--- In
cybalist@yahoogroups.com, "stlatos" <stlatos@...> wrote:
>
> But doesn't the form galondrina also exist? It seems that
> *arondrina existed in Iberia first, with met. in Portuguese but
> dissim. r>l, then contamination with gallina 'hen' in Spanish, then
> (in most dia.?) assim. a-o > o-o.
>
>> It's possible that er > ar was an intermediate stage of er > r in
> those environments where e>0 would occur, but was sometimes
> preserved when met. or analogy occurred (or maybe it was just
> sporadic).
>
What you say is probably true (though the details of the process will
remain unprovable, I suspect). I simply gave a linear account of
Spanish golondrina from Latin hirundinem. There are undoubtedly many
complications to that simple, linear story. For instance: what was the
original popular Latin word from which the rest derive: hirundo:? the
accusative hirundinem? a diminutive? And, in the latter case, was it
hirundi:na or hirundella? Were all current at the same time in all
territories? Did diminutive (or masculine) forms coexist all along the
story? everywhere? and how? I notice that some forms probably
coincided (as Spanish golondre, erondina and possibly some other) and
some interfered badly in another's development, as can bee seen (in
Spanish) in the cross-interference between the different swallow words
and the lark word (which had to be aloda but, crossed with olondre,
gave out alondra, at the same time that olondre had to differentiate
into golondre or whatever similar by-form).
Further complicating the scene, there were possibly (as you point out)
altered or mixed forms like galondra and galondrina, and a former
arondrina, and who knows what other intermediate or regional form.
(Notice that Italian loses the initial e- while it becomes a- in
French; I assume hesitations took place in both languages before
standardization froze one into the present dictionary avatar.)
So, in all, any linear description is probably a best rational
approximation and a excellent teaching aid but won't fit the actual
development of the nightingale word in Romance, more describable as a
thicket of alternates and half-baked, interfering forms than as neat
lineal inverted tree. One thing that is to be stressed, though, --and
my aim in what I wrote-- is the clarity and self-containedness of the
hirundo: -> golondrina /andorinha development, and the likely
irrelevance of substrate or adstrate considerations (like Gm. gal-
bird words in Spanish golondrina or Basque ander in Portuguese
andorinha) in this case. (I even doubt the possible contamination with
gallina that you point out --too dissimilar birds, I think, to
interfere-- unlike the larks and swallows.)
>
> *passero+ > pajaro (if *ssr wasn't allowed at the time,
> so e>0 couldn't occur).
>
For that, I think the usual explanation (i.e. that Spanish tends to
make an -a- out of the penultimate vowel in first-syllable stress
trisyllabic words such as relámpago or ciénaga, in application of the
"law of typical word-finals") is more natural than looking for some
phonetic rationalization (which, I'm sorry to confess, reminds me in
this case of Procust's bed).
Ton Sales
Ton Sales