What the process is here?

Question

Here is a phonological rule:

-ViC(-) > -VCʲ(-),
where i both /i/ or /j/;

and its vice versa:

-VCʲ(-) > -ViC(-).

(I think that -VeC(-) is possible too).

But I don't know what is the name of such rule, maybe it is particular case of the metathesis or the transphonologization.

And I don't remember where I found it. Maybe it was the description of Romanian language, or Hungarian, or whatever else.

So my question is about the term of such process and in which languages it can be observed historically or in loanwords.

UPD There was two responses for clarifying but I don't know how so I can give only examples. For the phonological rule they are (as pseudowords):

the historical stage in some language:
bein > beny /bɛin > bɛnʲ/;

or:

the adaptation of loanword in some another language:
rarrohaatat' > rerugatajt /rarːoɦaːtatʲ > ɾəɾʊɣatai̯t/.

UPD2 It seem to be kind of dissimilation in second case:

the /tʲ/ dissimilate into the /i̯t/ sequence;

but I don't understand its regressive character, i.e. why not to the /ti/ (maybe it is the forbidden coda).

Also there are more examples for metathesis:

Eng. foliage > foilage.

Answer 1

There isn't a name, but there is a phonetic explanation. The difference between a syllable [ViC] and [VCʲ] is somewhat subtle, having to do with how you detect palatal formants at the end of a syllable, and localize them on a vowel or a consonant. If you don't have a fairly robust consonant release burst, it's hard to identify Cʲ in the coda, so you have to listen for i-like transitions at the end of the preconsonantal vowel. Of course, that is what you also have to listen for in identifying [ViC], the difference being mostly one of time (i.e. over how long a span do you have palatal formants). This makes it difficult (not impossible) to maintain a contrast between [ViC] and [Vcʲ]. When there is no actual phonemic contrast, there is the potential for phonemic reanalysis. This, then, may explain a Greek sound change orjV > oirV (and an analogous example in a number of Bantu languages, known as "imbrication"), if the steps were orjV > or:ʲV > oʲr:ʲV > oir(:)V, that is, the palatal constriction tended to occur progressively earlier across generations. The contrary tendency ViC > VCʲ is less likely and fairly uncommon, because it involves moving the palatal formants from a fairly audible position to a position where they are much harder to identify. One should look into the details of the bein > bɛnʲ progression, which, I expect, is related to an avoidance of diphthongs.