reversal of merger

[Larry Trask]

----------------------------Original message----------------------------
Miguel Carrasquer writes:
 
[on the Basque case]
 
> I like it too, but I like it slightly better reformulated as
> follows.  What do you think?
 
> (1) The devoiced /j/ genuinely did merge completely with the
> inherited <x> in Gipuzkoan.
 
> (2) When the backing of <x> was introduced, *all* instances were
> affected in principle, but those instances of <x> bearing expressive
> value were re-palatalized back to /S/.  However, <gajo> and
> <gizarajo>, in spite of their etymology, did not undergo
> "re-palatalization" because speakers no longer perceived their <x>
> as having expressive value.
 
Both scenarios are possible, and I know of no way of deciding between
them, in the absence of any textual evidence.  If we had some early
Gipuzkoan texts in which <x> were used in place of modern <j>, that
would probably settle things in favor of Miguel's scenario, but we
have no such texts.  Unfortunately, Gipuzkoan was one of the last
dialects to be written; we have nothing before the 18th century and
hardly anything before Larramendi, who published his main works
between 1729 and 1745, and these works exhibit the modern state of
affairs.
 
This little dilemma is reminiscent of the familiar problem of
"analogical maintenance" versus "analogical restoration", though our
case is phonological, not morphological.
 
In the absence of any evidence to the contrary, I prefer to invoke
Occam's Razor, and go with Michelena: no change to inherited <x>,
rather than change followed by reversal of change.
 
Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK
 
larryt at cogs.susx.ac.uk