Part thirty-six of a tour through Greek inflectional morphology to help get students thinking more systematically about the word forms they see (and maybe teach a bit of general linguistics along the way).

We’ve now spent a lot of time looking at distinguishers and inflectional classes within each of the indicative personal ending paradigms (present active, present middle, imperfect active, and imperfect middle) of each lexeme. We also checked the consistency of the lexeme-specific part (the X or “theme”) in each paradigm.

But we haven’t really talked about the consistency of the lexeme-specific part across the PAs, PMs, IAs, and IMs. Perhaps not surprisingly, the same theme is used by a lexeme for both the PA and the PM (if both voices are used) and likewise the IA and IM. In other words, voice is not indicated by the theme in the present and imperfect, only by the set of endings used.

But what about the theme consistency between the present and the imperfect? That’s what we’ll look at now.

Given that the present and imperfect differ in their sets of endings (other than in the 1PL and 2PL) there is not a huge need to use any additional mechanism to express present versus impefect.

But as mentioned when we first started with the imperfects, there is difference besides the endings, namely the augment in the imperfect: typically either a prefixed ε before an initial consonant or a lengthed initial vowel.

The situation is made slightly more complex by the fact that this augmentation applies before the incorporation of any prepositional “preverb”. Greek had a quite productive way of forming new verbs by prefixing base verbs with certain prepositions (a topic worthy of some posts another time).

But for our discussion of augments and the relationship between the present and imperfect forms, we will firstly look at the cases where there is no preverb.

There are 108 lemmas in the SBLGNT without preverbs that start with a consonant in the present and so in the imperfect are just prefixed with ἐ- (e.g. βλεπ- ~ ἐβλεπ-; διδ- ~ ἐδιδ-).

The situation is a little different when the present starts with a vowel. In such cases, the vowel essentially lengthens.

In three cases (ἐάω, ἕλκω, ἔχω) ε becomes ει (e.g. ἐχ- ~ εἰχ-)

In six cases (ἐγγίζω, ἐλπίζω, ἐργάζομαι, ἔρχομαι, ἐρωτάω, ἐσθίω) the ε becomes η (e.g. ἐρχ- ~ ἠρχ-).

We’ll explore the difference in a later post.

Initial ευ stays as ευ in two cases (εὐδοκέω, εὐπορέομαι) but becomes ηυ in one (εὔχομαι ~ ηὐχόμην). Sometimes this happens within a single lexeme too (see the end of this post).

In the five cases of an initial ο- (ὁμιλέω, ὁμολογέω, ὀνειδίζω, ὀρθρίζω, ὀφείλω), it becomes ω. Note that it does not become ου.

οι in οἰκοδομέω becomes ῳ (οἰκοδομ- ~ ᾠκοδομ-).

The one case of an initial η (ἥκω) stays as η.

The vowels α, ι, and υ which can be short or long just become their long variety but of course long α generally becomes η in Attic and Koine without a preceding ι, ε, or ρ.

There are 19 cases of α becoming η (e.g. 2PL ἀγαπᾶτε ~ ἠγαπᾶτε).

There are two cases of ι becoming a long ι (ἰάομαι, ἰσχύω) and one of υ becoming a long υ (ὑμνέω).

In one case (αἰτέω) αι becomes ῃ and in two cases (αὐλίζομαι, αὐξάνω) αυ becomes ηυ.

We now turn to those verbs with a preverb (or which augment as if they do).

Because a lot of prepositions end in vowels or other sounds that interact with the start of the verb root or with the augment there is often elision or assimilation.

For example ἀναβαίν- (ανα + βαιν-) in the present becomes ἀνεβαιν- (αν’ + ε + βαιν-) in the imperfect. The ἀνα- is intact in the present but elided in the imperfect. ἀνεχ- in the present becomes ἀνειχ- in the imperfect: ἀνα- elided to ἀν’- in both present and imperfect.

The ε augment often has the effect of breaking the consecutive consonants that assimilate in the present. For example ἐμβλεπ- in the present becomes ἐνεβλεπ- in the imperfect, the ν in ἐν- no longer becoming μ in the presence of the following labial β.

συν- is particularly worth observing because you get things like:

  • συγχαιρ- ~ συνεχαιρ-
  • συζητ- ~ συνεζητ-
  • συλλαλ- ~ συνελαλ-
  • συμβαιν- ~ συνεβαιν-

At some point it might be fun to whip up the exact finite-state model for all this but for now, I’ll just note the counts.

There are 100 examples with preverbs plus a consonant-initial verb stem.

Preverbs plus a vowel-initial verb stem follow the same sound rules as without a preverb and the expected elision can be found.

There are 16 cases of α>η with a preverb (e.g. περιαγ- ~ περιηγ-; ἀπαγγελλ- ~ ἀπηγγελλ-). There are two cases of αι>ῃ with a preverb, 8 cases of ε>ει, 11 cases of ε>η, one case of ευ>ευ, 4 cases of η>η, 6 cases of ι>ι, and 3 cases of ο>ω.

Here’s a summary of all these counts:

  no preverb with preverb
C > εC 108 100
ε > ει 3 8
ε > η 6 11
ευ > ευ 2 1
ευ > ηυ 1 -
ο > ω 5 3
οι > ῳ 1 -
η > η 1 4
α > η 19 16
ι > ι 2 6
υ > υ 1 -
αι > ῃ 1 2
αυ > ηυ 2 -

There is the interesting case of εὐαγγελίζω which is treated in Acts as if εὐ were a preverb and the imperfect form εὐηγγελίζοντο is found (wth α>η). This is not counted in the 16 above.

We also in Acts find προορώμην which does not have an augment but which clearly has the imperfect middle ending.

Not included in the counts above are εἰμί and compounds which are probably worth their own post at some point (although note that for the most part the imperfect stem is just η). Also not included are εἶμι and compounds where the imperfect stem is ῃ.

We have a handful of cases where the lemma having multiple inflectional classes prevents a trivial mapping between the present and imperfect stems in all instances (ἀφίημι, δέω, συγχέω, ἵστημι, and πλέω, each of which we’ve discussed before.) Once the right stem is chosen to map to, the rules above apply cleanly.

That leaves us with five imperfect stems whose relationship to the present stem has not yet been covered. They are:

  • δύναμαι having imperfects in both ἐδυν- and ἠδυν-
  • μέλλω having imperfects in both ἐμελλ- and ἠμελλ-
  • θέλω having an imperfect in ἠθελ- (because the present was originally ἐθελ-)
  • εὐκαιρέω having an imperfect in εὐκαιρ- and ηὐκαιρ-
  • εὑρίσκω having an imperfect in εὑρισκ- and ηὑρισκ=

Other than these and the ε>ει versus ε>η distinction, augmentation of the present stem to form the imperfect is entirely consistent and predictable.

We’ll dive into more detail on the augment later on but we’ve now reached a good time to leave behind the present/imperfect and start to look at the aorist in the new year!