Hostname: page-component-848d4c4894-v5vhk Total loading time: 0 Render date: 2024-06-13T15:00:12.120Z Has data issue: false hasContentIssue false

’Ενευλογηθήσονται as a Speech Action Middle in Genesis 12:3b LXX

Published online by Cambridge University Press:  29 January 2024

Dan York*
Durham University;
Rights & Permissions [Opens in a new window]


Interpreters unanimously read ἐνευλογηθήσονται in Gen 12:3b LXX as a passive. Good evidence, however, exists to challenge and problematize this conclusion. Recent linguistic studies on the ancient Greek middle voice reveal that aorist and future -θη- forms express a semantically middle domain. When we reexamine the word ἐνευλογέομαι within this light, a better option emerges for seeing its -θη- forms as manifestations of speech actions within this middle domain. In their own unique ways, the LXX as well as Philo, Paul, and Acts further corroborate this alternative. As a result, we may read ἐνευλογηθήσονται in Gen 12:3b LXX as a speech action middle: “to pronounce blessings.” The proposed reading promotes a better understanding of Abraham within Genesis LXX. Rather than a means to an end, Abraham remains at the center of God’s blessing as the earth’s families cry out: “God make me like Abraham!”

Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (, which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
© The Author(s), 2024. Published by Cambridge University Press on behalf of the President and Fellows of Harvard College


General consensus maintains that the Hebrew Bible (HB) and the Septuagint (LXX) represent two different understandings of Gen 12:3b. Although interpreters typically render the Hebrew niphal (ונברכו) with reflexive force, most straightforwardly accept the Greek future passive (ἐνευλογηθήσονται) as passive.Footnote 1 Here are the different readings:

ונברכו בך כל משפחת האדמה

And all the families of the earth shall bless themselves by you [i.e., Abraham].

καὶ ἐνευλογηθήσονται ἐν σοὶ πᾶσαι αἱ φυλαὶ τῆς γῆς.

And all the tribes of the earth shall be blessed in you [i.e., Abraham].Footnote 2

According to the two English translations, the HB and the LXX diverge on the roles of Abraham and the earth’s families apropos the blessing.

In the HB, Abraham models divine blessing. The earth’s families admire Abraham as the paragon of blessedness. Longing to obtain Abraham’s favor and blessing, the earth’s families bless themselves by his name. The reflexive sense—“bless themselves”—thus indicates something like a speech action, conveying the idea of blessing pronouncements.Footnote 3 Genesis 48:20 expresses a similar idea:

בך יברך ישראל לאמר ישמך אלהים כאפרים וכמנשה

By you [i.e., Ephraim and Manasseh] Israel shall pronounce blessings, saying: “God make you like Ephraim and Manasseh.”Footnote 4

We may read this blessing formula along the lines of a threefold speech act: 1) uttering the words “God make you like Ephraim and Manasseh” (locutionary act) 2) to bless Israel (illocutionary act) in the hopes of 3) bringing about the experience of blessing in Israel (perlocutionary act).Footnote 5 The emphasis, however, falls on Israel performing the locutionary and illocutionary acts, thereby setting Ephraim and Manasseh as paragons of blessing.Footnote 6 Returning to Gen 12:3b, we may understand its content analogically, inferring that the words uttered as blessing by the earth’s families resemble something like “God make you like Abraham,” thereby treating the patriarch as the paragon of blessing.Footnote 7

In the LXX, however, Abraham exists for the sake of the world. Rather than a paragon of blessing, Abraham becomes a source of blessing. As the means by which the world experiences blessing, Abraham brings blessing to the earth’s people. In the words of Francis Watson: “For the [LXX] translator, the calling and destiny of [Abraham is] to bring blessings to the entire world, and his rendering seeks to bring this out as clearly as possible.”Footnote 8 Rather than pronouncers of blessings, the earth’s people thus become the objects of blessing, remaining entirely passive in their reception of the Abrahamic blessing.

Although the passive reading of Gen 12:3b LXX remains dominant, this paper challenges and problematizes the reading, arguing that the LXX verse accords more with its HB counterpart than interpreters typically grant. The consensus reading largely relies on morphology alone to explicate verbal meaning. Thus, if a verb appears morphologically passive, we should then render the verb passive. If the morphologically passive verb happens to express nonpassive force, then that occurrence represents an outlier. Rarely does such an occurrence prompt us to consider the limits of morphology.Footnote 9 But recent linguistic studies apropos the ancient Greek middle voice call such assumptions into question by providing fresh alternatives for understanding aorist and future -θη- passive forms within a semantically middle domain. A study of the word ἐνευλογέομαι then reveals that its aorist and future passive forms may fit within this semantic middle domain, often as speech acts. As a result, we may render ἐνευλογηθήσονται in Gen 12:3b LXX (along with its four other appearances in 18:18; 22:18; 26:4; 28:14) as a speech action middle, that is, “to pronounce blessings.”

-θη-: Passive in Form, Middle in Domain

Building on recent linguistic studies, Rachel Aubrey and Carl Conrad argue convincingly that overly relying on morphology to understand verbal meaning in ancient Greek leaves us with a deficient and oversimplified understanding of the middle voice.Footnote 10 We typically understand voice to describe a relationship between subject and verb. As a result, we often assume a morphological change in voice corresponds with a shift in the semantic role of subject to verb. Active voice depicts a subject as an agent performing the verbal action. Passive voice depicts a subject as a patient suffering the force of the verbal action. A middle construction often then describes a combination of both active and passive voice: a subject here performs the verbal action as an agent but then suffers the effects of that same action as a patient. Thus, the middle voice overlaps semantically with reflexive meaning where two participants—agent and patient—participate in one coreferential verbal action. However, such an understanding of the middle voice is too narrow. Although reflexive meaning manifests a middle voice, it only represents one meaning among many in what we may call a middle domain. Furthermore, our typical understanding of voice also goes hand in hand with what we make of transitivity. While we often interpret voice as relating subject and verb, we take transitivity to relate object and verb. A transitive verb takes an object, whereas an intransitive verb does not. Such an understanding, however, sometimes becomes too constricting. For example, it disables a subject from fully participating in the developmental stages and processes (i.e., the “hows”) of verbal events in any meaningful way unless that subject morphs into an object.Footnote 11

When grappling with middle meaning in ancient Greek, difficulties arise due to the presence of three distinct morphological forms in aorist and future paradigms. Unlike present, imperfect, and perfect paradigms, which express one nonactive form, aorist and future paradigms express two nonactive forms: a sigmatic form (aorist: -σάμην, -σω, -σατο, etc.; future: -σομαι, -σει, -σεται, etc.) and a -θη- form (aorist: -θην-, -θης-, -θη, etc.; future: -θήσομαι, -θήσει, -θήσεται, etc.). To this tripartite morphological division, we typically attach three corresponding voices: an active voice, a middle voice marked by the sigmatic form, and a passive voice marked by the -θη- form. As a result, we assume verbs marked by -θη- safely signify passive meaning, thereby describing a verbal event where a subject receives the force of an action from someone else.Footnote 12

Aubrey and Conrad, however, reveal these assumptions to be too neat and tidy. Ancient Greek’s voice system represents a polarity of active–middle rather than one of active–passive. Ancient Greek inherits this polarity from its Proto-Indo-European language system, an active-middle voice system. As a result, ancient Greek inherits not three but only two morphological paradigms (i.e., active and middle) from its Proto-Indo-European ancestor. Although the middle form may express passive force, ancient Greek does not exhibit a unique passive form. According to Conrad, the so-called passive -θη- forms, supposedly intended to bear distinct passive meaning, were:

(a) relatively late developments in the history of ancient Greek and (b) were originally derivative from intransitive aorist forms and, far from ever bearing per se a distinct passive sense, competed with and ultimately supplanted the older Middle morphoparadigms in the aorist and the future tenses.Footnote 13

This insight helps us account for the reason why verbs attested in the active voice often express nonpassive meaning in the passive form. For example, ἐγείρω and εὐφραίνω both appear in passive form in Isa 26:19 LXX but convey nonpassive meaning: ἀναστήσονται οἱ νεκροί καὶ ἐγερθήσονται οἱ ἐν τοῖς μνημείοις καὶ εὐφρανθήσονται οἱ ἐν τῇ γῇ (The dead shall rise, and the ones in the tombs shall rise, and the ones on the earth shall rejoice).Footnote 14 In Isa 40:7 LXX, ξηραίνω appears in passive form but expresses nonpassive meaning: ἐξηράνθη ὁ χόρτος καὶ τὸ ἄνθος ἐξέπεσε (The grass withers, and the flower fades).Footnote 15

Furthermore, Conrad contends that “the so-called ‘middle/passive’ endings in the present, imperfect, and perfect tenses are fundamentally subject-focused in meaning and only secondarily came to assume any conventional passive function—and never did assume exclusively a passive function.”Footnote 16 The sigmatic middle and -θη- passive forms represented in aorist and future tenses then correspond to their present, imperfect, and perfect tense counterparts. In Aubrey’s words: “While voice is highly relevant to lexical semantics, directly altering the nature of the action described by the verb, tense is less relevant; it does not alter the meaning of the verb, but only distinguishes when it takes place.”Footnote 17 Moreover, very few aorist and future verbs exhibit both sigmatic and -θη- forms.Footnote 18 In fact, some aorist and future verbs default to -θη- forms without necessarily expressing passive meaning. Thus, the morphological passive form does not necessarily indicate passive meaning, since passive meaning is not intrinsic to a verb’s morphological form. Instead, we must determine meaning from a verb’s use in context and not rely only on assumptions arising from verbal morphology.

While Conrad maintains that morphological paradigms express ambivalent meanings (i.e., a verb may denote passive or middle meaning depending on context), Aubrey places the -θη- form within an overarching middle domain. This middle domain consists of middle verbs, some being more agent-focused and others being more patient-focused.Footnote 19 Following Suzanne Kemmer’s typology of verbs generally denoting middle voice, Aubrey and Conrad demonstrate that similar types of verbs fit within a middle domain in ancient Greek, even when these verbs appear passive in form.Footnote 20 This middle domain encompasses a spectrum of middle verbs. Some verbs land on the patient-like end of the middle spectrum: spontaneous processes, collective motions, bodily motions, physical processes, mental processes, and passives (i.e., actions that may arise without the subject’s volition or control and yet are still actions experienced by the subject). Some verbs land on the agent-like end of the middle spectrum: mental activities, speech acts, perception, direct reflexives, grooming acts, and reciprocals (i.e., actions that arise from the subject’s volition). As a result, many -θη- forms—often labeled “pseudo-passives” or “pseudo-reflexives”—make better sense when understood within this middle domain: ἐφάνη (appear/become visible), ὤφθη (appear/become visible), ἐκρύβη (hide), κατεποντίσθη (sink), ἐπαλαιώθη (grow old), ἐπωρώθη (become hard), ἐπληθύνθη (increase), ἐξηράνθη (dry up), συνήχθη (gather together), and ἐφοβήθη (become frightened).Footnote 21 Regardless of where a verb lands in the middle spectrum, the middle voice indicates that the subject remains deeply involved in the processes and activities of verbal events. Although English renderings (e.g., appear, grow old, etc.) sometimes aid us in illuminating middle meaning, we must remember that modern English and ancient Greek operate differently. Some things are simply hard to understand and communicate in translation. This difficulty, however, does not disqualify us from indicating a middle meaning in -θη- forms. Instead, the key to detecting middle meaning remains the subject’s deep agent-like or patient-like participation in the processes and activities of verbal events.

We now home in on mental activities and speech act middles. Like direct reflexives and reciprocals, a mental activity represents an event where a volitional participant experiences change via a verbal process. A speech act is like a mental activity, although reflected out loud and manifested in a verbal utterance. The volitional participant utters something, signifying a locutionary act. What the subject does with that utterance signifies the illocutionary act. What the subject generates as a result signifies the perlocutionary act. Through agent-like participation, the volitional participant may experience the speech act as both its inceptive source of energy (via the illocutionary act) and its endpoint (via the perlocutionary act).Footnote 22 Aubrey indicates that ἐυλογέω may sometimes express a speech act.Footnote 23 Building on what we have learned here, I will now argue that ἐνευλογέομαι—as a close cousin to εὐλογέω—may also denote a speech action middle and does so in Gen 12:3b as ἐνευλογηθήσονται.

’Ενευλογηθήσονται as Speech Action Middle

A. ’Ενευλογέομαι as Neologism

As we now consider ἐνευλογέομαι as a speech action middle, I propose that we treat the word as a neologism.Footnote 24 The proposal remains tentative due to the difficulty in LXX scholarship to define neologisms. For example, Johan Lust (LEH) classifies a neologism as a word proper to the LXX and literature contingent on it. For a word appearing in the LXX but also in contemporary papyri and/or literature (starting with Polybius in the second century BCE), he classifies it as a neologism with a question mark. Noting the tentative nature of his suggestions, he defines a neologism as a word probably not used earlier than the LXX.Footnote 25 For another similar example, Takamitsu Muraoka (GELS) classifies a neologism as a word not attested before the LXX. If such a word appears both in the LXX and Polybius, he posits that incomplete attestation explains its absence before the LXX, since the LXX does not shape Polybius.Footnote 26 We must notice a subtle dissimilarity between LEH and GELS. The former describes a neologism as a word “not used” before the LXX’s time. Taking a more careful approach, the latter prefers “not attested” prior to the LXX. An unattested word does not necessarily mean an unused one. John Lee indicates that words unattested in literature until the LXX are sometimes simply normal words used and understood by average ancient Greek speakers.Footnote 27 As Nikolaos Domazakis suggests, we cannot judge with much confidence whether words appearing in the LXX are new at the time of translation/composition or are created by the translators/authors.Footnote 28 As a result, Domazakis defines a neologism as a word not attested before the LXX. The word cannot appear in any extant Greek literature, epigraphy, or papyri dated prior to the respective LXX book’s accepted date. This word may reflect a translator’s/author’s morphosemantic coinage or may already exist in written and/or oral language at the LXX book’s compositional time but appears unrecorded in both literary and nonliterary texts predating the book, insofar as we can know.Footnote 29

An electronic search of the word ἐνευλογέομαι in the Thesaurus Linguae Graecae (TLG) yields 206 hits.Footnote 30 The first sixteen results are germane for our current purposes since they provide the word’s earliest extant chronology. Likely sometime in the third century BCE, ἐνευλογέομαι emerges in Gen 12:3; 18:18; 22:18; 26:4; 28:14. The word then appears in other LXX texts throughout the second century BCE in 1 Kgdms 2:29; Ps 9:24; Sir 44:21.Footnote 31 Quoting the Genesis texts, Philo of Alexandria (ca. 20 BCE–50 CE) records the word six times sometime in the first century CE.Footnote 32 In his letter to the Galatians, the apostle Paul then makes use of the word in the 50s CE (Gal 3:8).Footnote 33 And finally, sometime in the late first to early second century CE, Acts 3:25 also uses the word.Footnote 34 Like Philo, Paul and Acts both quote the Genesis texts. A further electronic search of ἐνευλογέομαι in the Searchable Greek Inscriptions, the Supplementum Epigraphicum Graecum, and the Papyrological Navigator databases supplies no results, leaving the word unattested in extant Greek epigrapha, papyri, and literature beyond what we find in TLG.Footnote 35 The word ἐνευλογέομαι thus appears to fit Domazakis’s definition of a neologism above, thereby letting us treat it as such.Footnote 36

Besides being “new” words, what do neologisms do? Robert Hiebert helpfully suggests that neologisms “remedy a perceived deficiency.”Footnote 37 In LXX Genesis, the translator renders ברך with εὐλογέω in Gen 1:22, 28; 2:3; 5:2; 9:1, 26; 12:3a, but with ἐνευλογεόμαι in 12:3b.Footnote 38 I wonder what “perceived deficiency” the translator feels needs remedying in 12:3. James Aitken points out that the LXX stereotypically renders ברך with εὐλογέω, whether the verbal subject expresses a human praising God or God blessing a human.Footnote 39 But what might the LXX do with verbal events depicting human-to-human blessing? Apropos texts where humans pronounce blessings on other humans, the LXX may appear interpretively elastic in its translations. For example, with the word ברך, Deut 10:8 (cf. 1 Chr 23:13) depicts Aaron and his sons pronouncing blessings by YHWH’s name. Here the LXX renders ברך with ἐπεύχομαι, a word found just twice in the LXX but in fairly wide use since Homer.Footnote 40 Rather than using the stereotypical εὐλογέω, the LXX attempts to make explicit what might appear vague unless specified with ἐπεύχομαι. The LXX wants to clarify for its readers that ברך here depicts prayer to God rather than human-to-human blessing. In other places, such as Jer 20:14, the LXX neologizes with ἐπευκτός in place of the expected εὐλογέω for ברך.Footnote 41 Jeremiah’s birthday is not a day after which to long. Again, the intent behind using a different word is to clarify and to specify. I suggest something similar is at work in Gen 12:3. John Wevers notes that the Genesis translator often makes every effort to elucidate the perceived Hebrew meaning in his Greek translations.Footnote 42 Moreover, a LXX translator may also coin a form with the aim to render each part of the Hebrew into Greek.Footnote 43 In this light, the Genesis translator’s “perceived deficiency” with εὐλογέω might pertain to the word’s broad semantic range, something the translator wants to narrow with ἐνευλογέομαι in 12:3b to clarify what he thinks the Hebrew means. As a result, when the Genesis translator annexes ἐν to εὐλογέω, the new formation may now express a built-in instrument or cause—i.e., to bless by means of someone/something—and mirror the parallel use of the preposition ב in Hebrew and Aramaic.Footnote 44

At this point, someone might object, insisting that ἐνευλογηθήσονται ἐν σοί is simply a stylistic appearance of a preverb (ἐν-) with a homonymous preposition, similarly expressed elsewhere in the LXX, and should thus be taken as an incidental corollary of εὐλογέω. For instance, Exod 14:4 manifests something akin: ἐνδοξασθήσομαι ἐν Φαραὼ. The word ἐνδοξάζομαι, however, also appears without a homonymous preposition in the same book, retaining its own nuance apart from δοξάζω (Exod 33:16). Furthermore, ἐνευλογέομαι itself appears without a homonymous preposition in 1 Kgdms 2:29; Ps 9:24 LXX. And when we return to Genesis, we discover that the stylistic pattern is inconsistent throughout the book. For example, Gen 48:20 displays a construction akin to 12:3; 18:18; 22:18; 26:4; 28:14 but with εὐλογέω instead of ἐνευλογέομαι: Ἐν ὑμῖν εὐλογηθήσεται Ἰσραὴλ. I suspect that the translator interprets 48:20 as straightforwardly depicting a speech pronouncement, while wondering if 12:3b might be open to interpretation unless specified. As a result, the translator feels no need to neologize in 48:20. In 12:3b, however, he feels the need to add clarity and nuance to a potentially vague event, and thereby narrow the broad semantic range of ברך with ἐνευλογέομαι.

B. Semantic Meaning of ’Ενευλογέομαι

We now turn to the semantic meaning of ἐνευλογέομαι, especially in its Genesis context. If ἐνευλογέομαι indeed narrows the semantic range of ברך, what might the word mean? The major lexicons express the semantic range of ἐνευλογέομαι: “to confer special benefits, act kindly, bless” (BDAG); “Pass., to be blessed in … Med., to take a blessing to oneself” (LSJ); “M: to take a blessing to oneself, to bless oneself. P: to be blessed in” (LEH); “to make happy … to enjoy the benefit of + gen.” (GELS); “(1) mid. to bless (for oneself) … (2) pass. to be blessed” (BrillDAG); “pass…. be renowned or blessed” (CGL).Footnote 45 The lexicons disagree on the default morphology. Where BDAG opts for an active form (ἐνευλογέω), LSJ, LEH, GELS, BrillDAG, and CGL prefer a middle (ἐνευλογέομαι). BDAG, LSJ, LEH, and BrillDAG notice the verb’s speech aspects but adhere to strict morphosyntactical categories, assuming the -θη- form conveys passive meaning. Of the word’s eight LXX appearances, six express -θη- forms (future: Gen 12:3; 18:18; 22:18; 26:4; 28:14; aorist: Sir 44:21) and two are present middle (1 Kgdms 2:29; Ps 9:24). Both NT occurrences are future -θη- forms, being composite LXX quotes (Acts 3:25; Gal 3:8). If Conrad and Aubrey are correct above, we should note as significant that ἐνευλογέομαι never expresses the so-called middle sigmatic morphology when aorist or future.Footnote 46 Even if we extend the scope to include εὐλογέω, neither the LXX nor the NT render the word with a sigmatic morphology in aorist or future tenses.Footnote 47 Before the LXX, εὐλογέω only manifests once as a sigmatic middle in antiquity (Isocrates, Evag. 9.5). As a result, we may infer that the LXX defaults to the -θη- form for ἐνευλογέομαι—as well as εὐλογέω—when expressing aorist and future tenses. At the same time, however, we should not infer that the LXX then intends to convey passive meaning every time the -θη- form appears.

In classical Attic the -θη- form signifies mental processes, physical processes, bodily motions, collective motions, and passives, thereby expressing a more patient-like end of the middle domain. During this era, speech acts normally occur in the sigmatic form. However, numerous speech acts also abound in the -θη- form, thus already revealing a fluidity between the two forms in aorist and future tenses.Footnote 48 The Hellenistic period, however, amplifies this fluidity. In Hellenistic Greek the -θη- form enlarges in scope and even begins to replace other middles, namely, those typically expressed with the sigmatic form: reflexives, reciprocals, and more agent-like volitional activities, and so on. The Hellenistic period thus encompasses two middle forms operating side by side for aorist and future tenses.Footnote 49 As a result, rather than conveying passive meaning, ἐνευλογηθήσονται likely signifies a speech action middle in its five Genesis appearances beginning in 12:3b. We may read 12:3b as follows: “And all the tribes of the earth shall pronounce blessings by you.” Landing on the agent-like side of the middle domain’s spectrum, a speech action depicts a verbal event where a participant volitionally performs something. A speech act’s volitional nature has important ramifications for Gen 12:3b in its LXX context. In Gen 12:1–3, God assures Abraham that God will make good on his word. If Abraham leaves his country, kindred, and father’s house, God will in fact bless him. Up to our own day, beginning with the fear of danger in an unknown land among unknown people, refugees face major setbacks. God calls Abraham to leave behind a life in a familiar land with familiar people and to become a refugee in an unknown land amid unknown people. And to culminate the call, God assures Abraham that the unknown land’s surrounding peoples will not hurt him but rather voluntarily pronounce blessings with his name.

Reinforcing Speech Action Middle within the LXX

To strengthen the likelihood of the speech action middle, we now examine the overlap between Gen 12:3b and other LXX texts making use of the blessing formula. But before we move forward, we should keep in mind what we imply when we call ἐνευλογέομαι a neologism. We suggest that εὐλογέω and ἐνευλογέομαι are different words. The former expresses a broader semantic range, whereas the latter narrows the former’s range. For these reasons, even as a -θη- form, εὐλογέω often conveys a different meaning from a speech action. For example, we see this occur in Ps 49:19 (48:19 LXX); 112:2 (111:2 LXX); 128:4 (127:4 LXX); Prov 20:9b; 28:20; Sir 1:13; Isa 65:16. At the same time, in the light of what we see above on ancient Greek -θη- forms, we should recall that these -θη- forms do not necessitate a passive reading. In fact, we may render all these cited occurrences here as “experience blessing” rather than “be blessed.” Although difficult to grasp with the English language, this rendering helps us see the verbal events as middle mental or spiritual experiences where the subjects remain involved and not objectified. When we locate analogous Greek constructions such as the blessing formula, however, sometimes a LXX translator does not neologize and instead retains εὐλογέω, conveying the force of a speech action middle like the narrower ἐνευλογέομαι as we see below.

We first consider Ps 72:17 (71:17 LXX) as the psalm applies the blessing formula to the Solomonic king.

ויתברכו בו כל־גוים יאשרוהו

And may they pronounce blessings by him; may all the nations pronounce him happy.Footnote 50

καὶ εὐλογηθήσονται ἐν αὐτῷ πᾶσαι αἱ φυλαὶ τῆς γῆς, πάντα τὰ ἔθνη μακαριοῦσιν αὐτόν.

And all the tribes of the earth will pronounce blessings by him; all the nations will pronounce him happy.

While the HB certainly connects Gen 12:3b and Ps 72:17 (71:17 LXX), the LXX psalm makes the overlap unmistakable by adding and making πᾶσαι αἱ φυλαὶ τῆς γῆς the subject of εὐλογηθήσονται. By inserting this phrase from Gen 12:3b, the LXX psalm explicitly analogizes the king to Abraham.Footnote 51 But the LXX psalm also appears to link itself to Gen 48:20, rendering the hithpael (ויתברכו) with εὐλογηθήσονται instead of Gen 12:3b’s ἐνευλογηθήσονται.Footnote 52 Although the concept of blessing pronouncements might need some teasing out in Gen 12:3b and Ps 72:17 (71:17 LXX), Gen 48:20 leaves no ambiguity. Blessing the two sons of Joseph, Jacob pronounces: Ἐν ὑμῖν εὐλογηθήσεται Ἰσραὴλ λέγοντες Ποιήσαι σε ὁ θεὸς ὡς Eφράιμ καὶ ὡς Μανασσή (Israel shall pronounce blessings by you, saying: “God make you like Ephraim and Manasseh”). Although we might be correct to say Israel will be blessed as a result, if we express εὐλογηθήσεται as passive, we miss the speech act’s force here. By Ephraim and Manasseh, Israel shall pronounce blessings. As we saw above, the force lands on Israel performing the blessing (illocutionary act) by uttering the words “God make you like Ephraim and Manasseh” (locutionary act), thus elevating the two boys to paragons of blessing.Footnote 53

Returning to Ps 72:17 (71:17 LXX), we must now notice the parallelism between the two lines: καὶ εὐλογηθήσονται ἐν αὐτῷ πᾶσαι αἱ φυλαὶ τῆς γῆς and πάντα τὰ ἔθνη μακαριοῦσιν αὐτόν. Akin to its Hebrew counterpart, the lines form a couplet where the latter clarifies and informs the meaning of the former.Footnote 54 In the latter line, the nations clearly notice the king’s blessing, making pronouncements about his fortunate state. Prayer is continually offered on the king’s behalf, and all day long people bless him (71:15 LXX). The king’s blessed name endures through all the ages; his name endures longer than the sun (71:17 LXX). The nations thus reckon the king favored, admiring him and yearning for a blessing like his. This all illuminates how we read the former line of our couplet. The earth’s tribes pronounce blessings by the king’s name, saying something along the lines of: “God make us like the king!” As in Gen 48:20, we might be right to say the earth’s tribes will be blessed as a result, but if we render εὐλογηθήσονται as passive, we misunderstand the verbal event. Psalm 72 (71 LXX) prioritizes the king over the nations, and εὐλογηθήσονται climactically underlines that priority when the nations take the king’s name upon their lips in blessings. Akin to Ephraim and Manasseh but in a much higher register, Ps 72:17 (71:17 LXX) depicts the king as a paragon of blessing. All this then reinforces our reading of Gen 12:3b, especially in the light of the LXX psalm’s effort to unite the two texts. Abraham and the king remain analogues, and the earth’s tribes now pronounce blessings by the king, patterned after the blessings they first pronounce by Abraham.

Sirach 44:21 presents a similar point, depicting the blessing formula in Hebrew MS B and the LXX.

על כן בש[בו]עה הקים לו לברך בזרעו גוים

διὰ τοῦτο ἐν ὅρκῳ ἔστησεν αὐτῷ ἐνευλογηθῆναι ἔθνη ἐν σπέρματι αὐτοῦ.Footnote 55

We may translate the Hebrew: “Therefore by an oath he [i.e., God] established for him [i.e., Abraham] to bless nations by his seed.” But we may also read it: “Therefore by an oath God established for Abraham that nations might bless [i.e., pronounce blessings] by his seed.” When the preposition ל joins an infinitive, the union may indicate intentionality and alter the verb’s subject joined to the preposition. For example, in Isa 10:2 iniquitous people decree evil (10:1) with an intention that widows might become their spoil (להיות אלמנות שללם).Footnote 56 Mirroring its Hebrew counterpart, Sir 44:21 LXX may read: “Therefore God established for Abraham by an oath that nations might bless [i.e., pronounce blessings] by his seed.” Furthermore, the end of Sir 44:21 (both MS B and LXX) and Ps 72:8 (71:8 LXX) resonate verbally. This resonance is not accidental, since the blessing formula appears both in Sir 44:21 and Ps 72:17 (71:17 LXX). But the LXX also inserts a version of Gen 22:17 amid Sir 44:21, tightening the analogies between the king, Abraham, and Israel as paragons of blessing desired by the nations.Footnote 57 Due to these various reasons, we may render the aorist passive ἐνευλογηθῆναι in Sir 44:21 as a speech action middle, mirroring the Hebrew source.

An Afterlife of ’Ενευλογέομαι as a Test Case

A. Philo of Alexandria

To provide a test case for our thesis, we now move beyond the LXX to consider the afterlife (Nachleben) of ἐνευλογέομαι and the blessing formula in Greek-speaking antiquity. Here we ask whether our reading of ἐνευλογέομαι fits with how ancient Greek readers use the blessing formula texts. Although the ancient Greek readers below all read the blessing formula with their own nuances, we may interpret their readings of ἐνευλογέομαι as expressions of the middle domain.

Philo of Alexandria appears first, making use of ἐνευλογέομαι when he reflects on the Genesis patriarchs.Footnote 58 For our purposes, we primarily examine De Migratione Abrahami, where Philo quotes Gen 12:3 three times (Migr. 1, 118, 122).Footnote 59 But to provide a fuller picture, we also read what Migr. says about ἐνευλογέομαι in the light of Philo’s wider corpus.Footnote 60 Before approaching Gen 12:3, we should notice at the outset two features of Philo’s take on Abraham and the word εὐλογία. First, Philo regards Abraham as a paragon of character. Abraham is “a rule of nobility for all strangers/foreigners” (Virt. 219: ἅπασιν ἐπηλύταις εὐγενείας ἐστὶ κανών). These strangers reckon his life worthy of emulation, treating him as a king and recognizing his soul’s grandeur (Virt. 211, 216–18). They abandon their former lives in paganism to pursue a life patterned and modeled after Abraham’s (Virt. 219).Footnote 61 Second, Philo dissects the word εὐλογία into εὖ and λόγος.

τὸ γὰρ εὖ πάντως ἐπʼ ἀρετῆς· λόγος δὲ ὁ μὲν πηγῇ ἔοικεν, ὁ δὲ ἀπορροῇ, πηγῇ μὲν ὁ ἐν διανοίᾳ, προφορὰ δὲ ἡ διὰ στόματος καὶ γλώττης ἀπορροῇ.

For εὖ certainly concerns excellence of character. As for λόγος, one aspect resembles a spring, the other an outflow: that which is in the intellect resembles the spring, but the utterance by mouth and tongue resembles the outflow. (Philo, Migr. 70–71)

Whereas εὖ straightforwardly connotes excellence, λόγος expresses two aspects concerning reason and speech. As a result, when God declares the words εὐλογήσω σε to Abraham (Gen 12:2), God promises to gift the patriarch with excellent reason and speech, the blessing’s central features (Migr. 70).Footnote 62 Philo’s breakdown of εὐλογία comports conveniently with our study thus far on ἐνευλογέομαι. Within the middle domain, ἐνευλογέομαι pertains to a mental activity arising volitionally from its subject and manifesting in a speech action.

When we turn to his reflections on Gen 12:3 in Migr. 109–27, Philo converges these two points. Abraham remains a model of excellent character for onlookers, and blessing manifests excellence in mental activities and speech utterances. For Philo, Gen 12:3a depicts an event where others notice the righteous human’s blessed state and desire it. As a result, the onlookers reckon the righteous human with honor and pronounce blessings upon the Abrahamic figure (Migr. 109–10). The uttered blessings alone, however, do not generate blessing for the onlookers. Harking back to Migr. 70–71, Philo again analogizes the mind to a fountain (πηγή) as the standard by which humans test their words (Migr. 117). Thus, the mind’s internal activities must correspond to the mouth’s external speech pronouncements in order to experience blessing. Here Philo draws on the figure of Balaam, someone whose internal intentions do not match his external words (cf. Leg. 3.210). At the same time, however, Philo’s demand to unite the mind and the mouth creates a context for imperfect but well-intentioned humans to experience and grow in blessing, something that spills over into Gen 12:3b (Migr. 111–17).Footnote 63

We now arrive at Philo’s reading of Gen 12:3b (Migr. 118–27). In Migr. 118–19, Philo sums up his take on the blessing formula:

μέγιστον δʼ ἑξῆς, ὅταν ἡσυχάζωσιν ἐκεῖνοι, τὸ μηδὲν μέρος φύσεως λογικῆς ἀμέτοχον εὐεργεσίας ἀπολείπεσθαι· λέγει γὰρ ὅτι “ἐνευλογηθήσονται ἐν σοὶ πᾶσαι αἱ φυλαὶ τῆς γῆς.” ἔστι δὲ τοῦτο δογματικώτατον· ἐὰν γὰρ ὁ νοῦς ἄνοσος καὶ ἀπήμων διατελῇ, ταῖς περὶ αὐτὸν ἁπάσαις φυλαῖς τε καὶ δυνάμεσιν ὑγιαινούσαις χρῆται, ταῖς τε καθʼ ὅρασιν καὶ ἀκοὴν καὶ ὅσαι αἰσθητικαὶ καὶ πάλιν ταῖς κατὰ τὰς ἡδονάς τε καὶ ἐπιθυμίας καὶ ὅσαι ἀντὶ παθῶν εἰς εὐπάθειαν μεταχαράττονται.

But the greatest follows: even when they remain silent, no portion of the rational nature remains free from a benefit. For it says: “All the tribes of the earth shall bless themselves by you.” And this saying is very instructive. For if the mind continues free from sickness and harm, it employs all the healthy tribes and powers around it: those pertaining to sight and hearing and all others pertaining to sense-perception, and again those pertaining to pleasure and desire and all changing from passionate emotions to good ones.Footnote 64

Philo reads Gen 12:3b in two different and overlapping registers, one higher and the other lower. As we see in the quote above, the higher register interprets the verse as an allegory of the soul. Here Abraham and the tribes respectively typify the mind and the soul’s lesser parts, pertaining to sense, desire, and passion. The lower register sees the verse depicting a righteous human within a less-enlightened society. Here Abraham and the tribes straightforwardly represent the righteous human and less-enlightened humanity. Although Philo begins with the allegory of the soul, he illustrates its meaning by fleshing out the righteous human. As the human devotes themself to virtue and righteousness, they become humanity’s pillar, bringing out (προφέρων) everything they have into society’s midst for the benefit of those who might use it (χρησομένων), and giving without any grudges (Migr. 120–21).Footnote 65 The righteous human also intercedes for the less virtuous, becoming a supplicant and offering their word of supplication (τὸν ἱκέτην ἑαυτοῦ λόγον), something God does not ignore. Here Philo recalls Num 14:20, where God answers Moses’s prayer to forgive the Israelites with the words: “I am merciful to them in accordance with your word (τὸ ῥῆμά σου).” Philo immediately equates these words with Gen 12:3b and then adds Abraham’s intercession for Sodom as another example (Migr. 122).Footnote 66

After these various examples, one might conclude that Philo reads ἐνευλογηθήσονται as a passive.Footnote 67 But I suggest otherwise. We may also read the above examples as prompts to the less enlightened. For them to experience blessing, they must look at the righteous human and take advantage of the benefit in their midst. For example, although Philo brings up the intercessions of Moses and Abraham, he knows that neither story depicts God withholding judgment from people deemed obstinate or wicked (Gen 18:23–19:29; Num 14:20–23). Or, when the righteous human offers benefits to the surrounding society, Philo mentions that the onlookers must make use of them (χρησομένων; Migr. 121). Reflecting elsewhere on Abraham’s intercession, Philo hopes that such acts might lead the less enlightened to make use of them (χρήσασθαι) for a better and more stable life (Sacr. 123). Again, elsewhere, where good things are graciously held out to the imperfect, it is to challenge them to zealously pursue and participate in virtue (προκαλούμενος αὐτοὺς εἰς μετουσίαν καὶ ζῆλον ἀρετῆς; Leg. 1.34). In both cases, if the less enlightened forgo the opportunities held out to them, they do not experience the righteous human’s blessing. But the opposite remains true as well. If the imperfect but well-intentioned people lay hold of the offered benefits, they grow in virtue and righteousness, becoming more like the righteous human they desire to emulate in Gen 12:3a. Thus, Philo seems to assume that the less enlightened participate as subjects deeply involved in the verbal event described in ἐνευλογηθήσονται, something that comports with our understanding of the middle domain.

To specify further this deep involvement described in ἐνευλογηθήσονται, we close our discussion of Philo with Migr. 124. Here Philo returns to the soul’s allegory, converging his higher and lower registers of Gen 12:3b. He now calls “us” to pray (εὐχώμεθα) that the mind in our souls parallel the righteous human’s place in humanity for the healing of “our” moral maladies. As long as they remain healthy (i.e., the mind and the righteous human merged together), the hope of healing survives, because God holds out (προτείνας) the all-healing medicine to supplicants and worshipers (τῷ ἱκέτῃ καὶ θεραπευτῇ) to use (χρῆσθαι) in order to heal the ill and their souls’ wounds. Like the mind and the righteous human, here the soul’s lesser parts and less-enlightened humanity also merge in “us” and now become supplicants and worshipers, paralleling the righteous human above as they pray that God might heal soul-wounds, both their own and those of others. To benefit from what God offers, “we” need to use the all-healing medicine by praying for “our” sake that the righteous mind/human might remain among “us.” Through the utterance of prayer, the soul’s lesser parts heal and become more like its righteous mind. It might help to think again about the mind as Abraham and the “us” as the tribes of Gen 12:3b. If we apply Philo’s logic, through the utterance of blessings, the earth’s tribes bless themselves in order to become more like Abraham. As the soul’s lesser parts conform themselves to its righteous mind, the tribes conform themselves to Abraham the righteous human. As a result, in both cases, the inferred prayer or blessing may sound something like “God make me like Abraham” or, more allegorically, “God make my soul’s lesser parts like its virtuous mind.” For this reason, we may read Philo as someone who interprets ἐνευλογηθήσονται within a middle domain as an internal mental activity and an external speech action.

B. Paul

For our next test case, we briefly consider Paul’s letter to the Galatians. In Gal 3:8, Paul applies the Genesis blessing formula to the justification of gentiles.Footnote 68 Quoting an amalgamation of three different texts where ἐνευλογέομαι appears, Paul expounds on the meaning of the Abrahamic promise:

προϊδοῦσα δὲ ἡ γραφὴ ὅτι ἐκ πίστεως δικαιοῖ τὰ ἔθνη ὁ θεός, προευηγγελίσατο τῷ Ἀβραὰμ ὅτι ἐνευλογηθήσονται ἐν σοὶ πάντα τὰ ἔθνη.Footnote 69

And the scripture foreseeing that God would justify the gentiles by faith, in advance announced the gospel to Abraham, saying: “All the gentiles will pronounce blessings by you.”

Although many—even those sympathetic to reading Paul within Judaism and Israel’s story—read ἐνευλογηθήσονται as a passive here, I suggest we may read Paul with middle or reflexive force.Footnote 70 Before the blessing formula appears, Paul reminds the Galatian gentiles that they receive the Spirit by hearing rather than circumcision (Gal 3:1–5). He then directs their attention to Abraham who believes God and is counted righteous, holding the patriarch up as a model to emulate (3:6). But for Paul, the Galatian gentiles do not simply imitate Abraham, they mysteriously attach to him.Footnote 71 Through faith, they become family, as his sons (3:7). I propose this explains why Paul announces the blessing formula next, equating it to the gospel and the justification of the gentiles by faith. As I have argued above, the divine words spoken to Abraham announce that the earth’s tribes will pronounce blessings by the patriarch, uttering something like: “God make me like Abraham.”Footnote 72 For Paul, this day has arrived in Christ who brings the blessing of Abraham to the gentiles in order that they receive the promise of the Spirit by faith (3:14). Christ thus makes faith possible for the gentiles, so they might pronounce that God make them like Abraham apart from circumcision. In this way, they experience blessing together with Abraham, receiving the Spirit and becoming the patriarch’s gentile sons without needing to be circumcised (3:7, 9). As a result, Paul appears to pass our test, as someone who reads ἐνευλογηθήσονται in a middle domain expressing a speech action.

C. Acts

Our final test case brings us to Acts 3:25–26 where the blessing formula applies to Jews instead of gentiles. As Peter preaches to the Jewish people gathered around Solomon’s portico, he quotes an amalgamation of two texts containing the blessing formula:

ὑμεῖς ἐστε οἱ υἱοὶ τῶν προφητῶν καὶ τῆς διαθήκης ἧς διέθετο ὁ θεὸς πρὸς τοὺς πατέρας ὑμῶν λέγων πρὸς Ἀβραάμ· καὶ ἐν τῷ σπέρματί σου [ἐν]ευλογηθήσονται πᾶσαι αἱ πατριαὶ τῆς γῆς. ὑμῖν πρῶτον ἀναστήσας ὁ θεὸς τὸν παῖδα αὐτοῦ ἀπέστειλεν αὐτὸν εὐλογοῦντα ὑμᾶς ἐν τῷ ἀποστρέφειν ἕκαστον ἀπὸ τῶν πονηριῶν ὑμῶν.Footnote 73

You are the sons of the prophets and of the covenant God made with your fathers, saying to Abraham: “And all the earth’s families will experience blessing by your seed.” Raising up his servant, God sent him to you first to bless you so that each of you might turn away from your wickedness.

The use of πατριά is significant here. Unlike Paul above, who chooses ἔθνη to single out the gentiles, Acts broadens the scope with πατριαί to convey that Abraham’s blessing must be realized in all people, both Jews and gentiles. This explains what Peter claims next, namely, that God raised up the divine servant (i.e., Jesus) and sent him to the Jewish people first in order to bless them. Here Acts adopts two Pauline principles: Jesus is the promised seed (τῷ σπέρματι) and the gospel comes to Jews first (Gal 3:16; Rom 1:16). Despite these similarities, Acts applies the blessing formula to address a different concern from Paul’s justification of the gentiles. For Acts, the formula addresses how Israel, as God’s elect people, might experience the blessing of Abraham’s seed before it expands out to all the earth (cf. 1:8).Footnote 74 Jesus blesses them by providing an opportunity for them to turn away from their wickedness. But they only experience blessing when they repent and take advantage of the opportunity (cf. 3:19).Footnote 75 As a result, ἐνευλογηθήσονται does not express a speech action in Acts 3:25. The word, however, exhibits a mental activity, describing the voluntary event of repentance and changing one’s mind, and thereby still manifests the middle domain. For these reasons, we may infer that Acts 3:25 mostly passes our test. Although not as a speech action, ἐνευλογηθήσονται still depicts a verbal event where a subject remains deeply and volitionally involved.Footnote 76 But as we see above, Acts emerges late, possibly as late as the early second century CE.Footnote 77 As a result, its subtle differences might manifest the beginnings of what eventually becomes the more standard Christian passive reading of Gen 12:3b and the blessing formula. At the same time, however, Acts does not itself necessitate such a reading.


In this article, I have challenged and problematized the long-held consensus that we should read ἐνευλογηθήσονται as a passive in Gen 12:3b LXX. Of course, since texts may be read in a variety of ways, the consensus reading remains an option, if one wants to rely on morphology alone. However, in the place of such a reading, I have proposed an alternative and better option, interpreting the word as a speech action middle that conveys the sense “to pronounce blessings.” Fresh linguistic studies on ancient Greek voice undergird my proposal, demonstrating that verbs in -θη- form—typically considered passive—actually manifest a middle domain. In various ways, the LXX and Philo, Paul, and Acts provide reinforcement. By reading forwards rather than backwards, my study avoids imposing later Christian interpretation upon the blessing formula, thus making better contextual sense of Genesis LXX as a narrative with its own integrity. Thus, rather than a means to an end, Abraham remains at the center of God’s blessing as the earth’s families cry out: “God make me like Abraham!”


1 Although this remains the general consensus for the HB, serious challengers persist. For recent efforts to interpret the niphal as passive, see Keith N. Grüneberg, Abraham, Blessing, and the Nations: A Philological and Exegetical Study of Genesis 12:3 in Its Narrative Context (BZAW 332; Berlin: de Gruyter, 2003); André Flury-Schölch, Abrahams Segen und die Völker. Synchrone und diachrone Untersuchungen zu Gen 12, 1–3 unter besonderer Berücksichtigung der intertextuellen Beziehungen zu Gen 18; 22; 26; 28; Sir 44; Jer 4 und Ps 72 (FB 115; Würzburg: Echter, 2007). For a recent defense of the classical reflexive rendering, see R. W. L. Moberly, The Theology of the Book of Genesis (Cambridge: Cambridge University Press, 2009) 151–55. For the passive Greek rendering, see John W. Wevers, Notes on the Greek Text of Genesis (SCS 35; Atlanta: Scholars Press, 1993) 164; John W. Wevers, “The Interpretative Character and Significance of the Septuagint Version,” in From the Beginnings to the Middle Ages (Until 1300) (ed. Magne Saebø; vol. 1.1 of Hebrew Bible/Old Testament: The History of Its Interpretation; Göttingen: Vandenhoeck & Ruprecht, 1996) 84–107, at 97; Marguerite Harl, La Genèse. Traduction du texte grec de la Septante, Introduction et Notes (2nd ed.; La Bible d’Alexandrie 1; Paris: Cerf, 1994) 56, 153; Susan Brayford, Genesis (Septuagint Commentary Series; Leiden: Brill, 2007) 289–90; La Biblia griega. Septuaginta. I. Pentateuco (ed. Natalio Fernández and María Victoria Spottorno; Salamanca: Ediciones Sígueme, 2008); Septuaginta Deutsch. Das griechische Alte Testament in deutscher Übersetzung (ed. Wolfgang Kraus and Martin Karrer; Stuttgart: Deutsche Bibelgesellschaft, 2009) 14; Francis Watson, Paul and the Hermeneutics of Faith (2nd ed.; London: Bloomsbury, 2016) 169. James K. Aitken, The Semantics of Blessing and Cursing in Ancient Hebrew (ANES 23; Louvain: Peeters, 2007) 104–5, 114, however, notes the possibility of a nonpassive rendering intended in the future passive form. Nevertheless, to the best of my knowledge, I am unaware of any fully developed nonpassive interpretations of Gen 12:3b LXX.

2 Unless I note otherwise, all translations are my own. All LXX texts are from Septuaginta. Vetus Testamentum Graecum Auctoritate Academiae Scientiarum Gottingensis editum (20 vols.; Göttingen: Vandenhoeck & Ruprecht, 1931–). Genesis Hebrew texts are from Genesis (ed. Abraham Tal; vol. 1 of Biblica Hebraica Quinta; Stuttgart: Deutsche Bibelgesellschaft, 2015). All other HB texts are from BHS. All NT texts are from NA28.

3 Throughout the article, I pragmatically interchange “speech action” and “speech act” to mean the same thing.

4 Although the Masoretic Text (MT) points יברך as a piel (יְבָרֵךְ), we may also vocalize the word as a niphal (יִבָּרֵךְ). See Wevers, Notes on Greek Genesis, 818. Genesis BHQ (ed. Tal), 194*, notes that the Samaritan Pentateuch reads יברך as a hithpael with an assimilated ת (יִבָּרֵךְ), which two old Samaritan Targum manuscripts make visible by rendering the word as יתברך. Thus, the MT only provides one option amid others. The alternative niphal and hithpael possibilities, however, tighten the overlap between Gen 12:3; 18:18; 22:18; 26:4; 28:14; 48:20.

5 For a classic understanding of speech acts, see J. L. Austin, How to Do Things with Words (ed. J. O. Urmsson and Marina Sbisà; Cambridge: Harvard University Press, 1975); John R. Searle, Speech Acts: An Essay in the Philosophy of Language (Cambridge: Cambridge University Press, 1969). For how speech act theory might inform biblical interpretation, see Richard S. Briggs, Words in Action: Speech Act Theory and Biblical Interpretation (Edinburgh: T&T Clark, 2001). See also my comments in n. 22.

6 Although the blessing pronouncement showcases the role of Ephraim and Manasseh as paragons of blessing, the locutionary and illocutionary acts are not simply intended to bring honor to the boys. Rather, the blessing concerns Israel’s ongoing life as Israel pronounces blessings on itself (or one another) using the words “God make you like Ephraim and Manasseh” (see Ephraim A. Speiser, Genesis [AB 1; New York: Doubleday, 1964] 358). Thus, Gen 48:20 indicates and provides an ongoing blessing formula intended for Israel beyond Ephraim and Manasseh’s own space and time.

7 This move is classically expressed by Rashi. For a fuller discussion, see Moberly, Theology of Genesis, 150–56.

8 Watson, Paul and the Hermeneutics of Faith, 169. By “rendering,” Watson refers to the morphological -θη- passive form of ἐνευλογέομαι. Although he uses “Israel” for “Abraham,” the point remains the same for both figures.

9 For example, Rachel Aubrey, “Motivated Categories, Middle Voice, and Passive Morphology,” in The Greek Verb Revisited: A Fresh Approach for Biblical Exegesis (ed. Steven E. Runge and Christopher J. Fresch; Bellingham: Lexham, 2016) 563–625, at 567, makes the point well: “Applying labels such as ‘deponents’ or ‘passive in active sense’ creates a sense of legitimacy in the midst of the inconsistencies in its behavior, giving scholars occasion to invent new subcategories and further rules to justify their existence as leaks in the system.”

10 As an outworking of Suzanne Kemmer’s book, The Middle Voice (Typological Studies in Language 23; Amsterdam; Philadelphia: John Benjamins, 1993), Rutger J. Allan, The Middle Voice in Ancient Greek: A Study in Polysemy (Amsterdam Studies in Classical Philology 11; Amsterdam: Brill, 2003), specifically applies her approach with respect to ancient Greek. Afterwards, the approach expands in Carl W. Conrad, “New Observations on Voice in the Ancient Greek Verb,” Ancient Greek Voice: Propositions Concerning Ancient Greek Voice, 19 November 2002,; Aubrey, “Motivated Categories,” 563–625; Rachel Aubrey, “Hellenistic Greek Middle Voice: Semantic Event Structure and Voice Typology” (MA thesis, Trinity Western University, 2020). Being especially indebted to Aubrey and Conrad, I try to summarize and distill their research here.

11 Aubrey, “Hellenistic Greek Middle Voice,” 6, 13–14.

12 Ibid., 12–13.

13 Conrad, “New Observations,” 6. He relies on Andrew L. Sihler, New Comparative Grammar of Greek and Latin (Oxford: Oxford University Press, 1995) §§414, 508; Guy Cooper, Attic Greek Prose Syntax (Ann Arbor: University of Michigan Press, 1998) §§52.6.

14 Cf. Matt 24:7, 11: ἐγερθήσεται γὰρ ἔθνος ἐπὶ ἔθνος … καὶ πολλοὶ ψευδοπροφῆται ἐγερθήσονται (For nation will rise against nation … and many false prophets will arise).

15 For more examples, see Aubrey, “Motivated Categories,” 566–67.

16 Conrad, “New Observations,” 8 (italics omitted).

17 Aubrey, “Hellenistic Greek Middle Voice,” 63.

18 For a list of verbs that exhibit both forms in the NT, see Conrad, “New Observations,” 15.

19 Aubrey, “Motivated Categories,” 565, 612–16; Conrad, “New Observations,” 11.

20 Conrad, “New Observations,” 9–10; Aubrey, “Hellenistic Greek Middle Voice,” 78–141.

21 Aubrey, “Motivated Categories,” 575–76.

22 Aubrey, “Motivated Categories,” 606–10. Before Aubrey, Kemmer, Middle Voice, 133, 269, and Allan, Middle Voice in Ancient Greek, 105–12, both explain how speech acts denote and manifest a middle domain, more generally across various languages (Kemmer) but also more specifically within ancient Greek (Allan). For this reason, I draw on classical studies concerning speech act theory (mentioned above in n. 5) to build on Aubrey, bringing these to bear on our discussion to help us better grasp how this particular expression of middle meaning works. Thus, while speech act theory does not itself lead us to a middle meaning, it helps us explicate how middle meaning operates once we have located speech acts within a middle domain.

23 Aubrey, “Hellenistic Greek Middle Voice,” 116.

24 In addition to ἐνευλογέομαι, the LXX also neologizes in other renderings of ברך into Greek. For example, Aitken, Semantics of Blessing and Cursing, 33, 105, judges both εὐλογητός (a rendering of the Qal Passive Participle ברך) and ἐπευκτός (Jer 20:14) to be neologisms.

25 Johan Lust, Erik Eynikel, and Katrin Hauspie, A Greek-English Lexicon of the Septuagint (3rd ed.; Stuttgart: Deutsche Bibelgesellschaft, 2015) xiv.

26 Takamitsu Muraoka, A Greek-English Lexicon of the Septuagint (Leuven: Peeters, 2009) xiii. For an extensive account beyond LEH and GELS of how LXX scholarship defines and identifies neologisms, see Nikolaos Domazakis, The Neologisms in 2 Maccabees (Studia Graeca et Latina Lundensia 23; Lund: Lund University [Media-Tryck], 2018) 71–85.

27 John A. L. Lee, A Lexical Study of the Septuagint Version of the Pentateuch (SCS 14; Chico, CA: Scholars Press, 1983) 40–50. See also James K. Aitken, “Neologisms: A Septuagint Problem,” in Interested Readers: Essays on the Hebrew Bible in Honor of David J. A. Clines (ed. James K. Aitken, Jeremy M. S. Clines, and Christl M. Maier; Atlanta: SBL Press, 2013) 315–29. Aitken calls “for more descriptors of so-called new words, identifying them as semantic extensions, unattested compounds, morphological extensions, foreign loans, and so on” (321).

28 Domazakis, Neologisms, 75.

29 Ibid., 95. Although his study pertains to neologisms in 2 Maccabees, I adopt his understanding mutatis mutandis.

30 Thesaurus Linguae Graecae ® Digital Library,

31 Domazakis, Neologisms, 359–63, provides a tentative chronology of the LXX books, relying on Marguerite Harl, Gilles Dorival, and Olivier Munnich, La Bible grecque des Septante. Du judaïsme hellénistique au christianisme ancien (Paris: Cerf, 1988); The T&T Clark Companion to the Septuagint (ed. James K. Aitken; London: Bloomsbury, 2015).

32 Philo, Migr. 1; 118; 122; Her. 8; Somn. 1.3, 176.

33 J. Louis Martyn, Galatians: A New Translation with Introduction and Commentary (AB 33A; New York: Doubleday, 1997) 19–20, locates the letter in the 50s CE.

34 Craig S. Keener, Acts (New Cambridge Bible Commentary; Cambridge: Cambridge University Press, 2020) 46–48, suggests a date in the 70s–80s CE, whereas Richard L. Pervo, Acts: A Commentary (ed. Harold W. Attridge; Hermeneia; Minneapolis: Fortress, 2009) 5–7, argues for ca. 115 CE.

35 Searchable Greek Inscriptions: A Scholarly Tool in Progress, by The Packard Humanities Institute,; Supplementum Epigraphicum Graecum Online,; Papyrological Navigator,

36 LEH 203; GELS 237; Harl, Genèse, 56, 153; Robert J. V. Hiebert, “Textual and Translation Issues in Greek Genesis,” in The Book of Genesis: Composition, Reception, and Interpretation (ed. Craig A. Evans, Joel N. Lohr, and David L. Petersen; Leiden: Brill, 2012) 405–26, at 410–11, also all classify ἐνευλογέομαι as a neologism.

37 Hiebert, “Textual and Translation Issues,” 410.

38 Wevers, “Interpretative Character and Significance,” 95, posits that LXX Genesis is the work of one translator.

39 Aitken, Semantics of Blessing and Cursing, 33.

40 For example, ἐπεύχομαι appears twenty-one times in Homer, Il., and nine times in Od.

41 Aitken, Semantics of Blessing and Cursing, 105, calls ἐπευκτός a neologism, appearing elsewhere only in Pss. Sol. 8:16.

42 Wevers, “Interpretative Character and Significance,” 100.

43 Aitken, “Neologisms,” 326. He conjectures how this might be the case with ἔντριτος and המשלש in Eccl 4:12.

44 See Heinrich von Siebenthal, Ancient Greek Grammar for the Study of the New Testament (Oxford: Lang, 2009) 277, 288, 294. At 288, however, he suggests that prepositional prefixes (“preverbs”) rarely alter the meaning of a word unless the prefix is a double preverb (e.g., ἀντι-παρ-ἐρχομαι). But in Gal 3:8, we see three words consisting of a single preverb—προοράω, προευαγγελίζομαι, and ἐνευλογέομαι—which all express more specific meaning than their base words.

45 Frederick W. Danker et al., A Greek-English Lexicon of the New Testament and Other Early Christian Literature (3rd ed.; Chicago: University of Chicago Press, 2000) 336 (italics omitted); Henry George Liddell, Robert Scott, and Henry Stuart Jones, A Greek-English Lexicon (9th ed.; Oxford: Oxford University Press, 1996) 564; LEH 203; GELS 237; Franco Montanari, The Brill Dictionary of Ancient Greek (ed. Madeleine Goh et al.; Leiden: Brill, 2015) 694; James Diggle et al., The Cambridge Greek Lexicon (Cambridge: Cambridge University Press, 2021) 494.

46 After preverbs attach to verbs, the new words may shift from active to nonactive morphology. For example, διαλέγομαι (διά + λέγω) and ἐνδοξάζομαι (ἐν + δοξάζω) never appear in the LXX or the NT in the active voice. Even more, the sigmatic (so-called middle) and -θη- (so-called passive) forms of διαλέγομαι both convey nonpassive meaning. The same shift from active to nonactive morphology appears manifest in ἐνευλογέομαι.

47 Eὐλογέω appears in the LXX ten times as future -θη- (Gen 48:20; 2 Sam 7:29; Ps 48:19; 71:17; 111:2; 127:4; Prov 20:20; 28:20; Sir 1:13; Isa 65:16) and five times as aorist -θη- (Judg 5:24; 9:19; Tob 4:12; Isa 36:16). Its fifty-six other middle appearances are all either present (three times) or perfect (fifty-three times). The NT renders the word twelve times with middle form as either present (twice) and perfect (ten times), never as aorist or future.

48 For a discussion and a list of examples, see Allan, Middle Voice in Ancient Greek, 160–4, 169. Although Allan’s discussion here revolves around the aorist tense, we may still use it to convey our point, recalling the words of Aubrey, “Hellenistic Greek Middle Voice,” 63: “While voice is highly relevant to lexical semantics, directly altering the nature of the action described by the verb, tense is less relevant; it does not alter the meaning of the verb, but only distinguishes when it takes place.”

49 See Aubrey, “Hellenistic Greek Middle Voice,” 89; Geoffrey Horrocks, Greek: A History of the Language and Its Speakers (2nd ed.; Chichester: Wiley-Blackwell, 2010) 103, 130, 256.

50 Although I retain the MT’s verse division, another reading remains possible: “And may all the nations pronounce blessings by him; may they pronounce him happy.”

51 In his tentative LXX chronology, Domazakis, Neologisms, 360, locates the development of Psalms LXX somewhere between the beginning of the 2nd cent. and the 1st cent. BCE, helping to reinforce and explain Ps 71:17 LXX’s dependence on the earlier Gen 12:3b LXX.

52 Interestingly, codex A, papyrus 833, and minuscules 72´, 569, 343, and 59 replace ἐνευλογηθήσονται with εὐλογηθήσονται in Gen 12:3b. Moreover, Ps 72:17 (71:17 LXX) renders a hithpael (ויתברכו) with a future -θη- form (εὐλογηθήσονται), whereas Jer 4:2 interprets a hithpael (והתברכו) as a future active (εὐλογήσουσιν): καὶ εὐλογήσουσιν ἐν αὐτῇ ἔθνη (and nations shall bless by him [i.e., Israel]). In Jer 4:2 the nations will pronounce blessings by Israel. Here the same blessing formula appears as in Gen 12:3b; 18:18; 22:18; 26:4; 28:14; 48:20; Ps 72:17 (71:17 LXX). Unlike the others, however, Jer 4:2 expresses active voice. Like Gen 48:20, Jer 4:2 and Ps 72:17 (71:17 LXX) forgo the preverb ἐν- to εὐλογέω.

53 As noted above, the LXX does not feel the need to render ברך with a neologism in Gen 48:20. However, minuscule 57 reads εὐλογηθήσεται as ἐνευλογηθήσεται, making the link to 12:3b more explicit. Minuscule 708 reads εὐλογηθήσεται as future middle εὐλογήσεται.

54 Flury-Schölch, Abrahams Segen, 220, helpfully discusses the parallelism in the psalm’s Hebrew version.

55 The Hebrew text is from The Book of Ben Sira in Hebrew: A Text Edition of all Extant Hebrew Manuscripts and a Synopsis of All Parallel Hebrew Ben Sira Texts (ed. Pancratius C. Beentjes; Leiden: Brill, 1997).

56 See HALOT 2:510 (s.v. ל 26a).

57 Although our translations and interpretations differ, see the helpful discussion in Bradley C. Gregory, “Abraham as the Jewish Ideal: Exegetical Traditions in Sirach 44:19–21,” CBQ 70 (2008) 66–81, at 77–80. In his tentative LXX chronology, Domazakis, Neologisms, 361, locates the development of Ben Sira LXX between 132 and 117 BCE or slightly before or after 117 BCE, helping to explain the dependence of Sir 44:21 LXX on the earlier Gen 22:17 LXX.

58 All Greek quotes of Philo are from Philo (trans. F. H. Colson, G. H. Whitaker, and J. W. Earp; 9 vols.; LCL; Cambridge: Harvard University Press, 1929–1962). For recent works on Philo’s understanding of Abraham and the nations, see Phoebe Makiello, “Abraham and the Nations in the Works of Philo of Alexandria,” in Abraham, the Nations, and the Hagarites: Jewish, Christian, and Islamic Perspectives on Kinship with Abraham (ed. Martin Goodman, George H. van Kooten, and Jacques T. van Ruiten; TBN 13; Leiden: Brill, 2010) 139–61; Sean A. Adams, “Abraham in Philo of Alexandria,” in Abraham in Jewish and Early Christian Literature (ed. Sean A. Adams and Zanne Domoney-Lyttle; LSTS 93; London: T&T Clark, 2019) 75–92.

59 As seen above, εὐλογηθήσονται also appears in Philo, Her. 8; Somn. 1.3, 176.

60 Adams, “Abraham in Philo,” 76, suggests this interpretive move.

61 C. T. R. Hayward, “Abraham as Proselytizer at Beer-Sheba in the Targums of the Pentateuch,” in idem, Targums and the Transmission of Scripture into Judaism and Christianity (Studies in the Aramaic Interpretation of Scripture 10; Leiden: Brill, 2010) 17–34, at 23 n. 13.

62 For a full discussion, see Makiello, “Abraham and the Nations in Philo,” 139–44.

63 For a helpful discussion, see ibid., 144–47. At 146–47, Makiello rightly compares Abraham in Gen 12:3 with the king in Ps 72:17 (71:17 LXX), even granting reflexive meaning to the Hebrew hithpael (ויתברכו) and noticing the nations desire to reckon the king happy/blessed. But when shifting to the LXX, she reads the Greek -θη- form εὐλογηθήσονται as a passive, suggesting Philo follows the same route with Gen 12:3’s ἐνευλογηθήσονται. Thus, from the outset Makiello precludes a reading open to reflexive and middle possibilities at work in Philo’s take on Gen 12:3b.

64 In accordance with our understanding of the middle domain above, the tribes here remain involved in the verbal event expressed in the nonactive μεταχαράττονται. The verbal event here depicts a mental experience, a category pertaining to the middle domain as we see above. As a result, I render μεταχαράττονται with “changing” rather than “undergoing change.” In the English language, “undergoing change” might convey a passive reading, thus obstructing the fact that the tribes remain involved in the verbal event as the subject, whereas “changing” keeps their involvement intact.

65 Citing a variation of Gen 12:3b, Philo conveys the same point in Somn. 1.176, describing the refined human (ὁ ἀστεῖος) as offering/holding out (προτείνων) the benefit of their own accord.

66 Although Philo cites Gen 12:3b directly, since he makes intercession central, I wonder if he means Gen 18:18 here even though the quotes are slightly different.

67 For a passive reading of Gen 12:3b in Philo, Migr., see Makiello, “Abraham and the Nations in Philo,” 148–59.

68 Matthew Thiessen, Paul and the Gentile Problem (New York: Oxford University Press, 2016) 106, argues that Gal 3–4 accounts for the justification of gentiles specifically rather than of all people.

69 The three texts are Gen 12:3 (ἐνευλογηθήσονται ἐν σοὶ πᾶσαι αἱ φυλαὶ τῆς γῆς); 18:18 (ἐνευλογηθήσονται ἐν αὐτῷ πάντα τὰ ἔθνη τῆς γῆς); 22:18 (ἐνευλογηθήσονται ἐν τῷ σπέρματί σου πάντα τὰ ἔθνη τῆς γῆς). Paul combines ἐν σοὶ from 12:3 and πάντα τὰ ἔθνη from 18:18 and 22:18. Paul’s quote is most like 18:18 but does not refer to Abraham in the third person (ἐν αὐτῷ).

70 For just three recent examples of passive readings, see Thiessen, Paul and the Gentile Problem, 106–8; Matthew V. Novenson, Christ among the Messiahs: Christ Language in Paul and Messiah Language in Ancient Judaism (New York: Oxford University Press, 2012) 124–25; Paula Fredriksen, Paul: The Pagans’ Apostle (New Haven: Yale University Press, 2017) 14–15, 105. Although Fredriksen does not explicitly mention it in Gal 3:8, her reading of Gen 12:3b is passive, which then gets assumed into Paul. As someone sympathetic to all three works, I believe my reading actually reinforces their overall arguments.

71 Thiessen, Paul and the Gentile Problem, 106–7, rightly stresses this point.

72 Novenson, Christ among the Messiahs, 124–26, argues for a reading of ἐν σοὶ as “in you.” But whether we read ἐν σοὶ as “by you” or “in you” is irrelevant for my point. If we choose “in you,” a speech pronouncement still remains possible. For as we see above, Gen 48:20’s blessing formula couples ἐν ὑμῖν with an explicit blessing pronouncement, regardless of whether we read “in you” or “by you.”

73 The two texts are Gen 12:3 (ἐνευλογηθήσονται ἐν σοὶ πᾶσαι αἱ φυλαὶ τῆς γῆς) and 22:18 (ἐνευλογηθήσονται ἐν τῷ σπέρματί σου πάντα τὰ ἔθνη τῆς γῆς). Acts combines ἐν τῷ σπέρματι from 22:18 and πᾶσαι αἱ φυλαὶ τῆς γῆς from 12:3 but replaces φυλαί with πατριαί. The textual versions express variety apropos ἐνευλογηθήσονται. Codices A*, B, Ψ, minuscules 323, 945, 1739, and Irenaeus’s Latin translation (apparent reading), instead use the broader εὐλογηθήσονται. Codex C uses the newer ἐπευλογηθήσονται.

74 C. K. Barrett, A Critical and Exegetical Commentary on the Acts of the Apostles (2 vols.; ICC; Edinburgh: T&T Clark, 1994–1999) 1:212–13, makes this point.

75 Friedrich W. Blass, Albert Debrunner, and Robert W. Funk, A Greek Grammar of the New Testament and Other Early Christian Literature (Chicago: University of Chicago Press, 1961) §§308, 404.3 n. 3, read ἀποστρέφειν as intransitive (“if you turn away”) and note that ἐν τῷ introduces the content and the process of Jesus’s blessing (“to bless you in that you turn”). Although Barrett, Acts, 1:214, notes the convenient symmetry between an intransitive reading and Acts 3:19, he opts for a transitive reading (“to turn you away”), suggesting that the notion of blessing comports better with divine over human action. Pervo, Acts, 103–4, 109 n. 59, however, agrees with my reading, noting that ἀποστρέφειν is almost certainly intransitive.

76 Even if we read ἐν τῷ ἀποστρέφειν as transitive (“to turn you away”), ἐνευλογηθήσονται still expresses the middle domain. Instead of a mental activity, ἐνευλογηθήσονται would describe an event pertaining to mental process. Exhibiting patient-like rather than agent-like force, ἐνευλογηθήσονται would denote an action arising without the subject’s volition or control while still being experienced by the subject. As Jesus turns them away from their wickedness, they experience blessing.

77 See n. 34.