The following is from Bruce Chilton, A Feast of Meanings: Eucharist Theologies from Jesus through Johannine Circles (Supplements to Novum Testamentum LXXII; Leiden: Brill, 1994), 169-71:
THE
CONSTRUCTION ΟΥ ΜΗ . . .ΕΩΣ
[ΑΝ] IN THE ASSEVERATIONS OF JESUS
In Matthew 5:26//Luke 12:59, the
warning that you will not leave the prison until you have paid your last penny
is no promise of departure after payment, but an injunction to avoid Roman
litigation (cf. Matthew 5:25; Luke 12:57, 58), which might sot you both your
wealth and your freedom. The statement in Matthew 23:39 and Luke 13;35, “You
will not see me until you say, Blessed is he who comes in the name of the Lord,”
is manifestly odd. Within the narrative context, it is evident that Jesus will
be seen in Jerusalem; within the context of early Christian eschatology, it
is equally plain that joining in Psalm 118 will not effect an appearance
of the son of man. In other words, the apparently causal relationship between
the phrases breaks down on closer inspection. Their linkage is rather in their
assertion together: you will one day join in saying the (eschatological)
Hallel, and I will not show my face in Jerusalem. So understood, Matthew
26:29//Mark 14:25//Luke 22:18 would mean both that Jesus envisaged drinking new
wine in the feast of the kingdom, and that he bound himself not to drink wine
which was not associated with the celebration of the kingdom.
The passages which have been cited
are adduced by Beyer as among those in which the emphasis must surely be taken
to fall on what comes after “until,” [1] and that it is an important insight.
It is equally important to recognize that the usage of congruent with that of
the Septuagint. Genesis 28:15 . . . instances the usage of ου μη . . . εως
for לא-עד, as one might expect, and somewhat similar cases are presented in
Genesis 19;22; 29:8; Numbers 23:24b; Ezra 4:21; Nehemiah 13:19. In fact,
however, the last usages cited do not repeat the pattern of ου μη with εως
[αν]; the redundant negation does not
appear, and the temporal particle varies (between εως in Genesis
19:22: 29:8; Numbers 23:24b through εως οπισω in 2 Esdras 23:19 and οπως in 2 Esdras 4:21). In most of
those cases, moreover, an evidently temporal and causal relationship between
the phrases does obtain. The single exception is that rule might be
instructive. Ezra 4:21 in the Aramaic reads:
וקריתא דך לא תתבנא עד־מני טעמא יתשׂם
The Septuagint at 2 Esdras 4:21
gives us:
και η πολις εκεινη ουκ οικοδομηθησεται ετι, οπως απο τηε γνωμης.
The translators clearly evidence
an awareness that לא-עד need not be taken in a temporal sense. [2] Unfortunately,
Beyer does not distinguish between temporal and non-temporal usages of the construction,
and even conflates the usage of אם-עד/ει-εως (in Psalm 132:3-5) within his
list. But the evidence of the Septuagint appears clear: לא-עד need not be taken
in a temporal sense, and ου μη-εως might
be the rendering when it is not to be so taken.
Beyer’s conflation of usages, and
of the same basic usage with distinctive meanings, obscures a crucial feature
of the Septuagintal meaning of ου μη-εως. Not only, as we have seen, may
the construction be used to convey a non-temporal usage of לא-עד; in Isaiah
55:11, the construction is introduced in order to render לא-כי אם, and
the meaning is clearly a temporal, an insistence that God’s word will not
return to him empty, but accomplish what he desires. Just as in the cases of
Jesus’ sayings, the phrases coordinated by ου μη and εως amount to
vigorous assertions in tandem: God’s word will perform what he desires and will
not return empty to him (ever). The usage of the construction in the Septuagint
demonstrates that it may be used without a temporal reference, and even that a
translator might introduce the usage of ου μη-εως simply in order to render two coordinated assertions, one
positive and one negative.
Eyer observes that the usage of “until”
is not to be pressed temporally:
Jedoch bezeichnet "bis"
auch nach einer Negation im Sem, und im Griech, öfters nur die Grenze,
innerhalb derer die Haupthandlung betrachtet wird, ohne dass damit gesagt sein
soll, dass sich danach etwas ändert . . .(RB: English: However, even after a
negation in Sem, and in Greek, "until" often only denotes the
boundary within which the main action is considered, without implying that
anything changes afterwards. . .)
He then instances Genesis 28:15;
Isaiah 22:14; 42:14, among other passages, the second of his examples is an
instance of the type we have noted above, in that ουκ-εως renders אם-עד. The cognate examples from the Synoptics he proposes include
16:28//Mark 9:11//Luke 9:27; [3] Matthew 24:34/Luke 21:32; Matthew 5:18, 10:23.
[4] He is correct in insisting that the construction opens the possibility that
the intended meaning simply involves assertion (by means of the clause following
εως) and negation (by means of the clause following ου μη), without the postulation of a temporal relationship between the
two.
Other considerations will also
need to be brought to bear, to determine whether or to what extent a temporal
aspect is operative within the usage. It is of not, for example, that the
Markan parallel of Matthew 24:34//Luke 21:32 employs, not ου μη-εως, but ου μη-μεχρις ου (Mark 13:30), an emphatically temporal construction. Similar
usages of the relative pronoun but with εως) are associated with vigorously temporal assertions in Acts
23:12, 14, 21. [5] By that criterion, the usage in Luke 22:16, ου μη-εως οτου, and in v. 18, ου μη-εως ου, should be taken as in the nature of vows of abstinence . . . [6]
[1] Kalus Beyer, Semitische
Syntax im Neuen Testament: Studien zur Umwelt des Neuen Testaments 1 (Gōttingen:
Vandenhoeck und Ruprecht, 1962), 132, 133 (n. 1). There are other examples
cited from the Synoptics, but they do not belong to the category of
asseverations. They are simply assertions (cf. Matthew 24:39) or imperatives
(cf. Matthew 17:9); Beyer’s work may be consulted for other instances outside
the Synoptics.
[2] It may be of interest that, in
the appearance of the idiom in Leviticus Rabbah 24:3, the Aramaic reads
specific reference to time (זמן) in order to assure the temporal sense of the
construction.
[3] Beyer’s observation occasioned
my exegesis of the passage, on the understanding that those who would not taste
death were figures such as Moses and Elijah, who stood as guarantors that God’s
kingdom will come in power. Cf. God in Strength. Jesus’ Announcement of the
Kingdom. Studien zum Neuen Testament und seiner Umwelt 1 (Freistadt: Piōchl,
1979, reprinted within “The Biblical Seminar;” Sheffield: Sheffield Academic
Press, 1987); “The Transfiguration: Dominical Assurance and Apostolic Vision,” New
Testament Studies 27 (1980): 115-124.
[4] Beyer also cites Matthew 1:25;
1 Corinthians 4:5, but the formal hallmarks of the dominical asservation are
not present, the sayings are not attributed to Jesus, and the meaning of the
verses may in fact include a temporal dimension.
[5] John 13:38 marches in the
opposite direction, by employing ου μη-εως when the Synoptics do not (cf.
Matthew 26:34; Mark 14:30; Luke 22:34, the last usage being the closest
approximation to John’s). John perhaps represents a diminished sensibility of
the meaning of the locution.
[6] The sense is similar to Luke
1:26, although the construction is different.