Tagging

Carl W. Conrad cwconrad at artsci.wustl.edu
Sat Dec 14 09:25:31 EST 2002


Forwarded for "Iver Larsen" <iver_larsen at wycliffe.org> to replace
now-deleted HTML-formatted version of this message"
Date: Sat, 14 Dec 2002 03:12:43 -0500

Hi, Dale,

The following was written before I read Carl's response to the same post,
but I'll let it stand as it was written, since it complements what he said:

It must be an enormous work to morphologically tag all verb forms in the
GNT and LXX. But it is very helpful, especially in this computer age.
Let me make a couple of comments and suggestions below:

>[Dale:] I will look at this problem with DUNAMAI again; I've looked at it
>in the
> past and came to the conclusion that, since both middle (fut) and passive
> (aor) forms occur, one of the two has to be chosen as the tag for the
> "deponent" forms in the Pres/Impf/Pf.  The original editor of GRAMCORD
> chose passive, and at the time (coming up on ten years ago) I
> looked at it,
> I couldn't see any compelling reason to change it.  However, you, and
> fixing the MorphLXX along with Bernard Taylor, have convinced me that not
> only is "deponent" an incorrect term to apply to the
> form/function of Greek verbs [except for possibly verbs which truly only
> can bear one  voice, eg., ERXOMAI, viz., you can't be "goed" in English
> or in Greek, or  German...but  you can be in Hebrew, HLK: in the Hifil),
> but that middle is the overriding concept for these ambiguous verbs.

I would agree that the term "deponent" ought to be banished completely from
the teaching and tagging of Greek.

> The 9 Aor Pass forms of DUNAMAI are
> good examples of middle meaning in a passive form (I can find no other
> reason for these things at this point than that they just sounded
> better to a Greek ear than the middle form?).

It is common in IE languages that some of the most used verbs are irregular
in form. There is no synchronic explanation for this, only historical
development, or maybe just chance or accident. Common irregular verbs are
"go", "come", "be", "have", "can", "will".

 So I suspect that I'll change the
> ambiguous forms to middle...but that leaves the Aor Pass forms
> still tagged
> as Passive (the Morph Code: "tag only what is written, not what
> is meant!";
> which of course doesn't work all the time, but its our basic
> starting point).

I suggest you follow your basic starting point more closely. Some languages
have a simple affix indicating passive forms consistently. Greek does not.

When it comes to paradigms, I don't think it is helpful to talk about
ambiguous forms. The verb forms ought to be characterized according to
their place in the morphological paradigms, not according to any skewing
between form and meaning.
There is another kind of idiosyncratic ambiguity that should be maintained,
but that has to do with isolated cases, not whole paradigms. I am thinking
of the ambiguity between present ind. and imperative in 2 person plural,
and similar ambiguities.
>
> Also, you and Iver are using the terms MP1 and MP2...I must have
> missed an email...what are you referring to?

These were suggested by Carl and are based on a primary dichotomy between
active/basic forms on one hand (code: A) and middle-passive forms on the
other hand (code: MP). This dichotomy is basically syntactic and semantic,
rather than morphological. When needed, the MP forms can be further divided
into two groups based solely on the morphology. The MP forms that have the
-MAI etc. endings without the -QH- suffix are called MP1, and those with
the -QH- suffix are called MP2.

I used them because Carl introduced them, but I would not recommend them
for morphological parsing, partly because a one-letter code is more
practical, partly because I think the morphological codes should be based
on morphology rather than syntax and semantics, although I accept that
these grammatical categories are interdependent.

Let me suggest a system for morph tags that I would consider most helpful,
but first let us have a brief look at the morphology of Greek voice.

The basic morphological difference between what is traditionally called
active and middle, is not a difference in verb stems, but in the use of two
different sets of verb endings (pronominal suffixes). One distinctive
feature is that 1 ps. pl. in set I is -MEN, and 1 ps. pl. in set II is
-MEQA.
Set I is the basic set for most verbs and set II is normally used to convey
a middle or passive sense. But there are a many exceptions. For instance,
PASCW uses set I, but has the MP sense normally associated with set II.
Some verb forms use set II, but are active in sense.

These two paradigms of verb endings ought to marked in the morph tags, and
I would suggest to keep A for set I and use M for set II - even in those
cases where the sense is clearly passive, because whether they are passive,
middle or active in sense is a decision which is not or should not be made
on the basis of verb morphology alone. Since A is the first letter of the
alphabet, it can be used to stand for the basic form, as well as
maintaining a link to the former "active". The M stands for the -M- in -MAI
rather than Middle, but there is still a conceptual link to Middle.

Then we have verb stems with an added -QH- suffix in some aorist and future
forms. (The Q may be elided in certain environments and the H shortened to
E, but such forms are still to be counted as belonging to the -QH- stem).
The code for this stem could be either Q or H, but it should not be P. (If
two letters are allowed in the code, they could be termed -TH-, or when/if
everybody is using unicode, maybe use the Greek letter for theta. For now,
I am suggesting the b-greek convention for writing theta.)
The -QH- stem paradigm is a peculiar hybrid in that it takes set I
pronominal suffixes in the aorist and set II pronominal suffixes in the
future. (Future is marked by the addition of the -s- suffix to form a
future stem). The -QH- stem is often, but far from always, associated
with a passive sense. Sometimes it is associated with a middle sense, and
in a few cases even with an active sense. This needs to be determined for
each verb in question.

Using these three morph tags, Q for verbs with the -QH- stem, A for set I
pronominal endings apart for those verbs with a -QH- stem, M for set II
pronominal endings apart from the -QH- stems, should be fairly easy and
unambiguous.

In the case of DUNAMAI, the 9 aorist forms with the would be marked Q, and
all other forms marked M. A study of the verb in question and the contexts
in which it is used, will determine whether the meaning is active, middle
or passive.

The students can be taught that A forms tend to be active, M forms tend to
be middle, and Q forms tend to be passive, but this is only a rough
guideline. Or if they have been taught basic semantics, which I do hope
everyone will be taught, because it is much more helpful than grammar
alone, then they can be taught that
1) the A forms tend to represent a semantic verb with all its arguments
(roles) potentially present and usually explicit, and situations where the
agent is different from the patient, experiencer or beneficiary (PASCW and
a few other exceptions that defy synchronic explanation)
2) the M forms tend to represent a semantic verb where the agent role
coincides with the patient, experiencer or beneficiary role (again with a
few exceptions that defy synchronic explanation)
3) the Q forms tend to represent a semantic verb where the patient (or
beneficiary) is expressed as the syntactic subject. The agent is different
from the patient or beneficiary and is normally not made explicit. If it is
made explicit, it is by way of a secondary and added agent role - expressed
in the surface grammar by a prepositional phrase. (A few exceptions again).

Iver Larsen




More information about the B-Greek mailing list