Dear Greg and Yvan,
I just wanted to report a strange count from Phon.
The target (non)-word is /filpa/.
The actual production is [flipa].
For this word, the only errors that were reported are :
The deletion of [i].
The epenthesis of an [i].
No migration/metathesis is reported for that record.
No error was reported for /l/ even though the sound is in the same situation as [i].
I suppose Phon worked from left to right and detected the deletion of the [i]. During the alignment process, /l/ took the place of the deleted [i]. Then, /l/ was in the right place and was not reported as deleted. Finally, after the production of [l], the insertion of [i] was detected.
Regarding the lack of a reported migration or metathesis, I imagine that no check is made between a consonant and a vowel.
I did not check if this inconsistency in the count (errors counted for [i] and not [l], no migration/metathesis) is limited to that non-word or if it applies to all words of the shape CV-Liquid.CV produced as C-Liquid-V.CV.
I hope this makes sense to you and helps improve Phon.
Best,
Christophe dos Santos