Re[2]: [tied] Linguistic measurements

At 2:37:06 AM on Tuesday, June 24, 2003, Miguel Carrasquer
wrote:

> On Tue, 24 Jun 2003 00:41:38 +0200, Davius Sanctex
> <gruposdavius@...> wrote:

>>[Miguel] it's easy for a Portuguese-speaker to understand
>>Spanish, but for an untrained Spanish-speaker, it's
>>relatively difficult to understand Portuguese.

>>[Davius] What is the reason for this, Miguel?

>>Is the reason alike that this one: For any Peninsular
>>Spanish speaker is relatively easy to re-learn correctly
>>any variety of American Spanish, because the main
>>phonetical differences consists of mergings, but I really
>>know very few latin americans that could reproduce without
>>errors "th/s" alternances of Peninsular (and that have
>>merged in America).

> Mergers indeed would seem to be easier for speakers of a
> non-merging dialect.

Easier to learn to produce, but what about understanding?
It seems to me that this goes in the other direction: when
listening it's easier to ignore distinctions that you don't
normally make than it is 'uncollapse' a merger.

> I haven't really thought about all the consequences for
> intelligibility and learnability (not the same thing) of
> mergers, splits, losses and changes.

> Portuguese shows a greater number of losses in comparison
> to Spanish.

And the same here, I think: in one direction the listener
has to ignore extra material, while in the other he has to
create it.

Brian