What reasons are there for a Greek speaker to use a reflexive pronoun with a verb rather than the middle voice? If we assume, as most traditional approaches to Greek voice do, that the middle voice is, at its core, reflexive, then this becomes a very interesting question. Consider: Luke 10:29 or Titus 2:14.
- Ὁ δὲ θέλων δικαιῶσαι ἑαυτὸν εἶπεν πρὸς τὸν Ἰησοῦν· Καὶ τίς ἐστίν μου πλησίον;
But wanting to justify himself, he said to Jesus, “And who is my neighbor?”
- ὃς [Ἰησοῦς Χριστός] ἔδωκεν ἑαυτὸν ὑπὲρ ἡμῶν
who [Jesus Christ] gave himself on our behalf.
The first essential grammatical point to understand here is that neither of these two verbs (δικαιόω and δίδωμι, respectively) fall into any of the natural semantic classes of middle verbs (see Allan 2003,
Aubrey 2016, Kemmer 1993): spontaneous process, body action, perception, cognition, etc. Most of the time, those semantic classes are the primary determiner of whether middle morphology would be acceptable, “naturally middle semantics” you could call it.
For naturally middle events related to pronoun types, the big ones are:
- Naturally reflexive events (esp. grooming, & other body actions)
Naturally reciprocal events (receiving, buying, selling, etc.)
Most transitive verbs function within a different alternation: Caused vs. Non-caused action. Consider the verbs in the table below:
|ἀνακλίνω||cause to recline, cause to lie down||ἀνακλίνομαι||recline to eat|
|ἀναπαύω||make someone rest, have relief||ἀναπαύομαι||to rest, to find relief|
|ἀποκόπτω||to cut something, someone||ἀποκόπτομαι||to get castrated|
|βόσκω||cause cattle to eat, feed||βόσκομαι||to graze|
|βυθίζω||cause to sink||βυθίζομαι||sink (intrans.)|
|γενεαλογέω||trace a lineage||γενεαλογέομαι||be a descendant|
|ἐκχέω||cause to pour out||ἐκχέομαι||spill out|
|ἐγείρω||set/stand something up||ἐγείρομαι||stand up, rise (intrans.)|
|εὐοδόω||cause to prosper||εὐοδόομαι||prosper|
|πνίγω||strangle, choke (trans.)||πνίγομαι||drown, choke (intrans.)|
|σείω||to shake smth.||σείομαι||quake, shiver/shake (intrans.)|
|σκορπίζω||make hard||σκορπίζομαι||become hard|
|σκοτίζω||make dark||σκοτίζομαι||become dark|
|τήκω||melt smth (trans.)||τήκομαι||melt (intrans.)|
|φέρω||bring, bear, carry||φέρομαι||move, come|
|χωρίζω||cause separation, divide||χωρίζμαι||leave a group (separate oneself)|
|ψύχω||make cold, dry||ψύχομαι||become cold, dry|
While the semantic variation amongst these verbs might feel fairly wide, they do share are change in the number of participants and a shift in where the energy of the event is directed (Aubrey 2016). External causation is removed from the event’s structure and replaced with an internal, and often spontaneous, change or process.
The event structure and required participants of an active relative to its middle form needs to be accounted for in the use of the reflexive. Δικαιόω is a transitive verb that does not fit this pattern. The lexical semantics of the verb assumes an event structure wherein a judge declares an acquittal verdict on behalf of the defendant (Aubrey & Aubrey 2017).
In the case of δικαιῶσαι ἑαυτὸν, you have an event that leans more toward to the transitive prototype, where a judge declares an acquittal verdict on behalf of the defendant. Within that frame, there are necessarily two distinct participants: judge, defendant. A defendant declaring himself innocent would be an unexpected or even unnatural event structure given the requirements of the cultural frame. The judge and defendant function as two participant roles in the frame that do not lend themselves to merging together particularly well into a single participant in the same way that a natural middle event is. The participants are highly individuated (Kemmer 1993, 65-7; 2002, 108-9). We may view the distinguishability of participants as a scale—here Kemmer emphasizes distinguishability as relative (66). The roles of judge and defendant are quite high on that scale—highly distinct, a natural product of how they function within the cultural construct of the courtroom.
On the other side, there’s a coherence to the two participants in the active of ἐγείρω (Ι raise something up = cause it to be stood up) becoming one participant in the middle ἐγείρομαι (I stand up): raiser and raisee are readily mapped onto each other. Similarly, with νίπτω (I wash someone else = cause someone to be washed) vs. νίπτομαι (I get washed/wash myself), the agent and patient in the washing event may readily correspond with each other. But with δικαιόω, the roles of judge and defendant do not comfortably merge together at all. The Greek reflexive pronoun, then, provides a useful means for maintaining a greater degree of distinguishability and individuation. Then also, to an extent, it signals to the audience that this isn’t a normal transitive two participant situation. The same is true of ἔδωκεν ἑαυτὸν. ‘Giving’ is a verb that where the event frame obligatorily requires highly distinguishable participants. Since transference frames do adequately allow the merger of two participants into a single participant, middle voice morphology is not really a viable option for the event structure.
Of course, there are verbs that fall into the natural middle classes where a verb is still used in the active with a reflexive. This is a much smaller set. These usages have a different motivation that perhaps will be explored another time. But one such usage embodies the story of Good Friday and Easter well:
- ἐταπείνωσεν ἑαυτὸν γενόμενος ὑπήκοος μέχρι θανάτου, θανάτου δὲ σταυροῦ διὸ καὶ ὁ θεὸς αὐτὸν ὑπερύψωσεν, καὶ ἐχαρίσατο αὐτῷ τὸ ὄνομα τὸ ὑπὲρ πᾶν ὄνομα ἵνα ἐν τῷ ὀνόματι Ἰησοῦ πᾶν γόνυ κάμψῃ ἐπουρανίων καὶ ἐπιγείων καὶ καταχθονίων καὶ πᾶσα γλῶσσα ἐξομολογήσηται ὅτι κύριος Ἰησοῦς Χριστὸς εἰς δόξαν θεοῦ πατρός.
He humbled himself, becoming obedient to the point of death–even death on a cross. For this reason, God exalted him and graciously granted him the name that is above every name, so that at the name of Jesus every knee will bow in heaven and on earth and beneath the earth and every tongue declare that Jesus Christ is Lord to the glory of God the Father. (Phil 2:8-11).
(Rachel Aubrey also contributed to this discussion)
Allan, Rutgar. 2003. Middle voice in Ancient Greek: A study in monosemy. Amsterdam: Brill.
Aubrey, 2016. “Motivated categories, middle voice, and passive morphology.” Pages 563-625. The Greek verb revisisted: A fresh approach for Biblical exegesis. Edited by Steven Runge and Christopher Fresch. Bellingham, WA: Lexham Press.
Aubrey, Michael and Rachel Aubrey 2017. “Preposition usage and δικαιόω: A case-study of passivization.” Paper presented at the Greek prepositions workshop. Tyndale House, Cambridge.
Kemmer, Suzanne. 1993. Middle Voice. Amsterdam: John Bejamins. Kemmer, Suzanne. 2003. “Human cognition and the elaboration of events: Some universal conceptual categories.” Pages 89-118. The new psychology of language: Cognitive and functional approaches to language structure. Volume 2. Edited by Michael Tomasello. Mahwah, NJ: Lawrence Erlbaum Associates.