Professor Witold Man'czak, a famous Polish
Romanceist, in his "Fonética and morfología histórica del Español", wrote
(p. 33):
<<
§ 85. Desarrollo regular: ssi, sse entre voc.
> j > [s^] > [x]:
*bassia:re > bajar, russeum >
rojo
>>
Btw. single -si- yielded -s- in Spanish, not -j-,
like in ba:sio: > beso.
In other words, -j- in pájaro is regular if we
accepted the intermediate form *passiarum (with i = IPA [j]) or
perhaps *passiaro (with -m already disappeared, and with u > o). This
form is certainly irregular in comparison to Latin passerem (as a rule,
accusative, not nominative was the base for modern nominal forms in
Romance languages).
As for *-um (or -o) instead of -em we have
more examples in Spanish. Man'czak writes (ibidem, § 193):
<<
Cambios de tipo. Conforme a la ley II fueron
sustituídos:
cicer por chícharo
>>
(btw. chícharo is irregular or dialectal, as Latin
c [k] should not yield ch [č] in Spanish at all so any parallels between the
development of -c- and -ss- in cicer and passer are wrong; "low 2" is referred
to Man'czak's laws of irregular developments, based on Zipf's law (already
mentioned here and then not known by some of discutants, and causing their
unintelligible protests).
<<
os por hueso
passerem por pájaro
>>
[os is of neuter gender so the accusative form is
correct]
(there are also examples for substitution for
feminine substantives but they are not in question here)
Besides, in § 83 the author
writes:
<<
Desarrollo regular: s > 0
>>
(0 means no sound)
<<
1. ante s: grossum > grueso;
2. ante c + voc. anterior: asciolam > azuela,
piscem > pez;
3. ante ti + voc.: *u:stium > uzo;
4. ante esp. ch: masculum > macho.
Desarrollo irregular: pájaro por
passerem.
>>
In § 19 Man'czak formulates the following phonetic
rule:
<<
Desarrollo reglar: e > 0
1. protónica interna primera: supera:re >
sobrar;
2. postónica interna: litteram >
letra;
3. postónica final tras d, s, z, y, l, r,
n:
>>
[here numerous examples which are not in
question]
<<
Desarrollo irregular debido a la asimilación:
pájaro por passerem.
>>
In other words, in opinion of the specialist ALL
internal posttonic -e- (i.e. unstressed -e- in nonfinal syllable -
which is possible only in words stressed on the third syllable, counting
from the end, just like passerem) should have disappeared in Spanish.
However, it is under question (in my humble opinion) if such a
phonetic rule functioned independently on which consonants enclose the -e-
sound. The expected *passre (with double -ss- in the beginning, cf. notices on
different development of siV and ssiV groups) must have been too hard to
pronounce, and that is why the vowel stayed instead of disappearing. It is even
hard to say whether the development -e- > -ia- [ja] was regular or not (and
then caused by assimilation, like Man'czak suggests) - because the number of
examples is too little to formulate a phonetic rule (for example, the rule like
this one: internal posttonic -e- > -ia- between two consonants and one
consonant).
Anyway, the example of pájaro shows clearly how
valuable are certain reconstructions, including IE reconstructions. I see
no reason for supposing that there were differences in number of "exceptions"
between Romance and Indo-European development. And because the number of such
"exceptions" in Romance languages is quite fair, I am not surprised that plenty
of IE words developed irregularly as well.
Grzegorz Jagodzin'ski
--- In cybalist@... s.com,
"stlatos" <stlatos@... > wrote:
> The original paper
at:
>
> http://www.mec. es/redele/ revista4/ galinanes. shtml
>
> Rosso deriva del latín russu(m), mientras que rojo parte del
latín
> russeus.
The same given much earlier in _A Dictionary of
Selected Synonyms
in the Principal Indo-European Languages_ by Carl
Darling
Buck. I can't find anything earlier.