Oooops!

[BARudes at aol.com]

Dave,

Bob Rankin gave the history correctly in an earlier responses to RE: Andare.
Andare is not attested for Classical or Vulgar Latin.  Bob Hall proposed a
reconstruction *amDlare (where D represents a Greek Delta -- an arbitrary
symbol used to represent an uncertain sound) to account for It. andare, Sp.
andar, Pt. andar, as well as Franco-Provencal and Provencal anar and French
aller.  He related *amDlare to Latin ambulare, assuming syncope of the second
vowel and assimilation of mb to l.  There are a number of problems with the
reconstruction, among which are the fact that reflexes of purported *amDlare
are attested only in Italo-Western Romance -- no cognate is found in
Romanian, Dalmatian, Sardinian or even in Rheto-Romance (Western Romance) and
some Italian dialects (Italo-Romance).  Second, the cluster *mbl that
supposedly resulted from the syncope of the second vowel of ambulare  would
have been a unique cluster in Proto-Romance, and thus its outcome would be
unpredictable (which might explain the otherwise unique correspondene of
Northern French /l/ to /n/ in the rest of Western Romance and Italian -- Bob
Hall assumed that -mbl- > -mDl- > -ndl- > -nn- (Italo-Western except Northern
French)/-ll- (Northern French).  As Bob Rankin noted, *amDlare was not a very
satisfactory reconstruction for Proto-Romance.

Blair