> João_Simões_Lopes_Filho <josimo70@...> wrote:
>> But, for example, French have difficult in pronnounce English
>> "th". If a hypothetical French community that is annexed by an
>> English-languaged country, is it not plausible that this
>> population will maintain a shift T > s, or f, or alike?
Andy Howey wrote:
> I don't know if that's such a good example. The only
> reason the French have trouble pronouncing "th" is
> because they don't have that sound in their native
> language. They can learn to pronounce it, IF they want to
> learn.
Hehe, that reminds me the stubbornness of some French colleagues,
not willing to pronounce "Monthy Python" other than [mõ:tí pitõ]
(with stress on last syllable, ça va de soi). :-)
Piotr Gasiorowski wrote:
> But native French-speakers experience such a problem precisely
because
> their first language has no /þ/ or /ð/, not because of any innate
> predisposition against non-sibilant dental fricatives. It's a
trivial
> case of interference between L1 and L2 phonetics. Of course it may
give
> rise to substratal effects in appropriate circumstances, but there's
> nothing genetic about the kind of shift that you described. The
mismatch
> between the two phonological systems accounts for any such
interference.
> Actually, /þ/ and /ð/ are rather difficult even for young
> English-speakers; they are among the last phonemes acquired by
children.
Both answers touch the crucial point, that is: any person having to
cope with a foreign phonetical system has a natural tendency to
identify
every foreign phoneme with one from his' own familiar system. The
first
mental representation is always in terms of known units; even for
serious
phonetic differences, the mind analyses an L2 phoneme as being a
somehow
distorted L1 one. Of course, constantly exposed to L2 "correct"
phonetic
system, an L1 native might realize sooner or later that his
pronunciation
is fishy (especially if his fallacious imitation eliminates some
useful
opposition between phonemes as in "think/sink"), and he might even
manifest a certain amount of willingness to learn.
An aditional complication is introduced by the so-called "difficulty"
to pronounce correctly an L2 phoneme (that is: in such a way that a L2
native would not be able to differentiate it from usual L2 natives'
one).
This parameter has an important subjective part, depending on L1
system
but also on L1 native's ear, on his "mechanical" and learning
abilities
(not to forget his natural laziness). Of course, there are
articulatory
movements which are intrinsically more difficult since they require
more
energy, and the speakers will be naturally tempted to avoid them (most
of them are not unique phonemes but sequences); OTOH, for most
phonemes
"difficulty" is given by psychological and not by physical reasons.
Obviously, neither the tendency of identifying L2 phonemes with L1's,
nor real or supposed difficulties have something to do with genetics.
Brian M. Scott wrote:
At 9:45:29 AM on Thursday, October 9, 2003, Patrick C. Ryan wrote:
>> Let us take a concrete example. If a certain segment of a
>> population substitutes a fricative (/f/) for an aspirated
>> stop (/pH/), we can make one of two basic assumptions:
>> that segment finds it difficult (or impossible) to
>> replicate /pH/; or that segment does not properly hear
>> /pH/, and cannot distinguish it from /f/. Both scenarios
>> imply physical causes.
> No, they don't.
Actually, they might have also some physical causes: if articulation
of the aspirated stop requires supplementary movements, the natural
tendency would be to forget about them and substitute [pH] with
something sounding not-so-different-after-all and requiring less
effort -- for instance, an [f]. If the phoneme /f/ realized as [f]
fricative does already exist in the language, the result will be the
merging of the two phonemes in one of them; if not, speakers might
even not notice something has changed since one could speak only about
realization of /pH/ phoneme having had some shift from [pH] to [f].
But some possible physical causes are not an indication that genetics
should be involved. In fact, there are three different levels of
simultaneous incidence of things:
1. (simple) coincidences [which just happen, since the world is full
of more or less simultaneous events];
2. correlations [events that systematically occur together with good
statistics];
3. causal correlations [events that occur together because one of
them is the direct or the indirect cause for the other].
It is important to recognize and distinguish the three categories in
any pertinent analysis.
For the class of correlations (to which belongs the correlation
between languages in all aspects and genes), there is no direct or
indirect causality relationship in the sense of genes determining
or at least influencing languages' characteristics. There is indeed
some causal link for both genes and languages, and that is selectivity
of relations between individuals. In some sense, one could say it's
somehow the contrary: people have the natural tendency to live among
similar people -- and one important aspect of similarity is the tongue
they use. This selectivity is responsible for both: linguistic
convergent characteristics and genetical separations.
Regards,
Marius Iacomi