Of initial consonant clusters in Indo-European

[alice]

I don't know if this happens in practice, but I'm interested in the theoretical outcome, if it's known.

Imagine we have a root which starts with a laryngeal and ends with a stop; let's call it Heg. Now put this root into zero-grade and add a suffix which starts with a consonant: Hg-to-s. Initial laryngeals are lost in most IE languages (all except Greek and Armenian, IIRC), so we now have gtos. What happens to that initial gt?

[Soap]

The only root I know of offhand that is similar is *es- "be". In Spanish the past participle is sido, which is actually regular, since the infinitive form of the verb has changed to ser. It seems that this is an invention, though, and Latin has esum for the past participle. Perhaps this means that in such situations the verb stem did not go into a zero grade?

However there are at least some examples of words with unusual consonant clusters arising through suffixation. Old Latin had stlocus for "place", which evolved into Latin locus. I believe this is a zero grade.

[Dewrad]

I don't know if this happens in practice, but I'm interested in the theoretical outcome, if it's known.

Imagine we have a root which starts with a laryngeal and ends with a stop; let's call it Heg. Now put this root into zero-grade and add a suffix which starts with a consonant: Hg-to-s. Initial laryngeals are lost in most IE languages (all except Greek and Armenian, IIRC), so we now have gtos. What happens to that initial gt?

I would assume that it undergoes voicing assimilation, progressive or retrogressive depending on the daughter language.

[Grimalkin]

It's probably also likely that in some branches a) an epenthetic vowel was inserted either before the cluster or between the two consonants, or b) the cluster was reduced. Without any data at hand, I can't back this up though.