Observations on Ancient Greek Voice (LONG!)

From: Carl W. Conrad (cwconrad@artsci.wustl.edu)
Date: Tue May 27 1997 - 06:46:36 EDT

I have been altogether appalled at some of what has been said on the
"passive" infinitive in John 3:30 in the course of the past day, appalled
and also dismayed, because I believe very strongly that much of this is a
consequence of fundamental misconceptions governing the traditional
teaching of the voices in ancient Greek and that much of it has also been
perpetuated by incorporation of these misconceptions in standard reference
works consulted by students of NT Greek. In the course of the past two
years in this forum I have stated my views on several aspects of this
misunderstanding of the ancient Greek voices but I'm not sure that I've
ever tried to put it all together. I am doing so now, hoping to be brief,
knowing that I am bucking generations of traditional teaching and that I am
not likely to win many converts to my view of voice, although I am
convinced it is correct. I should add also that very little of what I am
saying originates with me, although I haven't seen it all put together this
way anywhere hitherto.

1. The Errors of traditional conceptions of the ancient Greek voices
        I think that there are two great and fundamental misconceptions
built into the traditional way that the Greek voices are understood,
taught, and represented in paradigm and parsing authorities:
        (a) Whether it derives from a sense of English grammar or perhaps
originally from an understanding (a misunderstanding, I'd say) of Latin
grammar, there is a mistaken assumption that the fundamental logical
opposition of the voice system is active/passive and that the middle voice
is an "intruder" that serves only to muddy and confound the understanding
of the voices. The very term "middle" seems to reinforce this misconception
because it apparently implies that the perspective on action represented by
the "middle" morphology is somewhere halfway between what is represented by
the Active and what is represented by the Passive. As a result students are
taught and come to learn (wrongly) that the middle/passive endings of
finite verbs, infinitives, and participles really express a Passive meaning
fundamentally, and that the Middle meaning is a rare and secondary notion
sometimes accruing to these forms. This is a serious and far-reaching
misunderstanding of the middle/passive endings and of their function.
        (b) The other chief factor in traditional terminology, teaching and
learning of the voices is the use of the term "deponent" to refer to verb
forms which are Middle or Passive in Greek but express verbal notions
represented in English (or whatever one's target language may be: Latin,
German, etc.) as either Active or Intransitive. Thus, for instance,
AISQANOMAI which may take a direct complement and which is translated into
English with an ACTIVE verb "I perceive, sense, feel," and similarly
ERCOMAI which is a verb of motion and which is translated into English as
an INTRANSITIVE verb "I come/go" are both called "deponent verbs"--why? I
can think of no other reason than the fact that their morphology is thought
to be abnormal--because it is alien to the common usage of English (or of
Latin, or whatever other language one thinks in and therefore deems
"normal/natural"). Another category where this term "deponent" is applied
in a misleading and confusing way is with regard to verbs that have an
aorist in -QH- and a future in -QHSOMAI; while most Greek verbs thus
inflected are in fact aorist and future PASSIVES, quite a few of them are
NOT Passives, as for instance, EBOULHQHN and BOULHQHSOMAI which are simple
aorist and future forms respectively of a verb BOULOMAI meaning "wish" or
"want"--but most traditional grammars and reference works will call this a
"Passive Deponent" and draw a distinction between it and other "deponent"
verbs that have Middle forms in the aorist as well as in the present and
imperfect (and perhaps perfect also) tenses but that have meanings which
are translated in English by active or intransitive verbs, e.g. MACOMAI,
aor. EMACESAMHN, fut. MACOUMAI, "I fight/do battle."

2. The Passive Voice is a Secondary and not altogether Normal Greek Verb
        (a) Indo-European never had a distinct passive voice morphology,
although it may possible have used the middle-voice forms to express
passive understandings of verbal action. But the real opposition of the
Indo-European verb is between Active and Middle (which I personally would
far prefer to call "reflexive" because it always involves or appears to
involve some notion of self-projection and also because "middle" seems to
imply that the natural antithesis of the Active is something other than the
Middle, which is not true or at least not true of the Indo-European verb.
Let me illustrate this by reference to a couple well-known facts of the
Romance Languages, facts best explained, in my opinion, by the resurgence
of a native Indo-European middle/reflexive voice:
        (1) While the antithesis of active and passive does seem to have
become normative in Latin (as it never really did completely become in
Greek), what is truly remarkable is the historical emergence of great
numbers of reflexive verbs over the course of time, especially in late or
"Vulgar" Latin, a development which accounts for the massive number of
reflexive verbs in French, Spanish, and Italian that gives English-speaking
students the same kinds of headaches as to the so-called "deponent" verbs
of Greek and Latin. While one can readily make sense, say, of French, "je
me lave les mains" for "I wash my hands" (even though it literally means "I
wash myself the hands", much like Greek LOUOMAI TAS CEIRAS), one who first
learns something like "elle s'en est allŽe" has a hard time making sense of
what appears to mean, in "literal" English, "she went herself from there."
        (2) In Miami and New Orleans, and doubtless in southwestern U.S.
border towns too one may read on store windows the phrase, "Aqui se habla
espa–ol," literally "Spanish speaks itself here," although normally
translated with a passive in English, "Spanish is spoken here." Students of
French and Spanish are well aware of numerous such expressions wherein a
reflexive verb construction serves to express the same idea as English
expresses with a passive construction. Who is to say that the English
passive is more "normal and natural" than the French or Spanish reflexive?
The reflexive of French and Spanish is actually better representative of
the original voice system of Indo-European.

        (b) The so-called "middle/passive" endings in the present,
imperfect, aorist and perfect tenses are fundamentally reflexive ("middle")
in meaning and only secondarily came to assume a common passive function. I
don't want to argue this at length; I shall use a couple examples to
illustrate how the distinct passive originated, in my opinion:
        ELUEN TON hIPPON MOU "He was untying my horse"
        ELUOMHN TON hIPPON "I was untying my horse"
        ELUETO hO hIPPOS MOU "My horse was getting loose"
        ELUETO hO hIPPOS MOU hUPO TOU FILOU MOU "My horse was being untied
by my friend"
        hH DOULH LOUEI TOUS PODAS MOU "The maid is washing my feet"
        LOUOMAI TOUS PODAS "I am washing my feet"
        LOUONTAI hOI PODES MOU "My feet are washing themselves" (I think
that's what this really means, although most people will want to put this
into normal English and say, "My feet are being washed.")
        LOUONTAI hOI PODES MOU hUPO THS DOULHS "My feet are being washed by
the maid."
        I would contend that it is only the presence of hUPO + genitive,
which became the standard agent construction in ancient Greek, that the
middle/reflexive forms here are actually recognizable as having a
distinctive passive function. Moreover, I would personally say that UNLESS
one can at least implicitly discern an agent (hUPO + genitive) or perhaps
an instrumental dative (which is not nearly so clearly an indicator of
passive voice, since it might be used even with a reflexive verb), then one
would do best to identify the verb in question as middle/reflexive rather
than passive. THAT IS: I would urge pedagogically that students not learn
to call the -MAI, -SAI, -TAI, -MHN, -SO, -TO endings "passive"
automatically, but either "middle/passive" or even preferablly

        (c) What I have already said in (b) refers essentially to
transitive verbs that have both active and middle/reflexive forms. I would
also point out, however, that this understanding of middle/reflexive forms
may be helpful for dealing with the apparent "anomalies" of the so-called
"deponent" verbs like ERCOMAI and POREUOMAI which have no active forms
normally at all but which we would term Intransitive verbs of motion.
Perhaps it's perilous to attempt to get very deep into the underlying
psychology of the Greek verb, but I would be inclined at least to
understand these verbs as originally involving a notion of self-projection,
self-propulsion, and that this is the reason why their basic forms are
middle/reflexive rather than active. Of course, it is certainly true also
that there are intransitive verbs with active forms, and in fact, HLQON
functions as the ACTIVE aorist of ERCOMAI and BAINW or PROBAINW may be used
to mean much the same thing as POREUOMAI (at least when one is traveling on
foot!). I think the same psychology is at work in the fact that so many
Greek verbs that are active in the present tense go into the
middle/reflexive in the future--they apparently involve a notion of
self-projection or self-propulsion in the thought of the agent, e.g.:
        PASCW, PEISOMAI (from root PAQ/PENQ)
and I think anyone competent in classical or Attic Greek could readily add
many others to this list. It's rather curious to me that the verb EIMI,
which is certainly active in form in the present and imperfect in classical
Attic and Homer, has a future ESOMAI (Homeric ESSOMAI) and already in
Hellenistic Greek shows (with imperfect HMHN) that it is on its way to the
modern Greek verb which is conjugated wholly in the middle/reflexive
(EIMAI, HMOUN, etc.).

        (d) -QH- forms did indeed BECOME standard passive morphology in the
classical period and continued to be that in the Koine. BUT this infix -QH-
does not appear to have had originally and association with passivity, and
the considerable number of Greek verbs with aorist in -QHN and future in
-QHSOMAI that do not have any passive meaning and that are traditionally
called "passive deponents" is, I believe, a pretty clear indicator that
-QH- does not have any fundamental and clear semantic association with
verbal passivity. I would content that, just as there are varied elements
added to a verb root that may be employed to create present stems, so -QH-
was one element added to verb roots used to form ATHEMATIC (SECOND/THIRD)
AORISTs. That is to say, there is no significant formal difference between
aorists--and all use, it should be noted, ACTIVE secondary endings--and
there is nothing PASSIVE in HDUNHQHN or hHSQHN (any more than there is
        I cannot prove my surmise, but I rather suspect that a process
similar to that outlined above for passive usage with middle/reflexive
primary and secondary endings may also have been in play in the gradual
emergence of the -QH- passive morphology. For example:
        EFANHSAN hOI STRATIWTAI "The soldiers appeared/showed up"
        EFANH TAUTA TOIS PAROUSI "These things became evident to those present"
revealed to those present by the orator."
        Here too, as earlier in the case of the primary and secondary
middle/reflexive forms, I would be somewhat hesitant to call a form
definitely passive unless there is a pretty clearly implicit or explicit
hUPO + agent construction or an instrumental dative. Here, for instance,
belongs Randy Leedy's ambivalent HGERQH which may be "was raised" (i.e.
genuinely passive) IF one findss an explicit or clearly implicit hUPO TOU
QEOU in the form HGERQH. Randy might be able to assert that HGERQH is BOTH
intransitive ("he arose") and passive ("he was raised") in this instance,
on grounds that "the line between intransitive and passive is so thin as to
be evanescent," but I see no reason to fudge the difference between what is
clearly passive and what is clearly intransitive (although I'll grant that
one may have no clear grounds for deciding in some cases whether HGERQH is
intransitive or passive, I certainly wouldn't try to affirm that it is
        I won't try to spell out the fact that what's been said about the
-QH-forms applies just as well to the future tense -QHSOMAI, etc. forms
Some, and probably MOST of the Hellenistic and Classical Attic -QHSOMAI
forms should be understood as passive, BUT one really ought to avoid saying
stupid things such as that DUNHQHSOMAI is "passive in form but active in
meaning"--the fact is that it is not really "passive in form" at all--it's
simply a future built upon the athematic aorist stem.

I have no way of knowing how successful this attempt to formulate a clearer
statement about the ancient Greek passive (and middle/reflexive) may have
been. There are probably numerous gaps that ought to be filled in, but I
think the most essential elements at least have been stated.

I would repeat what I said yesterday about ELATTOUSQAI in John 3:30. In my
own mind there's no question that this is middle/reflexive "to become
lesser," but I grant that it may be hard to accept that view in the face of
the mass of tradition and authoritative parsing guides. I would add that I
quite agree with Carlton Winbery that such a form could (assuming one has
reason to understand an agent construction implicit) be understood as
passive. Nevertheless, I would reaffirm my original proposition, such
forms should be taken normally as middle/reflexive and as passive only when
there is a good reason to see passivity in the particular instance.

Carl W. Conrad
Department of Classics, Washington University
One Brookings Drive, St. Louis, MO, USA 63130
(314) 935-4018
cwconrad@artsci.wustl.edu OR cwc@oui.com
WWW: http://www.artsci.wustl.edu/~cwconrad/

This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:38:17 EDT