* I am indebted to Professors K. Stendahl and D. J. Harrington S. J. for certain observations which have enriched the substance and improved the style of the following argument.

¹ This ostracon was first published by Naveh, J., A Hebrew Letter from the Seventh Century B.C., IEJ 10 (1960), 129–39Google Scholar and plate 17. See also, for the improved text we cite, the readings in Cross, F. M. Jr., Epigraphic Notes on Hebrew Documents of the Eighth-Sixth Century B.C.: II. The Murabba‘ât papyrus and the letter found near Yabneh-Yam, BASOR 165 (Feb. 1962), 34ff., esp. 42–46Google Scholar (an article apparently unknown to Berger) ; the reading in capitals represents a certain improvement by Cross. Of more recent editors, Gibson, J. C. L., in his Textbook of Syrian Semitic Inscriptions, 1 (Oxford: Clarendon Press) 1971, pp. 26–30Google Scholar , follows Naveh on the point under discussion, by reading a “responsorial” Amen; while Donner, H. and Röllig, O., in their Kanaanäische und Aramäische Inschrijten, 2² (Wiesbaden: Harrassowitz) 1968, pp. 199–201Google Scholar , feel forced to emend, with Albright, to 'm nqty.

² This postulate, that a prepositive 'āmēn is not part of normal Hebrew speech patterns, and could not be “naturally” part of Jesus' own way of speaking, is challenged by none of the reviews known to me: cf. on Berger, Hasler, V., ThZ 28 (1972) 362–64Google Scholar , Kearns, R., JBL 91 (1972), 267–68Google Scholar , Thrall”, M.E., JTS NS 23 (1972), 190–91CrossRefGoogle Scholar , Richter, G., Biblica 53 (1972), 290–93Google Scholar, as well as the reviews listed in Elenchus Bibliographicus Biblicus 53 (1972), § 3274Google Scholar ; and on Hasler, Kjeseth, P. L., JBL 89 (1970), 127–28Google Scholar , Greig, J. C. G., JTS NS 21 (1970), 161–64.CrossRefGoogle Scholar

³ To help the reader assess the plausibility of the various positions discussed, it may be useful for him to see the whole letter, and the context in which this use of 'mn occurs; so we give here the complete translation of F. M. Cross Jr. (cf. n.i above). “Let my lord the commander hear the word of his servant. As for thy servant, thy servant was harvesting in ῖaṣar 'Asam; and thy servant harvested, and took measure and stored (the grain) according to regular practice. Before Sabbath, when thy servant had measured his harvest and had stored (the grain) according to regular practice, Hoshaiah the son of Shobay came and took the garment of thy servant; when I had measured this harvest of mine in the regular way, he took the garment of thy servant. And all my brethren will testify for me, those who were harvesting with me in the heat of the sun; all my brethren will testify for me. Truly I am innocent of (any) gu[ilt. Pray return] my garment. And if not, it is (still) incumbent on the commander to retur[n the garment of thy servant; so let him gra]nt to him mer[cy …] thy servant, but do not drive him away.”

⁴ Presumably the following ω'm l' is to be understood “and if I am not guiltless, even so …” rather than “and even if they don't testify for me (or say ‘amen’ after me), even so … ”

⁵ Similar repetitions are found elsewhere in the ostracon — cf. especially the preceding sentence about Hoshaiah. Berger also objects that the word ‘nh, developing from its original meaning “to answer,” became used as a terminus technicus of legal language, which here does not mean “to answer X” (or “to answer on behalf of/for X” (?)) but “besagt dann jedes Vorbringen vor Gericht überhaupt.” For this sense Berger gives two examples with the preposition l, Ezek. 14:4, 7, but both examples are niph'als, and pretty incomprehensible at that. The remaining cases he cites occur with the accusative or with the preposition b, and the verb with this preposition usually (I Sam. 12:2 pace Berger, CD ix 7) means “to testify against someone”, and not “on his behalf” (so only in Gen. 30:33).

These observations of Berger in any case seem to me not to contribute much to confirming his own and Naveh's alternative understanding against that of Cross. Both Cross and Naveh take y‘nw to mean “will testify on my behalf” — and Berger's own translation “antworten mir,” if intentionally different, presents its own difficulties. As a matter of fact, BH usage does not attest ‘nh l “to witness on, behalf of someone”, although it surely lies within the semantic field of the verb and preposition. But we only have one case, Gen 30 33 where this meaning is found at all, there expressed by ‘nh b. Again, surprisingly, only once does ‘nh l mean “to answer someone”, in Ex. 15:21 — the standard construction is with ’t or a verbal suffix. Thus, ‘nh l would have only one Biblical parallel, Ex. 15:21, in the sense “antworten mir”; confessedly it has none at all in the equally natural-seeming sense “answer for me/on my behalf”. In brief, how can we know how one would in Classical Hebrew, using ‘nh, say “to testify on my behalf”? Yet surely the expression must have existed.

⁶ This is not a new discovery of mine, but a fact which needs to be made known to New Testament scholars. In an article “Amen as an Introductory Oath Introductory Oath Formula” written in a journal of Old Testament text criticism little read by N.T. scholars, Textus VII (Jerusalem: Magnes Press, 1969), 124–29Google Scholar , S. Talmon points to the existence of this locution not only in the Yabneh-Yam letter (an en passant in the New Testament) but also makes a strong case for the presence in the Hebrew Urschrift of the LXX at Jer. 15:11 of this variant, i.e., the reading 'mn yhwh rather than the MT's 'mr yhwh (readings which would in some second and first century B.C. hands be easily confused) and also for its superiority. He also defends its plausibility in Is. 25:1, Jer. 3:18 (19), Jer. 28:6 (unless this be a sarcastic liturgical 'āmēn,) and possibly in Is. 65:16, as well as (clearly) in mSoṭah 2:5 passim. In Rabbinic literature we find the affirmatory prepositive 'āmēn also in Tos. Soṭah 2:2, Tos. B.Q. 8:3, Tos. Sheb 15:8, 16:2, 20:16, and in bT Berakot 45a, bT Soṭah 17a. The present writer would add that it seems clearly attested in the Hebrew text underlying the LXX at III Kingdoms 18:39. (Parenthetically one might note that the simultaneously prepositive and postpositive āmên, which replies to a doxology but continues with its own statement, such as is found in Rev. 7:12 and 19:4, can be paralleled in Num R. 4 (142 d), cf. Strack-Billerbeck, III, 461).

Even if the above interpretation of the ostracon and of Jeremiah be correct, the question might be asked, why did the formula apparently disappear for six centuries, only to reappear with Jesus. I do not however feel this to constitute an objection to our explanation. Our knowledge of Hebrew in the intervening centuries is so partial that conclusions, from the silence of our rare documents to the infrequency or desuetude of certain locutions, are methodologically very dangerous. There are plenty of words only found in pre-exilic literature and then again in Mishnaic, for which it is not necessary to postulate a literary archaising revival, under Biblical influence, in the 2nd century A.D.!

⁷ In fact the peculiar element in the Dominical usage is the combination of the two independently well attested phrases a) prepositive āmên and b) legô hūmīn hoti.

⁸ Its continuation in later Christian apocalyptic tradition may indeed be derived from the exalted status that the phrase has, being recognized as once having been used on the mouth of the now Exalted Lord. If our thesis be accepted, we have not removed the necessity for traditio-historical research on these sayings; we have merely removed the specious start for such research which defines the origin of such sayings by postulating a non-Hebrew but bilingual (!) play on words originating in Hellenistic Judaism and surviving only in Hellenistic Jewish Christianity. But if the usage is normal Hebrew, the earlier specimens at least could be a remnant, of the ipsissima vox Jesu, even though one can demonstrate textually that this locution, for the Church characteristically Dominical, spread later to new sayings in the Synoptic tradition that originally didn't have the element (cf. Hasler, op. cit., 113–115) — the phenomenon is after all quite normal. So another traditio-redactio-critical thesis, built upon a more solid foundation than this illjudged hypothetical location, remains to be written. However it should be stressed that these criticisms do not affect Berger's discussion of later Christian apocryphal use; indeed his turning of his attention ever more fully to that neglected field (cf. JStJud 4 (1973)1–18) is greatly to be welcomed.