I read something interesting in thing Yechezkel Kutscher's Studies in Galilean Aramaic. On page 17, he writes:
k - q : kšţym (H) "archers" (=qštym) 611 (4 X). This is a particularly interesting example. The first t became ţ under the influence of the q.15 Then afterwards q dissimilated to k, since as, for example, in Akkadian, two emphatic consonants cannot coexist in the same root. Cf. kşyr "harvest" ( = qşyr) in Elephantine Aramaic (Aḥiqar 1, 127). We may therefore suppose that such a dialect also existed in Palestine/ We may pose the question whether the form tkšyţ is not really tqšyţ "decoration." This form occurs in YFG 106 22,26. See Epstein, MNM, p. 1227.To explain this in simpler terms, Kutscher is discussing the phenomenon of switching off between q and k, that is, ק and כ, in dialects of Galilean Aramaic, and specifically forms that appear in the Vatican manuscript of the Jerusalem Talmud. Here, his example is a Hebrew term, kšţym, כשטים, to mean archers, rather than qštym, קשתים. We would expect the latter, קשתים, because qšt, קשת, means a bow, from which the word קשתים arises. There are two changes here: from ק to כ (q to k), and from ת to ט (t to ţ). This word occurs 4 times on folio 611 of the Vatican manuscript.
15 See also in Palmyrene and in TY: qšţym, qštym YFG 181 9,10.
Kutscher offers an interesting analysis. First, the ק (q), which is an amphatic version of כ (k), influenced the ת (t), transforming it into an emphatic ט (ţ). This process is called assimilation. Then, since we dislike having two emphatic letters in a single root, the presence of ט (ţ) causes to ק (q) to dissimilate, and lose its emphatic quality, to become a כ (k). The purpose of footnote 15 is to show the intermediate stage, with both the ק (q) and ט (ţ), occurs in Talmud Yerushalmi, in Ginzburg's Yerushalmi Geniza Fragments. For the second transformation, he cites precedent in Akkadian, and also in Elephantine Aramaic - in the latter, the emphatic צ (ş) causes the original ק (q) to dissimilate to a כ (k).
Among other reasons I am interested in this is that I am in the midst of writing an essay for a take-home final in Theoretical Linguistics, and we need to show two concrete example demonstrating the importance of rule ordering. Here, it is clear that the rule causing assimilation must happen before the rule causing dissimilation, because otherwise, we would not have the emphatic ט (ţ) to cause the dissimilation of ק (q).