I came across the verbs כרע (krʕ) in Hebrew and ركع (rkʕ) in Arabic, both meaning something like "bow, kneel." Is it possible these words are cognate, and if so, why would the letters have switched order?
1 Answer
Different Semitic languages often have irregularly corresponding roots. Sometimes this leads to irregularly corresponding sibilants, sometimes to unexpected (de)voicing or (de)emphasis, and other times to metathesis like this (amongst many other types of irregular correspondence).
In some cases these are typically ascribed to borrowing between languages (in particular, a lot of instances where Arabic has an unexpected reflex of a sibilant are ascribed to Aramaic borrowing).
In most cases (as with this case) however, this does not work (as there is no language where this metathesis would be regular), and the correspondence is simply unexplained.
In my opinion this is likely a result of the root+template system leading to widespread systematic analogy reversing most conditional sound changes that would lead to an opaque root structure (something that would explain the relative lack of conditional sound changes in the Semitic languages, as well as the different correspondence sets exhibited within and outside of the root). We would then expect some instances where certain roots had the "wrong" alternant generalised, resulting in an irregular correspondence (also, in some cases, you could see a root split, as the two originally conditioned alternants become analysed as independent).
If correct, one or both of these languages would have (at some point in its past) undergone regular metathesis of /r/ & /k/ in certain positions, but when systematic analogy reversed other instances of this previously regular change it instead generalised the metathesised alternant (this part of course could only have taken place in one of the two languages).
