79. Of the groups not yet discussed, the most important are ct, gd, gn,
ks, which show palatalization. It is now generally assumed that the
Celts, who had turned their native ct into χt, pronounced Latin ct in
the same way when they learned Latin (Meyer-Lübke, _Einf._, § 186), and
likewise substituted χs for ks (Meyer-Lübke, _Gram._, I, § 650), and
probably χd, χn for gd, gn. The χ was attracted into a palatal spirant
by the following dental, and the dental itself was then palatalized.
Most philologists explain the development of cl, gl into l´ (cf. § 68)
in a similar way. Inasmuch as Indo-European pt had also been changed to
χt in Celtic, it is not unlikely that the Celts substituted χt, χs for
Latin pt, ps in a few words; the χ replacing p may sometimes have been
rounded.
1. To account for palatalization in the non-Celtic parts
of southern Gaul, we may assume either that the spirant
pronunciation spread from the Celtic to the other regions, or
that in the latter the palatalization came about simply through
the mutual attraction of the guttural and the dental.