but I have been unable to track down this precise reference in its original context because it does not match any text I can find. The text they quote is an expanded form of a text originally derived from the Didache, specifically Did. 2:7, but the specific form of the text of the Didache at most bears a passing resemblance to the structure and only some of the language of Jude without being a clear allusion to Jude.