[b-greek] Response to Ward Powers re Voice (2)

From: Carl W. Conrad (cwconrad@artsci.wustl.edu)
Date: Thu Nov 08 2001 - 10:02:46 EST

Here's the rest of my response to Ward's 20K message of 12:21 AM +1100
11/7/01 with subject-header "Re: MIDDLE AND PASSIVE VOICE"

At 12:21 AM +1100 11/7/01, B. Ward Powers wrote:
 . . .
>Again, Carl says:
><<(i) there appears to be sufficient historical evidence that the -QH-
>morphoparadigms increasingly came to supplant the older -MHN/SO/TO
>morphoparadigms in the aorist and the future tenses, although the older
>morphoparadigms in MHN/SO/TO continued in use in the aorist and future for
>many verbs in more common everyday use in the language;
>(ii) the GNT database displays no more than three handfuls of verbs that
>have BOTH the MHN/SO/TO AND the -QH- morphoparadigms in the aorist: and
>some of these are clearly or at least arguably identical in meaning
>-QH- morphoparadigm had already become the "standard" one for aorist
>middle-passive, and that the -MHN/SO/TO morphoparadigm survived and
>continued in use as a "strong" form simply because the verbs employing it
>were in more common usage among Greek speakers and writers.>>

And Ward asks:
>Yes: but why?? Here is the significant question: Why should these new -QH-
>forms emerge, and be so successful in being adopted "in the course of the
>historical growth and development of Ancient Greek"? I cannot recall Carl
>having given any answer for this, beyond drawing our attention to the fact
>that languages do change and develop.
>However, I have an explanation to offer: These "new" forms in -QH-
>developed and were used more and more widely, precisely because they met a
>need: they permitted Greek speakers to make a plain distinction between
>when they wished to convey middle meaning and passive meaning in aorist and
>future tenses. That is, the "new" paradigms became popular because they
>fulfilled a role: and that role was to differentiate the middle and passive

then adds:

>The wonder is that this morphological differentiation device between
>middle and passive did not spread to the present and perfect verb
>subsystems also, not that it became so widely used in the aorist and future
>subsystems. (Last month the question was posed on b-greek as to why the
>-QH- forms (or some other equivalent) were only developed for two of the
>four subsystems, and ideas and explanations for this were invited: if I
>recall correctly, there was not a single explanation tendered.)

I thought I HAD offered an explanation, and I shall re-state it, if I
didn't get it stated clearly before. But first, Ward's additional comment
that I've just cited contradicts the logic of the very reason which he
offers for the successful adoption of the -QH- morphoparadigms: if the
success is to be explained first and foremost by this morphoparadigm's
differentiation between the middle and passive, then WHY indeed did it NOT
spread to the other tense systems? I would say (a) that the differentiation
between middle and passive was not so important to Greek-speakers/writers
that they felt it necessary to have distinct morphoparadigms in the other
tenses, and (b) distinction of middle from passive wasn't the first and
foremost reason for adoption of the -QH- morphology.

What I've said about the origin and spread of the -QH- forms is essentially
as follows; I hope I can make a bit clearer my sense of how these changes
may have or probably occurred:

(a) In addition to older second-aorist thematic forms in -OMHN/ESO/ETO
which once were, I believe, as much middle AND passive in meaning as the
MAI/SAI/TAI forms in other tenses, there was a "third" non-thematic aorist
which tended to have intransitive or even "quasi-passive" semantic
functions. Thus for hISTAMAI and its compounds there was ESTHN and its
compounds, and for FAINOMAI there was EFANHN. A form such as KATESTHN
STRATHGOS from KAQISTHMI/KAQISTAMAI might mean "I became Strategos" or "I
was elected Strategos" (the only public office that Pericles ever held was
Strategos), while a form such as EFANH TOUTO TO SHMEION might mean "This
omen appeared" or "This omen was revealed."

(b) With the growing adoption of the sigmatic or first aorist active
morpholoparadigm, the opposition of voice-forms in the aorist for verbs
such as FAINW/FAINOMAI and hISTHMI/hISTAMAI became neatly:
"active/causative" EFHNA (originally EFANSA) and ESTHSA,
"intransitive/quasi-passive" EFANHN and ESTHN.

(c) At some point (and since there are a few such forms in Homer, it must
have been a fairly early but not immediately sweeping development), an
extended form of this morphoparadigm (long-vowel, normally Eta, secondary
ACTIVE endings: N/S/_/MEN/TE/SAN) emerged in -QH- conjugated in the same
fashion as the -HN/HS/H type. My hypothesis (and I honestly don't think I
invented this but I'm trying to show why understanding it may help us
understand the ambivalence of the -QH- morphoparadigm) is that these -QH-
aorist forms probably spread in common usage among Greek-speakers in
roughly the same time frame that the Sigmatic active aorists in -SA spread.
And of course, I'm also saying that these -QH- forms carried the same
ambivalence as the -HN/HS/H forms: they were
"intransitive/quasi-passive"--they might even convey "middle" semantics.

(d) It appears to be the case (but the evidence needs to be gathered to see
whether this is really the case) that as the -QH- aorist morphoparadigms
came increasingly into use, the older -MHN/SO/TO aorist morphoparadigm of
the same verbs became obsolete. It appears that there are VERY FEW KOINE
GREEK VERBS--certainly very few in the GNT--that display BOTH -MHN/SO/TO
aorists and -QH- aorists for non-active forms.

(e) Future-tense forms in Greek may be based upon a present stem, but
actually the present stem is most commonly formed with an extension of the
verb root that differentiates the present stem from other tense-system
stems. Not always, but more often than not, the future stem is built upon
the same form of the verb root as the aorist stem. I would guess that just
as, with expansion of the -SA aorist, futures in -SW/SOMAI came
increasingly to complement aorist actives in -SA/SAMHN; in the same manner
future-tense stems of the -QH- morphoparadigm developed -QHSOMAI/Hi/ETAI
forms to complement the -QHN/QHS/QH aorist forms and that these future
forms also carried the same ambivalence as the -QH- aorists: intransitive,
quasi-passive, middle semantics as called for by the particular verb and
the context. And here too, by and large, the futures in -QHSOMAI tended to
supplant the older middle futures in -SOMAI.

That's a hypothetical account of how I think these morphoparadigms may have
developed and spread--and I believe it helps to explain WHY they spread

>2. A small number of instances have been adduced where, in verbs which have
>(or could have) -QH- forms, -SAMHN (middle) forms are found in the NT with
>passive meaning; and in verbs which have (or could have) -SAMHN (middle)
>forms, some -QH- forms are found which have middle meaning. I wish to
>comment upon these instances further, but let it be noted here that the
>number of verbs where this is commonly or uniformly the case, or instances
>of verbs where it is occasionally the case, is very small. It is not good
>linguistic procedure to base a description of a language upon such a small
>number of instances, or to allow a few contrary examples to annul the
>acknowledgement of a description (a "rule") which applies to most word
>forms (in this case, verbs) most of the time. The standard linguistic
>procedure is to identify the pattern which can be observed to function in
>the language most of the time, and set it out, and then to see what one
>makes of the apparent exceptions. (Is there anyone on b-greek who has had
>formal training in linguistics who would disagree with this statement?)

I think that Ward has, probably unintentionally, misunderstood or misstated
what I actually was arguing. My point was not so much that aorist middle
forms were regularly or commonly used in a passive sense as that aorist
-QH- forms often have a middle sense. Those aorist middle forms that I
think occasionally show a passive sense are very few in the GNT, and they
are essentially older "second" thematic aorist middles (e.g. APWLETO and
one probably ought to consider EGENETO such an exception, particularly
where EGENETO means "was made" or "was done.").

>Carl acknowledges the large number of verb forms for which this
>differentiation holds good:

Not quite so: what I have acknowledged is that the -QH- morphoparadigm in
the Koine Greek verb more often than not signals the semantic passive. I
have NOT acknowledged that "the differentiation holds good." When I pointed
out that only three handfuls or less of NT verbs display BOTH aorist
'middle' AND aorist 'passive' morphoparadigms, what I was actually
intending to show was that the -QH- morphoparadigm has very extensively
SUPPLANTED the older aorist 'middle' morphoparadigm. And I've argued that
the sizable number of so-called "deponents" with 'passive' morphoparadigms
in the aorist, many of which are semantically intransitive and middle,
should NOT be viewed as exceptions to a proper characterization of the
function of the 'passive' morphoparadigm but as evidence for a broader
range of function of that morphoparadigm. Just perhaps, if we want tor
retain traditional terminology with minimal pain, we could retain the
"middle-passive" designation for the MHN/SO/TO morphoparadigm and designate
the -QH- morphoparadigm as "passive-middle" or even

I've already gotten this message up to 15K; just a word about those of
Ward's conclusions that I haven't already touched upon:

>Here then is my conclusions:
>1. The linguistic principle applies: A few instances to the contrary of a
>generally observed language phenomenon do not invalidate the recognition of
>the existence of that language phenomenon, and its utilization as a pattern
>in an explanation of that language, and in teaching students who are
>learning that language.
>APPLICATION: We recognize the correlation between the voice categories
>Active, Middle, and Passive, and the morphoparadigms (to use Carl's
>felicitous term for morphologically differentiated paradigms) of ELUSA,
>2. To just about any rule, exceptions exist. In his "Learning A Foreign
>Language" (1950; p.15), E A Nida says, "There is no reason but a historical
>one for any irregularities, and historical reasons are not really any
>reasons. They are just statements that the irregularity has been in the
>language for a long time. It does not tell us how the complexity arose. We
>do know something about these irregularities, namely, that they occur in
>all languages, that they are very persistent, especially if they occur in
>some frequently used form of the language, and that some irregularities are
>constantly disappearing and others being introduced."
>APPLICATION: We recognize the existence of exceptions to the above
>correlation, that is, that there occasions where a form which
>morphologically is Active, Middle, or Passive, is being used (either on a
>regular basis by a given verb, or in particular instances) in a way which
>in at variance with its morphological indications.

With regard to ##1 & 2, I wonder whether Nida really meant to cover such a
phenomenon as deponency in that statement cited by Ward. I don't disagree
with the statement as such, although I think that saying "historical
reasons are not really any reasons" is a principle of rather questionable
truth or value. Granted, one must learn irregular verbs ("a stubborn lot"
as Lewis Carroll's Humpty Dumpty called them), but for my part I've found
them easier to learn and teach in terms of their linguistic history. As for
the last point under #2 "application," my counter-point is that we are not
describing the morphological indications of the -QH- morphoparadigm
adequately by associating it exclusively with passive semantics.

>3. As cogently stated by Iver Larsen and Bryant Williams (see beginning of
>this posting), recognition of the meaning of a verb form must not be based
>purely upon "grammatical tagging", but must take into consideration syntax
>and semantics: and for voice, this means retaining the three tags, A, M,
>and P. We need to identify that a word is A, M, or P in form: but then we
>must not confuse "form" with "use" when discussing voice.
>APPLICATION: We start with a recognition that there are three
>differentiated morphoparadigms in the aorist and future tense systems, and
>that there are three different types of voice in koine Greek, Active,
>Middle, and Passive; and we recognize the correlation between the two: but
>we also recognize that there will be numbers of occasions when a verb is
>being used in a way at variance with its morphological structure.

This is not so much wrong as it is misleading, and the underlying reason I
find it disturbing, I think, is that THIS synchronic perspective on Koine
Greek is myopic: it fails to take adequately into account that NT Koine is
a language in flux with a surprising number of older and newer forms and
functions at play in speech and writing in the same historical era. I don't
think that the Greek verb can be adequately explained by that structure;
the deponent verbs are not simply irregular verbs but evidence that the
theory of verbs that perceives them as deviations from a norm is somehow
misrepresenting what the norm actually is.

>4. When verb usage is at variance with a verb form's morphological
>structure, sometimes this will be because of the lexical meaning of a
>particular verb, sometimes because of historical factors which persist in
>the language, and sometimes for no discernable reasons whatsoever. On the
>one hand this does not invalidate the primary pattern of verb use discerned
>in a language. On the other hand, the next step is to assess whether, and
>how, these exceptions to the usual pattern can be identified, explained,
>and categorized.
>APPLICATION: Patterns are to be discerned in those instances of verb use
>which depart from the standard "rule" of correlation of voice with
>morphological paradigms (as above). One of those is where morphologically
>the forms in the Greek are middle or passive (or both) and the
>corresponding English meaning is active: what would hitherto have been
>designated "deponent". Perhaps we have to define more precisely what we
>mean by this term to avoid misunderstandings; perhaps we need a new term:
>but this linguistic phenomenon definitely exists in koine Greek. There are
>however other situations where particular verbs exhibit specific semantic
>idiosyncrasies which need to be noted: e.g., where they have no active
>forms in the future, but use middle forms instead (what hitherto we would
>have classified as "verbs with deponent futures"). And verbs which differ
>in other ways (e.g., having no forms in use which have the -QH- morph; and
>isolated occurrences of a verb where it does not have the voice form which
>would have been expected of it in context). As has been suggested, it would
>seem that there is scope for further research in this area.

Yes, there's room for more research and more reflection here and it's a
matter of considerable importance. I remain personally convinced, moreover,
that a more adequate hypothetical perspective on ancient Greek voice is
urgently needed, and that student, scholar, and lay persons can all benefit
from a real reformation of the paradigm.

Carl W. Conrad
Department of Classics, Washington University (Emeritus)
Most months: 1647 Grindstaff Road/Burnsville, NC 28714/(828) 675-4243
cwconrad@artsci.wustl.edu OR cwconrad@ioa.com
WWW: http://www.artsci.wustl.edu/~cwconrad/

B-Greek home page: http://metalab.unc.edu/bgreek
You are currently subscribed to b-greek as: [jwrobie@mindspring.com]
To unsubscribe, forward this message to leave-b-greek-327Q@franklin.oit.unc.edu
To subscribe, send a message to subscribe-b-greek@franklin.oit.unc.edu

This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:37:11 EDT