explanation of aorist morphology (was, kappa aorists)

From: Ward Powers (bwpowers@eagles.bbs.net.au)
Date: Sat Oct 25 1997 - 08:24:44 EDT

Fellow b-greekers:

(Or at least: those fellow b-greekers with an interest in this thrilling
saga of the Explanation of Aorist Morphology - the rest of the list deleted
this post when they saw the subject heading.)

In our last exciting instalment, we saw how linguistic analysis of the
aorist forms in the GNT containing -SA- and -A- reveals that these are
allomorphs of the punctiliar morpheme (which also contains the variant I.A
allomorph when a liquid verb root has epsilon as its penultimate phonene;
plus three "kappa aorists" which have the allomorph -KA- as their
punctiliar morph). (Are you still with me?) But when a verb's punctiliar
morph (whichever of these it is) is followed by a vowel, it no longer
contains an alpha. What can possibly be the cause of this strange behaviour
on the part of the punctiliar morph? Is some other vowel moving into the
morph and taking the place of the alpha? Or has the alpha been indeed
banished from the morph, but not replaced? How do we explain what is going
on? (Indeed, does it really matter?)

Now read on ...

First of all, let us identify the things we need to explain, which are: the
forms without the alpha of -SA- in the verbs which usually take -SA- in
their aorist forms. These forms are: the 3rd p. sing. indicative ELUSE(N),
the 2nd p. sing. imperative LUSON, all the subjunctive forms (active and
middle), and a couple of others that also need a comment.

The explanation usually offered for the 3rd p. sing. indicative active form
ELUSE(N) is that in the indicative active -A- and -E- are alternatives.
Some grammar writers join this vowel to the -S- as together comprising one
"marker" (what linguistically would be called a morph), while others regard
the -S- and this vowel as being separate morphs. Whichever of these two
views is accepted, the point is that they both hold that the -S- plus -E-
is an alternative of the -SA- in the other forms of the verb's conjugation.
When subjected to examination from a linguistics perspective, this analysis
of the forms simply will not hold up.

First of all, as an explanation it is limited to just one form [EULUSE(N)]
out of the group which have the same factor needing to be accounted for
(that is, a vowel other than alpha following after the sigma). When we note
that the aorist 2nd p. sing imperative is LUSON, are we now going to say
that in THIS form the -SA- changes into -SO-? Why should the -SA- change
into -SE- on one occasion, and -SO- on another, and only ever make each of
these changes the once? From a linguistic perspective, such an explanation
involving a feature occurring only once in the language would be regarded
as highly suspect.

Furthermore, this explanation "-SA- changes to -SE- in the aorist
indicative 3rd p. sing." gives us no help at all in relation to explaining
the forms of the sorist subjunctive. We cannot there say that the vowel
following the sigma (which is either an omega or an eta, sometimes with a
iota subscript) is an alternative in these forms to the alpha, i.e. that
the alpha now changes into omega or eta or eta plus iota subscript: the
aorist subjunctive (active and middle) is IDENTICAL WITH the present
subjunctive, with just the addition of sigma following the verb root. To
understand this requires us to go back to look at the present flexion.

In the present active and middle flexions, each verb form (for example,
LUESQE) consists of three elements ("morphs"): the "meaning bit" (LU-), a
joining vowel which is often called "the thematic vowel" or "the connecting
vowel" (-E-), and the pronoun ending, which also indicates voice (-SQE). As
an aside at this point, I will confess that I have never been able to
understand why some grammarians call this vowel "the thematic vowel". What
is the intended meaning of this term "thematic"? How does this vowel show a
"theme" in a verb? It is just an empty word, a word without significance.
(I am wide open here to being corrected by anyone giving me a coherent and
logical explanation of the term.) Certainly the term has no meaning or
significance from a linguistic viewpoint. I will call it the "neutral
morph", for reasons which will shortly become apparent.

This neutral morph has two allomorphs, which occur in a regular and
predictable context: it is omicron when form final or followed by a nasal,
a iota, or an upsilon, and epsilon otherwise. (Actually, the upsilon in
LUOUSI(N) is compensatory lengthening of the omicron for the loss of
original -N- or -NT- from the stem, so in this form also -O- was originally
there because followed by a nasal. Being followed by a iota only occurs in
the optative and is a special situation.)

The subjunctive morph in Greek is a "process morph" - a process of change
to something already in the word. (We have lots of these in English also:
e.g., the way we make the plural of "man".) The present subjunctive (both
active and middle flexions) is formed by the lengthening of the neutral
morph, -0- into -W- and -E- into -H-. A following iota goes subscript. The
difference between these flexions and those for the aorist subjunctive is
only that in the latter there is a sigma inserted between the "meaning bit"
of the word and the lengthened neutral morph.

CONCLUSION: in the aorist subjunctive there is no way the vowel after the
sigma can be taken as a vowel replacing the alpha in the otherwise usual
-SA-; the aorist (punctiliar) morph is just -S-. The vowel which follows a
sigma is to be recognized as being, or being part of, A SEPARATE MORPH
conveying some other piece of information about the verb form. Thus the
lengthened neutral morph after the -S- is saying, "subjunctive form".
Similarly, the -ON in the form LUSON is an ending saying, "2nd p. sing.
imperative". And also, the -E- in the indicative 3rd p. sing. (whether with
or without the movable nu) is the pronoun ending meaning "he/she/it". (We
will see other parallels for this, below.)

Where does the foregoing leave us? The usual explanation of the indicative
3rd p. sing. form ELUSE(N), an explanation that links the -E- with the -S-
in some way as indicating "aorist", is just wrong. There is no cohesive
explanation offered along these lines that can cover all the observable
data in the Greek, that is, other forms which have a vowel other than -A-
after the -S- in the aorist. This is a dead end. What then is the way forward?

We go back to the observed data. Here are the facts:

1. The vast majority of Greek verbs contain -SA- in the aorist active and
middle between the verb root ("the meaning bit") and the pronoun ending.
When this -SA- occurs it is always either form final or followed by a
consonant. (A word to those who noticed it: The apparent exception of LUSAI
forms will be covered in due course. Patience.)

2. But there are some occasions when IN THESE SAME VERBS there are forms
which have -S- followed by another vowel, not -A-.

3. When that happens, the vowel after -S- is not part of an aorist morph
together with the -S-, but is (or is part of) the following morph in the word.

4. Therefore as a statement of fact covering the observable data, we can
say: The aorist (punctiliar) morph in a regular first aorist verb form is
-SA- when it is form final (that is, not followed by any other morph) or
when it is followed by a consonant, and -S- when followed by a vowel.

Now, is there any other situation in Greek which is comparable with this?
Yes, indeed there is. When a preposition is prefixed to a verb form, and
that verb form commences with a vowel, the final vowel of the preposition
elides. (PRO and PERI are exceptions to this rule; the other nine
prepositions ending in a vowel follow it.) We therefore observe the
existence of this concept of elision in the language, and how it operates.
What we observe in the behaviour of the punctiliar morph is absolutely
identical to the observed behaviour of the preposition morph. Therefore
this feature of the Greek verb can also be described as elision, because
everything which elision does in Greek operates in the same way in relation
to the punctiliar morph. So we can add a little more to what we have just

5. That is to say, the -A- of -SA- elides before a following vowel.


But that is not quite the end of the story. There is a little bit of
tidying up to do still. As at the end of any good detective tale, the loose
ends must be tied together neatly. Note the active infinitive LUSAI. This
contains -SA-, but the -A- cannot be part of the punctiliar morph, or under
Proposition 5, above, it would have elided in front of the iota. Therefore,
in accordance with this Proposition, the punctiliar morph must be just -S-,
and the -AI must be the following morph (which in this form would be the
infinitive morph), in front of which the -A- of -SA- has elided, in
accordance with the rule. We check this out (that is, we seek to see a
similar pattern elsewhere) by looking at other infinitives. We find lots of
them ending in -AI-: LUESQAI, LELUSQAI, LUQHNAI, and in other verbs
hISTANAI, EINAI, DOUNAI, GNWNAI, and so on and on. Thus we are confirmed in
recognizing -AI- as indicating "infinitive". (There are other infinitive
endings also, and it is possible to state a rule to account for them: but
that is not part of our present purpose.)

The -AI- in the optative has arisen from the regularizing in the koine
Greek verb of earlier less morphologically regular forms in its optative
flexions, with the iota suffixed to the punctiliar morph then coming to be
the "sign" (morph) of the optative. Note the same thing in the present
optative, where the iota is similarly suffixed as the optative morph to the
neutral morph -O-.

The -AI- in the aorist middle imperative LUSAI and the -W- in aorist middle
first p. sing. ELUSW have resulted from contraction.

We have also noted at the very beginning that the -S- of -SA- slides off a
liquid in the morph preceding it, leaving just -A- as the aorist morph in
those verbs, and that three verbs have -KA- as their aorist morph instead
of -SA-.


The punctiliar morph -SA- occupies Slot 7 of the Greek verb's nine morph
slots. An examination of parallel forms in other tenses of the active voice
will show that in that same slot in those forms will be found either the
neutral morph or -KA-, the perfect active morph. See for example ELUETE,
ELUSATE, LELUKATE. I mention here (I will not attempt now to demonstrate
it) that the verb roots of all verbs are inherently either durative or
punctiliar aspect. First Conjugation verbs (those with a first aorist) have
inherently durative roots, and form their verb systems from the durative
(present). That means they do not need anything in the verb form to MAKE
them durative. When they start life, as it were, they already ARE durative.
The neutral morph (-O/E-) in the aspect slot has a neutral affect upon the
verbs aspect. (Hence my preferred term for it, "neutral morph".) That is,
it leaves the verb's aspect unaltered. When this neutral morph is replaced
with the punctiliar morph, -SA-, the aspect of the form is switched to
"punctiliar". When Slot 7 is occupied by the morph -KA-, then the form is
switched to "perfective". (Note that this -KA- only occurs in the active
voice forms.)

It should be noted that the neutral morph and the perfective active morph
are also elision morphs. That is, like the punctiliar morph, they elide in
front of a following morph beginning with a vowel. In the case of the
neutral morph, this means it vanishes from view entirely [as in the
imperfect active 3rd p. sing. form, ELUE(N),], and in the case of the
perfective active morph the -KA- elides to -K- [as in the perfect active
3rd p. sing. form LELUKE(N)]. In both these forms the 3rd p. sing. pronoun
morph (with the meaning "he/she/it") is thus -E(N): exactly the same as we
have seen in the 3rd p. sing. aorist active form ELUSE(N)! This elision of
the aspect morph vowel also explains the pluperfect forms (E)LELUKEIN etc.
(the -EI- is the pluperfect morph, in Slot 8), and the participle LELUKOTOS
etc. (the -OT- is the perfect participle morph, in Slot 8). And we could
also note that Second and Third Conjugation verb roots are inherently
punctiliar. But all these things are part of another story, and go beyond
our present brief.

Why bother with all this detailed analysis, for (what may appear to some)
to be such minor results? Some people may decline all of the foregoing as
tedius, tedium, tedi-issimo (with "decline" perhaps becoming the operative

Is it tedious? Well, no more so than (say) the work of comparing variant
readings that is at the centre of textual criticism, which has the
important aim of arriving at a more reliable text. For myself, I find
linguistic analysis fascinating, and I suggest that this exercise is
worthwhile, for three main reasons:

1. Because of gaining knowledge for its own sake. Just to get a right
understanding of the matter. I reckon these things (as set out above) are
so. I share that conclusion with you, and present the evidence. You see
what you think.
2. Because the essence of what linguistic analysis is seeking to do is to
establish the simplest way of understanding and explaining the functioning
of a particular language. It is very much an "Occam's Razor" operation.
("Other things being equal, the simplest explanation of the observable data
will be the correct one.")
3. Because this simple explanation of first aorist morphology makes it so
much easier for students to understand a first aorist form, and to
recognize it in its variant disguises. Your base rule is that the
punctiliar morph is -SA-. You have an extension to cover the situation when
this morph is preceded by a liquid. You have a second extension to cover
the situation where this morph is followed by a vowel. At some stage you
can mention the five other verbs, the three kappa aorists, and ECEA and
HNEGKA. And that's it.

This rule has four things going for it:

1. It is very easy to understand.
2. It is very easy to use (to identify a verb form).
3. It is very comprehensive. (Only five special cases, and they can so
easily be slotted in. Are there other exceptions? I know not any.)
4. It fits neatly and exactly into the overall pattern of how parallel
morphs behave in the Greek language system.


Rev Dr B. Ward Powers Phone (International): 61-2-9799-7501
10 Grosvenor Crescent Phone (Australia): (02) 9799-7501
SUMMER HILL NSW 2130 email: bwpowers@eagles.bbs.net.au

This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:38:34 EDT