Re: Augment revisited (was: NUN+Verb.Aorist)

From: Don Wilkins (
Date: Thu May 01 1997 - 18:58:27 EDT

Apparently Carl and I both are trying desperately to find extra hours in
the day (I was up to 3:30 AM this morning working on program bugs, as I
often do), and for that reason it is all the more commendable that Carl
would offer such a thorough response to this issue. I will try to reply
with as much relevant detail as I can muster at the moment, keeping
documentation brief for now.

At 5:01 AM 5/1/97, Carl W. Conrad wrote:
>At 6:58 PM -0500 4/30/97, Don Wilkins wrote:
>>Jonathan, if you are referring to the absence of the augment in
>>non-indicative moods, then of course we would say that such aorists
>>inherently do not have reference to the past. As always, the problem is
>>with the indicative. Do you have in mind some aor. ind's that do not have
>>the augment? Offhand I can't think of any except those which are from the
>>Homeric time period or are "missing" the augment due to purely
>>morphological reasons (e.g. those verbs which begin with a long vowel).
>I would qualify Don's statement here somewhat to note that: (1) absence of
>augment in the non-indicative moods has never been a matter of dispute;

I'm sure we are in complete agreement here, so Carl's qualification
probably applies to what follows.

>(2) while it is true that the augment seems to be present or absent on
>imperfect, aorist, and pluperfect indicatives in Homer without any clear
>indications of a reason for inclusion or omission, some reasons other than
>metrical necessity or convenience have been adduced by some scholars. I
>would agree with Don, however, that Homeric practice ought not by itself to
>have any bearing on our judgment of the function of the augment in Koine or
>specifically NT Greek; BUT (3) whether the flexibility of Homeric usage
>with regard to augment on these indicative tenses is the major factor or
>not, it is nevertheless true that Greek poetry in all eras of antiquity
>remained flexible in the use or non-use of augments;

Quite right, and I take it that Carl (in point 3) is implying some
relevance to the present discussion. While one can make such an argument,
here is a comment from p. 502 of Threatte's recent _The Grammar of Attic
Inscriptions II: Morphology_, to which I've drawn attention in a previous

"Both the syllabic and temporal augment are frequently omitted in metrical
inscriptions of all periods whenever the omission facilitates the metre.
But the augment is virtually never omitted in Attic inscriptions in prose.
It is certainly omitted in EPERWTHSEN [my transliteration here and
following] in the Iobacchi regulations, II2 1368.20 (ca. 162/3 p.), and
while spellings with omitted augment certainly occur in papyri (cf. Mayser
1.2, p. 102; Gignac 2, p. 234, no. d1, with three cases of ERWTHS-,
including one dated 100 A.D.), the Attic example is so isolated that it is
tempting to consider it a scribal error caused by the preceding EPE |
ROTHSOMEN of ib. 19-20 (both readings L.)."

Threatte goes on to discuss a few other exceptions, none of which are
difficult to explain. The point for our discussion, as I see it, is that
poetry can reasonably be said to be flexible with augments due to the
nature and metrical requirements of the genre itself, and it does not
provide us with significant evidence for argument.

>(4) it ought not to be
>forgotten (it has been noted more than once, most recently by Edgar Krentz,
>I think, that the augment is sometimes absent from pluperfect indicatives
>in Koine. While this last fact may seem to be unimportant, since we have
>been talking mostly about imperfects and aorists, I think it's a crack in
>the wall of adamant assertion that the augment is necessarily an indicator
>of past time on an indicative verb form. I'm inclined to suspect (but of
>course cannot prove) that omission of the augment on some pluperfect forms
>MAY be occasioned by the length of these verb-forms in terms of syllables;
>in modern Greek an imperfect or aorist indicative is augmented only if the
>verb-form is two syllables or less long.

I think Carl's suspicion is well-founded. As to the pluperfect, Threatte
(in the place cited) states:

"The syllabic augment is not omitted in such few pluperfect indicative
forms as occur in the Classical and Hellenistic Periods, .... The reading
GE | GONEI TA has been recorded in the letter of a Roman official,
_Hesperia_ 10 (1941) p. 78, no. 34, lines 8-9 (s. II p.), but the very worn
condition of the stone makes this interpretation doubtful (L., GE |

So, while there is not a lot to go on for the pluperfect, the evidence
which exists indicates that there was no tendency to drop the augment
before (or perhaps until) the NT period. In view of this conclusion, I did
an Accordance search for the plupft in the NT and LXX. The LXX shows the
dropping of the syllabic augment in PAREMBEBLHKEISAN of Judg. 7:12; in
KEKRATHKEI of 4 Macc. 6:32; and in BEBHKEI of Wis. 18:16. Most of the
172 occurrences of the plupft in the LXX have temporal augments, but there
are also a considerably larger number of plupfts with the syllabic augment
than without (only 3, unless I overlooked something).
In the NT, Matt has TEQEMELIWTO lacking the augment in 7:25, and no other
instances of plupft that could take the syllabic augment. I should point
out for those who may be unfamiliar with the problem that verbs which begin
with a vowel seem to be fairly consistent throughout their history in the
handling of the plupft (such as OIDA going to Hid-, hISTHMI to hEIST-,
etc.), and so it is the syllabic augment verbs that are the focus of our
attention (Carl can correct me if I am oversimplifying). Mark has three
plupfts without syllabic augment (excluding the one in 16:9) and seems to
be more inclined to omit the augment in this tense than the other writers.
One word, DEDWKEI is used once by Mark (14:44) and Luke (19:15), while John
uses DEDWKEISAN in 11:57. This might indicate that there was a tendency to
lose the augment in the plupft of DIDWMI, rather than a tendency to lose
the augment across-the-board, so to speak. Excluding the disputed passage
16:9, that leaves two instances of the plupft in Mark where the syllabic
augment is omitted (15:7 and 10), of which the latter is a compound form
What I draw from this is that the absence of the augment for the plupft in
the NT is not very substantial given the existence of 82 plupfts, and the
instances of the same phenomenon in the LXX are roughly parallel to the NT,
despite the much earlier age of the LXX. SOMETHING seems to be going on,
but does the evidence really provide much ground for concern about a major
shift in the use and understanding of the augment during the NT period?
Another explanation that seems possible to me is that the augment seemed
very inconvenient in the construction of the plupft (along the lines of
Carl's observation about its multi-syllabic nature) and might easily be
omitted on some occasions because the plupft, like the impft, is uniquely
indicative and limited to past time. The fact that the same thing does not
occur for the impft may be due to the latter's frequency of usage and
simpler construction.

>I would not dispute the overwhelming documentary evidence for the presence
>of the augment on indicative imperfects and aorists in the written Greek of
>the NT period. I wish we had some way of knowing how consistent its
>presence may have been in the spoken language, but I don't think we have
>enough evidence about that to say anything one way or the other. I suspect
>that the wave of Attic revival toward the end of the first century may have
>retarded any changes in use of the augment that may have been developing in
>the demotic language.

Well, we *might* get some help from colloquial papyri of the period, if
some is extant. I have not yet taken a close look at the papyri, and with
the recent update of the PHI disc it may be possible to find something

>I think that what I would want to emphasize in this matter is (a) the flux
>in morphology and syntax that is always taking place in any language being
>used in an broad international and cultural realm, a flux that is evident
>even within the NT in alternative forms of the same words in the same
>authors (to mention only one indication of this), and (b) the
>questionability of the assumption that the enduring general use of the
>augment in the indicative of the imperfect and the aorist necessarily means
>that the augment itself was understood by its users to be a clear marker of
>past time. Quite frankly I have a strong suspicion that it is primarily
>secondary endings and especially the alpha endings that may be more
>important to the perception of the indicative imperfects and aorists as
>past tenses.
>The upshot of this is, I guess, that I don't question the overwhelming
>evidence that the augment continued to be used on indicative aorists and
>imperfects. What I do question is the claim that the augment was
>undisputably recognized as a marker of past time in these forms by those
>using it. I'm just not so confident that the endurance of a phonetic
>element in these verb-forms necessarily means that speakers and writers
>felt it had a semantic force.

This of course is where I would tend to part company with the generalist
linguistic approach. I think we need to look very carefully at the specific
pieces of evidence within the literature of a language itself and consider
that what may appear to be similar across languages is not necessarily so.
It's undoubtedly true, as Mari has noted, that grammatical forms like the
augment will last for some time after their meanings have become obscure,
but the actual timing of such an event should be determined objectively
from the language itself. The appeal of the generalist approach is (as it
seems to me) that one can really know all or most of what there is to know
about *any* language's syntax etc. by learning the supposed common traits
of language in general. But this seems too good to be true, and most likely
is. I hasten to add that I don't know what Carl thinks about this and am
going well beyond his comments immediately above. In that regard I would
agree with him that we can't know in effect what went on in the minds of
the speakers/writers; I just think that the written evidence that we have
paints a relatively clear picture of the use of the augment, at least
through NT and somewhat later usage.

Don Wilkins
UC Riverside

This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:38:14 EDT