Propositions Concerning Ancient Greek Voice (last revised October 13, 2005)

Previous versions of my formulations on Ancient Greek Voice:
“New Observations on Ancient Greek Voice”: PDF version of  my  Nov. 19, 2002 expansive revision of my original B-Greek message of May 27, 1997,  “Observations on Ancient Greek Voice.”
“Active, Middle, and Passive: Understanding Ancient Greek Voice”  This 9-page introduction to how the different voices in ancient Greek work was posted  Dec. 16, 2003;  it is intended to give practical guidance based upon principles set forth in the longer article indicated above, "New Observations on Ancient Greek Voice.”
Theta_Passive.pdf   (July 5, 2005) Extracts from Pierre Chantraine and Andrew Sihler on the origin of the aorist intransitive and “passive” verb forms, here made available as documentation for “New Observations on Ancient Greek Voice” above.

Pending completion of my expanded and revised “New Observations on Ancient Greek Voice,” I intend to post here more recent and hopefully clearer formulations of the chief propositions that I am trying to urge the community of students and instructors of ancient Greek to adopt in the hopes of coming to a better understanding and a more useful way of talking about the verbal voice system of ancient Greek. Links to PDF versions of my earlier work are posted at the foot of these propositions, but in the meantime, I hope to set forth in brief statements the chief notions that I think are fundamental to understanding these matters.

 1. Common use of the term “active” can be confusing or misleading: while it may simply indicate that the grammatical subject of a verb is acting or performing the action, it may seem to imply that the verb must be transitive and take an object. That is confusing or misleading when one speaks of a Greek verb form as being middle or passive in voice but having an “active” meaning. In fact, middle and passive verb-forms in Greek may be transitive and take direct objects, but it is perhaps more common that they are intransitive.

2. The terms “deponent” and “deponency” are not useful in a discussion of ancient Greek voice, a fact that has been noted at least since A. T. Robertson’s big NT Greek Grammar. Conventionally the term has been used to refer to verb forms that have middle or passive voice forms but no active voice forms, e.g. δύναμαι, ἀποκρίνομαι, πορεύομαι. A distinction commonly employed is that between “middle deponents” (verbs with middle forms in the present tense and in the aorist tense, e.g. μάχομαι, ἐμαχεσάμην) and “passive deponents” (verbs with middle forms in the present tense and passive forms in the aorist tense, e.g. πορεύομαι, ἐπορεύθην). Some verbs having active forms in the present tense but future tense forms that are middle (e.g. μανθάνω, μαθήσομαι) are often referred to as “future deponents.” The term is sometimes also employed to refer to verbs that have middle or passive voice forms in the present tense but active forms in the aorist or perfect tense (e.g. ἔρχομαι, ἦλθον, ἐλήλυθα; γί(γ)νομαι, γέγονα). The term “deponent” has been explained as deriving from the participle of the Latin verb depono, deponens with the sense “setting aside” or “laying aside,” the idea being that “deponent” verbs “set aside” or “lay aside” their present-tense forms. Although the term “deponent” seems to derive from Latin grammatical descriptions of Latin verb forms, it is really questionable whether the term is properly applicable even to Latin verbs — but that is another matter altogether. My objection to the terms “deponent” and “deponency” is that they seem to imply that Greek verbs having no active present tense forms do not conform to standard Greek patterns of morphology or that they are somehow irregular, while I would contend that these verbs are so numerous in ancient Greek that they should not be deemed less standard than the admittedly larger body of verbs that do have active present tense forms. Alternatively the term “defective” is sometimes used for verbs that lack forms in one or more of the morphological paradigms of the ancient Greek verb; that might be a better term, but I think it would be preferable to acknowledge that a great many of the verbs in common everyday usage are “irregular” in that they do not display the full array of verb-forms one sees in the paradigms of λύω or παιδεύω in the appendices of primers of ancient Greek.

3.  In lieu of the term “deponent” I would suggest that we speak of “middle verbs.” I would use this term for verbs whose primary present-tense form is middle-voice; that would include all the verbs that are traditionally or conventionally termed “deponent” but it would also include a sizable number of essentially intransitive verbs that display common middle-voice present-tense forms but also have a transitive active-voice form in the present tense, e.g. ἐγείρομαι “awake” or “rise up” with present active ἐγείρω “awaken” or “rouse” or “raise up,” or ἵσταμαι “stand up” with present active ἵστημι “cause to stand” or “establish,” ἀγείρομαι “assemble together” with present causative active ἀγείρω “bring together” or “cause to gather.”

4.  The verbal infix -θη-is traditionally or conventionally understood as the morpheme signifying passivity in aorist verb-forms in -θην, θης, θη, θημεν, θητε, θησαν and in future verb-forms in -θήσομαι, θήσῃ, θήσεται, θήμεθα, θήσασθε, θήσονται. It should be noted, however, that the future forms in -θήσομαι, κτλ. are derivative secondarily from aorists in -θην κτλ. that are conjugated with active endings and that are formally identically with non-thematic aorist active voice forms such as ἔβην, ἔστην. Indeed, the so-called “second passives” are clearly older and formally identical with these non-thematic aorist active voice forms, e.g. ἐφάνην “I appeared” (or “I was made to appear”), ἐβλάβην (“I got hurt”). While these forms in -θη-have, as I said above, traditionally or conventionally been deemed as markers for passive forms and meaning, they are essentially intransitive and were never used exclusively to express passive sense but rather to form normally intransitive aorist forms that could represent the aorist for either  “middle” verbs (ἠγέρθην “I rose,” aorist of ἐγείρομαι) or “passive” expressions of transitive (causative) active verbs (ἐποιήθη “was created,” aorist of ποιέω). While in fact the greater part of -θη- forms in ancient Greek do represent passive semantic force because they appear in the aorist- or future-tense forms of transitive causative verbs, nevertheless a very great number of the -θη- forms are simply the intrasitive aorist- or future-tense forms of “middle” verbs as defined in §3 above.

5.  Replacement of older aorist middle-passive forms by “passive” θη/η forms: It is important to understand that over the course of time aorist middle-voice forms in -μην.  -σο. -το and future middle-voice forms in -σομαι, -σῇ, -σεται came to be supplanted by --θη- forms in -θην, -θης, -θη (aorist) and -θήσομαι, -θήσῃ, -θήσεται (future). This process began early and is already apparent in Classical Attic in the verbs that have conventionally and traditionally called “passive deponents.” In the κοινή one may readily recognize this process in alternative forms in the LXX and the Greek NT such as ἀπεκρινάμην = ἀπεκρίθην and ἐγενόμην = ἐγενήθην. This process continued on in the course of the development of Byzantine and later forms of the Greek language.

6. “Active” and “Middle-Passive” paradigms in Proto Indo-European and in early Greek — the voice system in earliest Greek (prior to the development of  the standard verbal paradigms of Classical Attic): this needs to be set forth with sufficient clarity to indicate Homeric usage of the morphoparadigms, both their forms and their semantic functions. An appendix in any final version of the “New Observations” should list Homeric verbs displaying middle-voice forms with passive meanings and passive-voice (-θη-) forms with middle meanings.

7. “Middle” verbs: Suzanne Kemmer has shown that the middle voice is a linguistic category found in numerous languages spoken and written in the world, by no means Indo-European languages only, and she has shown that middle-voice semantics find quite varied expression in different languages: in distinct middle morphoparadigms in some, in reflexive forms in many others (as in the Romance languages, also in German); in Latin the -r, -ris, tur, -mur, -mini, endings commonly called passive really serve for Latin the same function as the -μαι, -σαι, -ται, -μην, -σο, -το forms in Greek -- that is to say, they convey both middle AND passive semantic force, even if it is most commonly passive. But there's nothing odd about fruor, patior, fungor, potior, and vescor -- they are authentic middle voice forms. My guess is that Hebrew Niphal and Hithpael forms would fall into the category of modes of grammaticalizing middle-voice semantics, just as in English forms with "get" and the participle have for some time become comparable modes.

8. Categories of “Middle” verbs:

        Suzanne Kemmer  offers the following categories of verbs that typically find expression in middle-voice morphology. I present a compilation from her listings and the illustrative verbs in Greek are, for the most part, my own compilation:

1. Grooming, body care: washing, dressing, shaving, bathing, undressing, cutting hair/nails, anointing self, buttoning clothes (Fr. se laver, Gr. λούομαι)
2. Change in body posture: sit down, lie down, kneel down, arise, stand up (Gr. κεῖμαι, ἕζομαι, ἐγείρομαι, ἀνίσταμαι)
3. Non-translational motion: stretch out, turn around, bend, shake head, move neck, clench fists (Gr. τρέπομαι, κάμπτομαι)
4. Translational motion: fly, flee, go away, run, hurry, go away from, climb, arrive, leave, come, go (Gr. πέτομαι, ἔρχομαι, πορεύομαι, Fr. s’en aller, Lat. se removere)
5. Indirect reflexive: break one’s arm, etc.; build oneself a house
6. Indirect middle, self-benefactive: choose, acquire for self, pray, attain, reach (Gr. αἱρέομαι, κτάομαι, εὔχομαι, ὀρέγομαι)
7. Naturally reciprocal events: meet, fight, greet, wrestle, embrace, quarrel, converse, agree with, mate, take stock together (Gr. ἀσπάζομαι)
8. Stative, naturally reciprocal: adjoin, be linked (copular), resemble one another, match (Gr. ἔχομαι)
9. Reciprocal marked naturally collective: assembling, gather, merge, be packed, accumulate, multiply, thin out, densify
10. Emotion: be frightened, hate, be angry, marvel at, delight in, take consolation, pity, care/worry, grieve/mourn, regret, be charmed, take pleasure, repent, be satisfied (Gr. ἥδομαι, Lat. misereor)
11. Emotive speech: complain, lament, blame (Gr. ὀλοφύρομαι, μέμφομαι)
12. Other speech actions: confess, boast, chide, accuse, lie, deceive, threaten, refuse (Gr. καυχάομι, αἰτιάομιαι, ὁμολογέομαι)
13. Simple cognitive events: ponder, meditate, interpret, devise, reflect, consider, deliberate, calculate (Gr. ἡγέομαι, βουλεύομαι, σταθμάομαι)
14. Perception: look at, feel, hear (Gr. δέρχομαι, αἰσθάνομαι, ἀκροάομαι, Lat. conspicor)
15. Complex cognitive events: believe, forget, remember, suspect, conjecture (Gr. οἴομαι, ἐπιλανθάνομαι, Lat. obliviscor)
16. Commissive, intensive: resolve, view, undertake, promise, intend, swear (Gr. ὑπιχνέομαι)
17. Spontaneous events: die, sink, develop, become light, change, dissolve, evaporate, germinate, dissipate, grow, burst, spread out, convalesce, thaw, melt, open, split, be born (a very large group, the mother of all of which is γίγνομαι)
18. Facilitative: inherent characteristic of patient allows action to take place: “soup eats like a meal."
19. Impersonal: generic agent (both this and the above have generic aspects).
        
        Neva Miller has drawn up her own categories of what I have called “subject-focused” verbs and placed into them just about all, if not absolutely all, the so-called “deponent” verbs found in the Greek New Testament. Whether or not one finds these lists of Kemmer and Miller fully satisfactory, I think they are very instructive for our understanding of what these verbs are and what they have in common.

Class 1: Reciprocity
A. Positive [i.e. friendly] Interaction (δέχομαι, δωρέομαι, εἰσκαλέομαι, ἐναγκαλίζομαι, ἐπιμελέομαι, ἐπισκέπτομαι, ἰάομαι, ἱλάσκομαι, χυναντιλαμβάνομαι, χαρίζομαι)
B. Negative [i.e. hostile] Interaction (διαμάχομαι, δράσσομαι, ἐνάλλομαι, ἐπαγωνίζομαι, ἐπιλαμβάνομαι, μάχομαι, μέμφομαι)
C. Positive and Negative Communication (αἰτιάομαι, ἀνατάσσομαι, ἀποκρίνομαι, ἀρνέομαι, ἀσπάζομαι, διαβεβαιόομαι, διακατελέγχομαι, διαμαρτύρομαι, ἐξηγέομαι, μαρτύρομαι, παραιτέομαι, ψεύδομαι)
Class 2: Reflexivity (ἀπολογέομαι, ἐγκαυχάομαι, ἐγκρατεύομαι, ἀπαναπαύομαι, μασάομαι, μιμέομαι, πειράομαι), directional: ἀφικνέομαι, διαπορεύομαι, διεξέρχομαι, ἐξάλλομαι, ἐπανέρχομαι, ἐπεκτείνομαι, ἔρχομαι, ὀρχέομαι, πορεύομαι
Class 3: Self-Involvement
A. Intellectual Activities (αἰσθάνομαι, διαλογίζομαι, ἐπυπνιάζομαι, ἐπιλανθάνομαι, ἐπίσταμαι, ἡγέομαι, λογίζομαι, οἴομαι, πυνθάνομαι)
B. Emotional States (βδελύσσομαι, διαπονέομαι, ἐμβριμάομαι, ἐμμαίνομαι, εὐλαβέομαι, μετεωρίζομαι, ὀμείρομαι)
C. Volitional activities (βούλομαι, ἐναντιόομαι)
Class 4: Self-Interest (διαδέχομαι, διαπραγματεύομαι, κτάομαι, ἐμπορεύομαι, ἐργάζομαι)
Class 5: Receptivity (γεύομαι, ἐπακροάομαι, θεάομαι)
Class 6: Passivity (γίνομαι, ἐπιγίνομαι, κοιμάομαι, μαίνομαι, μαντεύομαι)
Class 7: State, Condition (δύναμαι, ἐπίκειμαι, καθέζομαι, κάθημαι, κεῖμαι, παράκειμαι)

9. Transitivity and Causativity as factors bearing on the semantics of the morphological paradigms for voice in Greek. Egbert Bakker (Egbert J. Bakker, “Voice, Aspect and Aktionsart: Middle and Passive in Ancient Greek” in Barbara A. Fox, Paul J. Hopper, edd., Voice: Form and function (Typological Studies in Language 27) (Amsterdam/Philadelphia: John Benjamins Publishing Company, 1994), pp. 23-47) has clarified to an immeasurable extent how the Aktionsart of ancient Greek verbs bears upon the way in which the aorist forms in -σα and in -θην should be understood in terms of semantic voice.

10. Primitive verbs with aorist and perfect forms that are “active” in form, are generally intransitive; they correspond to verbs that are (or should be) lemmatized as “Middle” verbs: It has been observed by some that a full accounting of what have traditionally been called "deponent" verbs ought to clarify why it is that some "middle" verbs have active forms in the aorist and in the perfect tenses. For instance ἔρχομαι has a middle future derived from a different root (ϝελευθ) -- a root that appears in the aorist (ἦλθον) active and in the perfect (ἐλήλυθα) active forms; another is ἵσταμαι (usually lemmatized in the active form ἵστημι): there is an aorist active form ἔστην which is intransitive and a perfect active form ἕστηκα which is also intransitive. Why? The reason is that these aorist and perfect forms of these "primitive" verbs emerged long before the development of the middle-passive perfect tense and the distinctive -θη- forms of the aorist. They are not really "active" in meaning -- they are certainly not transitive and causative -- but rather they exemplify the "default" function of the "active,"  the form that is not distinctly marked for subject-focus.  Some of these verbs have causative transitive aorists in -σα, e.g. ἔστησα -- but the older “second” or  “strong” aorists really correspond to present-tense forms that are middle rather than active.

11. Parsing: in my view it is not the business of a parser to determine whether a middle-passive form should be considered middle rather than passive or vice-versa; the parser should indicate the form only, so that all -μαι, -σαι, -ται, -μην, -σο, -το forms will be marked as Middle, all -θη- forms will be marked as Passive. If I had my preference, I'd mark all “Middle" verbs MP1 and all “Passive” verbs MP2, since both morphoparadigms are capable of bearing either middle or passive semantic force. I know that the Bible programs (Bible Works, Accordance, Gramcord) base their parsing for voice on a judgment call regarding these matters, but I think they are wrong to do so; James Tauber's CCAT GNT Morph has got it right; I wish that the CCAT LXX were done the same way.

12. What's really new here? Perhaps at the outset of any final version of “New Perspectives” it needs to be stated very clearly that most of what is set forth here is already clearly indicated in such standard reference works as A.T. Robertson's big NT Grammar and in Smyth's Greek Grammar: reference to the relevant sections should be cited clearly and some of those relevant sections should be excerpted liberally. What is aimed at here is the simplification and clarification of what was already, to a considerable extent, understood by at least some traditional grammarians.

13. Appendices: these should include at least (1) A listing of Homeric “middle verbs” showing aorists in -θην; (2) a listing (based on Smyth's catalogue of irregular verbs) of "primitive” verbs that are essentially middle even if they have active causative forms; (3) a complete list of all verbs in the GNT that have MP1 ("middle") and MP2 ("passive") forms.