Re[2]: Res: [tied] Romance brother

At 7:56:27 AM on Monday, May 28, 2007, Piotr Ga;siorowski
wrote:

> tgpedersen wrote:

> [João:]

>>> Is g- > h- regular in Lat. germanus > Sp. hermano, and
>>> if not, what happened? And how about Pt. irmão (e > i)?

>>> It's not regular. Latin geminus > Portuguese ge^meo;
>>> Latin generalis > Portuguese geral. The anomalous shift
>>> could be explained as influenced by phrasal lenition:
>>> teu germanu > *teu yermanu

>> Given the etymogies Piotr mentioned, are your sure your
>> etymologies are not learned loans from Latin?

> My impression is that Portuguese generally retained the
> glide in unstressed initial syllables and hardened it into
> /z^/, which is what one sees in Port. joelho : Sp. hinojos
> (< genuc(u)lum 'knee'), Port. giesta : Sp. hiniesta (<
> genistam 'broom plant'). But I'm not sure about the
> details and the relevant literature is out of my easy
> reach at the moment. I may be missing something, but
> either João is right about phrasal effects, or perhaps
> irmão has been influenced by a Castilian-type dialect.

Edwin B. Williams, _From Latin to Portuguese_, says that
Latin initial /g/ followed by a non-low front vowel became
VL [j] and Ptg. <g> [z^]; he gives the sequence of changes
as [g] > [gj] > [j] > [dz^] > [z^]. He adds that
<germa:num> seems to have developed regularly at first,
citing <germão> 1282 and <germaho> 1288, and notes that the
loss of <g> 'has been attributed to its intervocalic
position in the frequent occurrence of <germão> and <germãa>
with possessive adjectives'; it's an old suggestion, as for
this he cites Gröber, _Grundriss der romanischen
Philologie_.

By the way, Penny's little history of Spanish says that OSp
had <ermano>, the <h-> being a later addition.

Brian