Part thirty-nine 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).
Now we’ll take an initial look at the aorist active infinitive and indicatives for λύω:
Probably the most common term for this type is first aorist but this implies some ordering (versus the “second aorists” in particular) that isn’t particular helpful in most cases.
If we’re contrasting the indicatives here with their imperfect equivalents, we’d have
The existence of the sigma is why these are often alternatively called sigmatic aorists.
But if we just look at the distinguishers within the paradigm, we can drop the sigma and get the following (with the thematic and root aorist distinguisher patterns shown for comparison):
Observe that the 2SG, 1PL, 2PL, and 3PL all appear to have the same ending as the thematic aorists but with an alpha instead of the ε/ο theme vowel. This is why these aorists are often alternatively called alphathematic aorists.
In the 1SG, the alpha ending is actually related to the ν in the thematic and root aorists. If a final ν (coming from a Proto-Indo-European *-m) follows a consonant, it becomes an α in Greek (coming from a syllablic *-m̥ in Proto-Indo-European). This is just a way of making an otherwise unpronounceable sequence pronounceable (in Greek). We see this same phenomenon in the accusative singular nouns (-ν when preceded by a vowel like in the 1st and 2nd declension, -α when preceded by a consonant in the 3rd declension).
So ἔλυσα makes sense instead of ˣἔλυσν. But now we have an interesting question: is the -α the ending or part of the aorist stem? Its origins are clearly as the regular ending but in light of forms like ἔλυσας, ἐλύσαμεν, one might reanalyse it as part of what distinguishes this type of aorist.
In the 3SG we find the bare ε (with movable nu in many cases) presumably by analogy with the thematic aorists. If an α had been used, it would be easily confused (without a nu) for the 1SG and (with the nu) for the 3PL
The 3PL is particularly interesting because it potentially explains the -σαν ending we see in the root aorists (and even athematic imperfects). Again this comes back to an interesting reanalysis.
ἔλυσαν, if thought about in terms of the historical 3PL ending, could be segmented ἔλυσα-ν. If thought about in the context of the other personal endings in its paradigm, it could be segmented ἔλυσ-αν. But it could also be segmented ἔλυ-σαν, particularly in comparison to the imperfect. The morph -σαν could then have been internalised as indicating 3PL in the aorist or even aorist/imperfect.
There’s more we can say about this once we’ve covered the perfect endings (probably a little while off!) as they’re likely involved in this as well but this is yet another example of how morphology isn’t really about concatenating morphemes with some compositional meaning resulting. It’s a complex interaction of forms within a system.