Voice Markers in Septuagint Greek in the Light of Hebrew Interference: A Corpus‐Based Study on the Aorist System of the Book of Genesis*

In this paper, we examine the behaviour of so-called passive and middle aorist forms in Greek reflected Genesis Septuagint. The Septuagint, Biblical more generally, displays a considerable aberration with respect to other varieties Ancient regarding relative frequency vis-à-vis forms. Here, explore feature Septuagint some detail, showing that there is or less consistent, binary system at stage, where preferred expression monoargumental detransitive clauses, while primarily appears biargumental clauses. A comparison original text Hebrew suggests its binyanim system, systematic distinction between transitive detransitivising stems, may have played key role accelerating process, which was already present earlier stages Greek. This process resulted general ousting middle, essentially became lexically restricted variant active voice. L'articolo esamina il comportamento degli aoristi passivi e medi nella lingua della traduzione greca Bibbia ebraica nota come Settanta. La Settanta – ma anche, più generale, greco biblico presenta una situazione molto diversa rispetto alle altre varietà di antico per quanto riguarda la frequenza relativa ai medi. Questo aspetto verrà indagato qui attraverso uno studio sui dati offerti dal libro Genesi. L'analisi mostra l'esistenza un sistema binario, o meno coerente, cui l'aoristo passivo è forma privilegiata frasi monoargomentali detransitivizzate, mentre medio ricorre principalmente biargomentali detransitivizzate. Il confronto con testo ebraico originale che binyanim, sua sistematica distinzione tra temi verbali transitivi detransitivizzati, può aver svolto ruolo chiave nell'accelerare tendenza, già presente fasi precedenti greca, verso voce binario. processo portò ad sostanziale eliminazione dell'aoristo medio, le forme residuali divennero variante lessicalmente determinata attivo. [Italian] paper deals voice markers Hellenistic Greek, as language translation Bible into made Greek-speaking Egypt from third-century BCE onwards. first five books Bible, Pentateuch (Genesis, Exodus, Numbers, Leviticus Deuteronomy), were translated first, presumably around year 250 (Jobes & Silva 2000: 29), by different translators (Tov 2015: 448). includes book Genesis, constitutes corpus research. For our purposes, important for following reasons. First, it reflects variety Koiné BCE, thus representing an source understanding development Classical later Koiné, reflected, instance, New Testament Second, regards use morphology and, specifically, aorist, shows both sources contemporary sources. Third, given text, question arises whether odd unexpected phenomena be due interference languages. Before proceeding analysis, let us summarize few characteristics system. two aspect (= imperfective) perfective) somewhat systems. present, opposition voice, marked verbal inflection. Verbal inflected occur constructions (both intransitive), broader range valency changing functions, example, passive, anticausative, reflexive (auto-)benefactive well structures subject direct object are linked relationships such possessor/possessee, whole/body part. contrast, stem exhibits ternary active, diathesis. include (a) inflection, namely endings, present; (b) stem-final suffixes -thē-/-ē-, distributed (some verbs take -thē- others -ē-). labels ‘active’, ‘middle’ ‘passive’ commonly used linguistics characterize three sets (cf. Rijksbaron 2002) follow conventional terminology. However, hasten add they mere does not represent category, strictly speaking. As known, aorists unaccusative1 (Benedetti 2005; Tronci 2005) overlap latter function. That say no one-to-one relationship especially aorists. An outline compared Table 1, employing verb lúein ‘to unbind, weaken’ (all quoted 1st person singular line lexicographic practices Greek): lú-ō ‘I I élu-sa unbound, weakened (trans.)’ lú-omai unbind myself / myself’, am unbound’, weaken (intrans.)’ elu-sámēn myself’ elú-thēn boundaries functional values forms, on one hand, clear-cut. One crucial factor morphological class verb's stem. sigmatic most productive aorists, together suffix -thē-. classes is, root thematic productive, so lexical items constitute closed class. They interesting research, however, residual older stage similar Benedetti 2017; 2018 further details). unproductive coexisted although exhibited number items. Some cases overlapping can noticed 2011). attested history language, Archaic via increase types tokens expense Besides their quantitative increase, also expanded range. result, replaced gradually disappeared. substitution concerned occurred unaccusatives reflexives. expansion area did concern 2018; Bruno 2019). picture complicated media tantum deponent verbs, exclusively middles but same syntactic actives Lavidas Papangeli 2007 (2017, 2018, 2020) has previously noted, restructuring processes voices We will back issue Section 2.3. contact Our interest topic arose observations grammars tend note show surprising texts, notably Helbing 1907: 97; Thackeray 1909: 193) Blass et al. 1961: 161). Quantitative data been collected discussed (2018) concerning (2019) piece evidence, calculated diachronic comprising representative texts Homeric (eight-century BCE) (second-century CE).2 Figures 1 2 when taken account respectively.3 difference remarkable calls explanation. illustrate how proportion fluctuates over time.4 need establish differences corpora reflecting each periods statistically significant. presents observed categories authors/periods.5 address distribution shown explained being chance. To clarify this, employed statistical significance tests, chi-squared test Fisher Exact test. Both tests evaluate hypotheses association variables, calculating probability obtaining under null hypothesis variables involved, case construction type. compares expected frequencies, frequencies actual numbers, 2, would obtained if independent. 3 gives data.6 945 [−] 72 641 607 66 216 3001 [+] 748 627 186 1327 101 899 852 93 304 4215 1050 880 261 Here following, employ notation indicate value greater than smaller value. When applied material, yields p-value 2.2e-16, 0.00000000000000022, 1604.2 9 degrees freedom (χ2(9) = 1604.2).7 below level 0.05. Thus, discard variables. provide any information about strength necessary measure effect size, Cramér's V value.8 ranges 0 1; 0.1 conventionally reported because too small. Reportable small (0.1–0.29) moderate (0.3–0.49) large (≥0.5). set 0.306, implying which, robust points. allows conclude significant size. fact makes possible go step patterns attractions repulsions constructions, attraction understood terms higher ([+]) repulsion lower ([−]). end, draw test, cell, calculates even deviation (cf., e.g. Janda 2013: 90–91). Specifically, make two-by-two contingency table cells 2. cell another row sum minus value, third column fourth total sums cells. Homer, 433 6676 second 1839 8144 cell. These numbers then furnish input results 4.9 < 2.2e-16 0.01426 2.895e-11 0.8104 0.9064 4 corroborates impression randomly distributed, though Plutarch what could expect Interestingly, observe all non-Biblical except towards On opposite tendencies sources, significantly attracted markedly rejecting clearly display analogous differ those displayed texts. striking authors, Callimachus Apollonius Rhodes, only lived worked within geographical cultural context, Ptolemaic Alexandria. authors attests exceptional character. widely accepted assumption had strong influence Testament, hardly these parallel specific morphosyntactic domain discussion. why previous tradition period. It tempting suggest peculiarity Semitic languages text. aim study investigate ones depend interference. concerns particular overlap, intransitive structures. translation, grammatical markers. structured follows. discuss data. devoted describing analysing 4, conclusions. research comprises occurrences Genesis. automatically searching electronic resource Thesaurus Linguae Graecae (http://stephanus.tlg.uci.edu/). searched sequences responding query combining features: tense: aorist; voice: passive; (c) mood: indicative, subjunctive, optative imperative (finite moods). found 153 620 occurrences: list them Appendix A. decided limit manageable sufficient hypothesis. Future studies tendencies.10 follows: Out items, 84 exhibit 61 8 aorist. Most part, 30 Deponent preferably hold inflection fact, slightly fewer half (25 items) belong ἐνθυμέομαι [enthuméomai] ‘think, notice’, λογίζομαι [logízomai] ‘count, reckon’ ὀσφραίνομαι [osphraínomai] ‘smell’. 5 summarizes sections, aorists: type (Section 2.2), 2.3) appear 2.4). out analyse relation (see above). behave like never form form, addition former Muraoka 2009 s.u. ἀποκρίνομαι [apokrínomai] πορεύομαι [poreúomai]). since ἀπεκρίνατο occurs rarer verb. indicative degree categories. supported Gospels (2018: 259–260). Comparable remarks apply [poreúomai], (e.g. Ge. 24.42) stem, Euripides, stems. Finally, συναναστρέφω [sunapostréphō] without exception verb, lemmatized (2009 s.u.). occurrence unexpected, usually combine argument, scrutiny here, psychological experiencer external whatever voice.13 explains correlates semantic subject, unaccusative constructions. thumóō ἐθύμωσεν [ethúmōsen] Os. 12.15) mimnḗskō epic poetry Pindar), causative meaning, ‘make angry’ ‘recall memory’ respectively. Unlike enthuméomai corresponding form. peculiar consistently verbs. ‘smell’, corpus, ἐλογίσατο [elogísato] Th. 6.31.5 ὤσφροντο [ṓsphronto] Hdt. 1.80.24).14 Despite inflectional distributional Deponency intriguing ancient linguistics. reference grammars, Kühner Gerth (1892: 324; 1898: 377), several deponency recent years, (2007); Ladewig (2010); Grestenberger (2014); Campbell (2015). do fit opposition, Because cannot evidence investigation Lazzeroni 1990 stance). particularly future stems 2020, Septuagint). many earlier: summarize, (a)–(e) aorists; therefore, prone disappear disappearance witnessed produced developed away 2018: 275). Middle increasingly (together aorists), expand intransitivity, unaccusative, unergative ones. instances lost function full marker express possessor possessee (14a) whole body part (14b) subject-referring pronouns added. redundancy decline Caragounis (2004: 152) points data, Kulikov 2017 increasing labile syntax). configuration common noun phrases clause (19a) (19b), should contain accusative, (‘double accusative’ construction, see 2022 references therein). clauses phrase denoting advancement, maintains accusative case. According (21), seems complementary distribution, (meaning ‘prevail’), modal ‘be able to’) governing infinitive. short, retains features item. time span approximately 3200 attestations twelfth-century BCE. extra-biblical sources.16 generally believed 1200 sometime after 450 1000–500 Three broad distinguished (ca. 1200–1000 BCE), Standard 1000–450 Late second-century CE) Hornkohl 2019: 534 discussion references). characteristic complex now turn. Example (26a) illustrates imperfective/long prefix conjugation characteristically denotes ongoing situation holds speech continue future. (26b), perfective/suffix typically completed located past. noted temporal readings implicatures rather entailments overruled contextual factors. (26c), consecutive imperfect narrative past tense (26c) participle regularly denote speech. 1.1 above, based tense/aspect inventories formed suffixes. BH, structure fundamentally different, voice-oriented affixes examples (26) comprise default labelled qal G-stem. philological distinguishes six (binyanim) valency-affecting distinctions. 6 survey various functions.18 anticausative factitive, intensive, resultative reciprocal Although definitions very simplified suggestive transitivising (piel, hiphil) (niphal, pual, hophal, hithpael). turn brief illustration salient functions binyanim.19 Waltke O'Connor main arguments favour (Waltke 1990: 374). Their argument comparative Arabic, Ugaritic, Tell Amarna letters possibly Aramaic dialects, systemic observation passives sometimes underlie pual hophal piel hiphal attested. alleged ‘asymmetric’ ‘suggest existence Qal semantics forms’ 374).20 Fassberg (2001) notes tendency replaces might facilitated passive. presented exist relics category dubious enough demonstrate stage. We, remain agnostic BH chosen among analysis. above analyzed successive cycles derivation, basic, niphal, piel, hiphil hithpael derived 7 absolute imperfect, perfect stems.21 8. Assuming frequent basic secondarily categories, correspond proposed There exception, hithpael, resembling suggesting category. yielded <2.2e-16, 965.55 12 (χ2(12) =965.55). 0.097, indicating low size.22 random. tense/asp