346 DAVID ARMITAGE overrides the indicative default) does create particular problems for “ literal ” translations which purport to work on a “word for word” basis. Indeed if Tov is right to claim that the Septuagint translation of the Psalms was “slavishly literal” 48, this might account for the apparent absence of translations based on the precative qatal in the Septuagint Psalms 49. Furthermore it might be tentatively suggested that ignorance of, or disdain for, the precative option on the part of the Septuagint translators could have contributed to exegetical emendation of Ps 22,22, if they were, like numerous modern commentators, uneasy with a perfect interpretation of yntyi"ˆ. i : n[ The final word in Ps 22,22 is thus best understood as a pre- cative perfect derived from hn[: “rescue me” 50. It retains the reading which is actually attested in the Hebrew text, whilst the text in the Septuagint must be derived from a Hebrew Vorlage that can at best be tentatively reconstructed and which might have had no independent existence outside the mind of an exegetically cre- ative translator. Given the possibility of a modal nuance there need be no intrinsic awkwardness to the MT reading, and understanding it as a reiteration of the preceding plea for help is both stylistically and contextually effective. Department of Theology and David J. ARMITAGE Religious Studies University of Nottingham NG7 2RD, United Kingdom Quoted in VILLANUEVA, Uncertainty, 87. 48 BUTTENWIESER, The Psalms, Chronologically Treated, see index, cites 49 numerous examples of precative qatal verbs in the Psalms. Almost without exception these are rendered as aorist indicatives in the Septuagint, apart from a couple of future indicatives (55,19; 65,12). The examples proposed by H. IRWIN, “Syntax and Style in Isaiah 26”, CBQ 41 (1979) 240-261, here 249, in Isa 26,15 are translated as imperatives in the Septuagint, suggesting that this use was not completely unknown. Alternatively (reading μymr yerQm as standing by ellipsis in relation to 50 i e n ˆ" i yi[yv/h) : “Save me from the mouth of the lion and from the horns of the wild ne i oxen – Answer!” This omits the pronoun from yntyi"ˆ but retains the semantic i : n[ ambiguity of the qatal form. “Answer!” like yntyi"ˆ, is best read as plea, but it i : n[ can (also like yntyi"ˆ) be read as announcing an answer already received. i : n[