Re: Some questions on phonology

On Mon, 13 Oct 2008 12:49:43 +0100, Falcata Lusa <falcata.lusa@...> wrote:

>If we could go back in time and see (and hear) the first speaking hominids,
>what do you think their sounds could be like?
This is a question I've mused on some time.  I don't have any concrete
answers.  I speculate, at least, that it would be a highly unmarked
phonology, in particular that it wouldn't display any of the deviations from
markedness that happen as the result of common sound changes.  But this is a
nice distinction to draw, and it gets into questions of how we know the
'truly' unmarked from merely the common outcome of a diachronic process.
(I've certainly seen a few phonological studies which go through great and
wince-inspiring contortions to explain on synchronic grounds oddities that,
if diachronic explanation was permitted, wouldn't cause such trouble --
missetting the bar for what synchronic markedness 'should' explain, as it were.)

If there's a type of phonological pattern that characterises mimetic or
sound-symbolic or such phenomena in current natlangs, maybe it's a good
candidate for also characterising the first spoken language.

(Perhaps I only speculate this once the language had settled into having a
phonology at all.  I have no idea whether the very first oral linguistic
behaviour would have a phonology, as we know it.)

Below you ask

>Is there a study on the chronology of consonants? Which ones appeared first
>and why?
>Is there a study on the chronology of vowels? Which ones appeared first and
>why?
Well, the biologists say that ontogeny recapitulates phylogeny; so if I were
to take this as a cue and misinterpret your question as "appeared first _in
child language acquisition_", that's a very well-studied question.  And it's
not inconceivable that the answer to this may be related to the question you
did ask.

>Would they have click consonants in their language?
Clicks are tricky.  I've read (forget where) that the indigenous populations
of south Africa where clicks natively occur also tend to have a different
shape of the alveolar ridge or palate or something, which makes the sound of
clicks more distinctive.  I don't know what this says on the question, but
something, certainly.

WP has this to say:

| Clicks are often portrayed as a primitive or primordial feature of human
language, but
| we have no reason to suspect that they are very old compared to other
speech sounds.
| In fact, given their complexity, they may be relatively recent. How clicks
arose is not
| currently known. Some linguists speculate that clicks were initially used
for taboo
| avoidance and then borrowed into regular speech. (Compare Damin.) Others
suggest
| that they developed from other complex consonants. For example, the
Sandawe word
| for 'horn', [t&#620;ana], with a lateral affricate, may be a cognate with the
root [&#331;&#449;a&#720;] found
| throughout the Khoe family, which has a lateral click. This and other
words suggests that
| at least some Khoe clicks may have formed from consonant clusters when the
first vowel
| of a word was lost; in this instance [t&#620;ana] &#8594; [t&amp;#620;na]
&#8594; [&#449;&#331;a] (= [&#331;&#449;a]).

>If I were to use this set of consonants and clicks...
>Consonants:
>p, t, k, q, b, d, g, G\
>m, n, J, (with N, N\ as allophones of n before velar and uvular consonants
>respectively)
>s, S, X, h
>r, R\,
>K,
>l, L
>Clicks:
>|\, !\, |\|\
>
>...which of this groups of vowels/semi-vowels would be appropriate and why?
Any of them, I think.  I don't know of any strong correlations between
consonant-sets and vowel-sets that would be relevant here.  Granted, your
third option

>vowels:
>@, a
>semivowels:
>w, j
is a marked one for not having any high or non-central vowels, but it's no
more or less marked wrt your consonants.

Incidentally re semivowels Chris Bates says
| It is somewhat difficult to find an uncontroversial example of a natural
language which
| lacks phonemic semi-vowels. For many languages there is a strong and clear
relation-
| ship between semi-vowels, like [j], and their equivalent vowels ([i]), but
there is often
| disagreement about whether this makes them allophones of the same phoneme.

>What other consonants do you think are missing in the group, or which ones
>are dissonant. Why?
Oh, if you're doing clicks, then definitely more clicks.  Clicks, in the
languages that 'honestly' have them (there are several Bantu languages have
a small set of clicks through borrowing, and there's Damin, both of which
I'm excluding), form two-dimensional systems.  They vary both in point of
articulation at the front and what sort of release is going on at the back.
 You've got the POA variation, but I'd recommend some more click releases.

For (an extreme) example, !Xoo allows each of labial, dental, lateral,
alveolar, and palatal clicks coarticulated with each of [k_h k k?) g kx) N_0
N_0_<_h N ?N) q q_> G\] and clusters [gk_h gkx) k_>q_> gq_> G\x].

One other thing, though perhaps it's just a question of analysis.  If /n/
assimilates before velars and uvulars but /J/ does not, I'd find that very
strange indeed.  If they both do then there's not really any claim to
calling the first phoneme in a sequence like [Nk] /n/ as opposed to /J/, or
vice versa (well, unless independent constraints on occurrence of these
consonants provide one).  What you'd have there then is a nasal archiphoneme.

Alex