semivocalic iota

[Lukas]

Χαίρετε!

Dr. Mastronarde's Introduction to Attic Greek mentions a semivocalic iota here and there. For example, on page 387:
"α. When y is added to dental stems, the result is a stem in ζ, such as
κομίζω from κομιδ- and ἐλπίζω from ἐλπιδ-. Note, however, that
-ίζω and -άζω became productive suffixes in their own right for
denominative verbs, as in νομίζω from νόμος or δικάζω from δίκη.
β. A ζ results similarly for a few other plosive stems, such as ἁρπάζω
from a root that appears in other forms as ἁρπαγ-.
γ. When y is added to velar plosive stems, the combination produces
Attic ττ and non-Attic σσ: e.g., φυλάττω (φυλακ-), πράττω (πραγ-),
τάττω (ταγ-).
δ. When y is added to stems in a liquid or nasal, the combination
produces either a doubling of λ or a compensatory lengthening and
alteration of the vowel preceding ν or ρ: e.g., ἀγγέλλω (ἀγγελ-),
βάλλω (βαλ-), φαίνω (φαν-), κτείνω (κτεν-), φθείρω (φθερ-),
κρῑν- (κρῐν-)."

Was the semivocalic iota an actual letter that was used, or was it more an added sound that certain consonants made when pronounced?

[mwh]

No it’s not and never was an actual letter. Linguists infer its erstwhile presence as a sound (not a letter) from morphological behavior. It explains why κομιδ- becomes κομιζ- for instance. But unless you’re interested in prehistoric phonological phenomena within the language you can forget about it. κομιζω is and always was written κομιζω (not κομιδyω or κομιδιω), and that’s how it’s pronounced.

[Lukas]

Ευχαριστῶ!