Thaptw: aorist passive etafhn (epsilon, tau, alpha, phi, eta, nu)
What's the explanation for this? The aspiration seems to have changed place.
Just guessing, I would say that the tau has dissimilated because of the phi. But why the phi? Perhaps it was expecting (or making the analogy with?) another theta aorist passive (luw etc), hence the aspiration?
Sorry if that sounds confused. I hope you get the general idea.
The explanation is that originally both t and p were aspirated. With the formation of the present stem, the p lost its aspiration, so that there was no reason for the t to dissimilate. In the aorist stem, the aspiration of p was retained; the t then lost its aspiration.
I don't have references to any grammars (Smyth, Kühner, Sihler, Rix &c.) at the moment, but if you need them, I could look it up.
This process Ptolemaios is describing is known as Grassman's Law, and applies to Greek and Sanskrit. When two aspirates follow one another, the first is unaspirated. However, when the -t- was added to the present stem, the final phi lost aspiration, so the theta remained at the beginning.
The same process accounts for the paradigm of "hair": [face=spionic]qri/c, trixo/j, h([/face].
original root [face=SPIonic]qaf[/face] + [face=SPIonic]t[/face] > [face=SPIonic]qapt[/face]. The initial aspirate is allowed to stand because the subsequent aspirate disappeared into the [face=SPIonic]pt[/face].
But Grassman's law operates in the 2d aorist passive giving [face=SPIonic]e)ta/fhn[/face] < [face=SPIonic]e)qa/fhn[/face].
There are cases where Grassman doesn't operate, e.g., 1st aorist passive, m-p perfect imperative and infinitive, respectively: