The
role
of
phonetic
naturalness
in
the
grammaticalization
of
phonological
processes
has
long
been
of
interest
to
the
study
of
sound
change.
In
particular,
the
phenomenon
of
post-‐
nasal
stop
devoicing
has
attracted
a
lot
of
attention
in
this
debate.
A
number
of
authors
(Pater,
1996,
1999;
Hayes,
1999;
Hayes
&
Stivers,
2000)
have
proposed
a
universal
bias
for
avoidance
of
post-‐nasal
stop
devoicing
(*NT),
articulatorily
grounded
in
the
fact
that
leakage
of
airflow
through
the
nasal
cavity
upon
closure
of
the
velum
should
promote
vocal
fold
vibration
rather
than
inhibit
it.
Post-‐nasal
voicing,
then,
is
considered
more
natural
than
post-‐nasal
devoicing.
That
said,
a
number
of
counter-‐examples
to
this
constraint
have
been
observed
from
Bantu
languages,
including
Tswana
(Hyman,
2001;
Coetzee
et
al.,
2007;
Coetzee
&
Pretorius,
2010)
and
Shekgalagari
(Solé
et
al.,
2010).
In
light
of
this
debate,
this
study
provides
novel
data
from
Panará,
a
severely
understudied
Northern
Jê
language
spoken
in
the
Brazilian
Amazon.
Panará
exhibits
a
process
of
post-‐oralization
and
devoicing
of
nasal
consonants
when
they
appear
before
oral
vowels
(1).
Nasal
consonants
are
only
fully
nasalized
when
they
precede
nasal
vowels
(2).
(1) /m,
n,
ɲ,
ŋ/
→
[m͡p,
n͡t,
n͡s,
ŋ͡k]
/
__
V
(2) /m,
n,
ɲ,
ŋ/
→
[m,
n,
ɲ,
ŋ]
/
__
Ṽ
Acoustic
data
from
post-‐oralized
and
devoiced
nasal
consonants
was
collected
from
a
total
of
12
speakers
of
Panará.
Stimuli
consisted
of
the
carrier
sentence
[ĩkjẽhẽ
katõsyrĩ
X]
(I
repeat
X),
where
X
is
one
of
13
target
words
beginning
in
an
underlying
sequence
of
nasal
consonant
plus
oral
vowel
(NV).
Participants
repeated
each
target
word
five
times,
and
this
yielded
a
total
of
780
tokens
of
post-‐oralized
and
devoiced
nasal
consonants.
Preliminary
findings
suggest
that
there
is
indeed
no
vocal
fold
vibration
during
the
oral
closure
and
release
of
Panará
NCs,
and
that
immediately
adjacent
vowels
are,
in
fact,
fully
voiced.
These
results
are
especially
interesting
in
light
of
comparative
data
from
other
Northern
Jê
languages.
The
Northern
branch
of
Jê
languages
includes
Panará,
Timbira,
Kĩsêdjê,
Tapayuna,
Mebêngôkre,
and
Apinayé.
Of
these
languages,
only
Panará
clearly
exhibits
devoiced
post-‐
oralized
nasals
at
a
synchronic
level.
Other
languages1
of
the
family
have
the
phonetically
natural
alternation
whereby
post-‐oralized
nasals
are
voiced
(3).
(3) /m,
n,
ɲ,
ŋ/
→
[m͡b,
n͡d,
ɲ͡ʒ,
ŋ͡g]
/
__
V
The
comparative
data
thus
suggests
that
Panará
may
also
have
had
voiced
post-‐oralized
nasals,
as
in
(3),
at
an
earlier
stage
of
its
phonology.
The
presence
of
a
synchronic
and
categorical
process
whereby
nasal
consonants
are
post-‐oralized
and
devoiced
in
Panará,
as
in
(2),
is
a
challenge
to
theories
claiming
that
a
diachronic
change
from
ND
-->
NT
is
phonetically
unnatural
and
unexpected
(Flemming,
1995;
Hayes,
1999).
Specifically,
this
diachronic
change
can
be
considered
a
“sound
change
for
the
worse,”
as
it
involves
a
natural
phonological
process
becomming
unnatural.
Furthermore,
there
is
no
clear
historical
or
funtional
motivation
for
such
a
process
to
have
arisen
in
Panará.
This
study
thus
contributes
to
the
body
of
literature
suggesting
that
sound
changes
may
not
always
be
phonetically-‐grounded.
|
Development Neurocognition Disorders
Linguistic Diversity and its Sources
