Zagadki dźwięczności w j. polskim

Download Report

Transcript Zagadki dźwięczności w j. polskim

Eugeniusz Cyran
KUL, Lublin
1
Aim: to understand…
 Phonetic properties of voicing
 Phonological properties of voicing:
 Representation of contrast, e.g. b/p
 Distribution of laryngeal contrast

Processes connected with voicing:
 Neutralization of contrast
 Final Obstruent Devoicing (FOD)
 Regressive Voice Assimilation (RVA)
 Progressive Voice Assimilation
 Role of sonorants as the target, source and barrier
 Relationship between phonology and phonetics
2
Two-way voicing contrast in Polish
#_V
V_V
 pić [pjit ] ‘to drink’
 bić [bjit ] ‘to hit’

rysa [r sa] ‘scratch’
ryza [r za] ‘ream’
#_SV
V_SV
 płotem [pw t m] ‘fence, instr.’
 błotem [bw t m] ‘mud, instr.’
oknie [ k ] ‘window, loc.’
ognie [ g ] ‘fire, pl.’
__(S)V
3
Neutralization and Final Obstruent Devoicing
a. [vaga]/[vak] waga / wag
[ aba]/[ ap] żaba/ żab
‘scale, nom.sg./gen.pl.’
‘frog, nom.sg./gen.pl.’
b. [muzgu]/[musk] mózgu/ mózg ‘brain, gen.sg./nom.sg.’
c. [d br ]/[dupr] dobro /dóbr ‘goodness, nom.sg./gen.pl.’
__ (S) #
4
Neutralization and Regressive Assimilation
a. [d x]/[txu] dech/tchu ‘breath, nom.sg./gen.sg.’
b. [pr
it ]/ [pr ba] prosić / prośba ‘to ask/a request’
c. [kf jad b g
ji] kwiat begonii ‘begonia flower’
d. [mEndrEk]/[mEntrka] mędrek/mędrka ‘smart-aleck,/gs.’
__ (S)C
5
Distribution of laryngeal contrast in Polish
C
(S)
Lar
V
a.
b.
c.
... C (S) V...
|
Lar
... C (S) #
... C (S) C...
Lar
Lar
= obstruent
= optional sonorant
= laryngeal contrast
= vowel
6
Binary representation of voice [+voi] / [–voi]
Simplified story:
everything that is phonetically voiced has [+voi]
everything that is phonetically voiceless has [-voi]
/a/
|
/b/
|
/m/
|
/p/
|
[+voi]
[+voi]
[+voi]
[–voi]
7
Neutralization and Regressive Assimilation
in [±voi] systems
a. liczba
/lj i t
- ba/
>
[ljid ba]
‘number’
>
[ apka]
‘frog, dim.’
[-voi]
[+voi]
b. żabka
/
a b
- k a/
[+voi]
[-voi]
8
Neutralization and Final Devoicing (FOD)
a. stóg /stu g/
>
[stuk]
‘haystack’
[+voi]
[-voi]
b. stuk /stu k/
default feature
>
[stuk] ‘knock’
[-voi]
[-voi]
default feature
9
Problems with binary representation
 It is able to describe everything
- without providing much insight (understanding)
 Feature [+voi] behaves differently in sonorants and
obstruents, e.g., asymmetry in:
 assimilations
 devoicing
 Being symmetrical, [± voice] ignores universally observed
asymmetries between [+voi] and [-voi] (markedness).
 implications
 distribution (direction of neutralization)
 frequency of occurrence
 order of appearance in acquisition, etc.
10
Ways to avoid binarity problems
 Rule specificity and rule ordering, e.g.:
 [+voi] can spread only from obstruents, and only onto
obstruents (assimilations)
 [+voi] spreads or is provided at the „right moment”
 Underspecification of sonorants
 [+voi] is added later in derivation
 especially that it comes in handy sometimes…
11
Markedness tendencies (puzzle?)
unmarked
(default)
marked
Obstruents
[-voi]
[+voi]
Sonorants
[+voi]
[-voi]
(Default rules, Markedness conventions)
[+sonorant] →
[-sonorant] →
[+voi]
[-voi]
12
The key to understanding voicing
is in phonetics
13
Aerodynamic conditions on voicing
oral and nasal exit
Larynx and vocal cords
sonorants
P1 > P2
obstruents
P1 = P2
Conclusion:
…
…
…
…
…
…
…
…
…
…
…
…
…
P2
P1 > P2
P1
Relaxing facial musles
Short closure
Larynx ↓
Voicing of sonorants is spontaneous
Voicing of obstruents requires additional active gestures…
P1 ↑
14
Privativity
 Voiced sonorants should be unmarked
– unless they are voiceless
 Voiced obstruents should be marked
- unless they are voiceless
If there is no contrast, no marking is necessary…
15
Phonetic categories based on VOT(Voice Onset Time)
closure
release
VOT lag
vowel
vowel
VOT lead
[d]
t
[t]
[th]
fully
voiced
voiceless
unaspirated
voiceless
aspirated
CL
Co
CH
16
Voicing and Aspiration languages
‘voicing’
‘aspiration’
Romance
& Slavic
Hawaiian
Polish
Icelandic
Thai
Hindi
Germanic
voiced
voiceless
unaspirated
voiceless
aspirated
[d]
[t]
[th]
/CL/
/Co/
/CH/
/dL/
/dL/
/dL/
/to/
/to/
/to/
/to/
/to/
/tH/
/tH/
/tH/
[d ] = /dL+H/
17
Privative models: Laryngeal Realism in
Element Theory (GP)
(Honeybone 2002, Gussmann 2007, Harris 2009)
3 types of voicing:
 Spontaneous (universal phonetics)
 No marking!!!
sonorants Vo, So
 Active
 Marked
obstruents CL
 Passive
obstruents Co
 No marking (voicing is system dependent)
Within one system, voicing in obstruents is either active or
passive, never both!!!
18
Neutralization and Regressive Assimilation in
Laryngeal Realism
a. liczba
/lj i t
o
- ba/
>
[ljid ba]
‘number’
>
[ apka]
‘frog, dim.’
/L/
b. żabka
/
a bo
- ko a/
/L/
19
Neutralization and Final Devoicing in
Laryngeal Realism
a. stóg /stu gg/o/
>
[stuk] ‘haystack’
/L/
b. stuk /stu ko/
>
[stuk] ‘knock’
20
Time for a real puzzle…
21
Cracow-Poznań Sandhi Voicing
Warsaw Polish (WP) vs. Cracow-Poznań (CP)
a.
jak oni
wkład odrębny
WP
k-o
t-o
b.
jak możesz
wkład mój
k-m
t-m
g-m
d-m
__S[+voi]
c.
jak dobrze
wkład własny
g-d
d-v
g-d
d-v
__C[+voi]
jak trudno
wkład stały
k-t
t-s
k-t
t-s
d.
CP
g-o
d-o
__V[+voi]
CP
WP
__C[–voi]
22
Formal analysis in binary feature models
 Spreading of [+voi] as in Regressive Voice Assimilation
 The target must be first neutralized
 The difference between WP and CP lies in the scope of the
spreading rule wrt the source/trigger
 WP: spreading [+voi] from obstruents only
 CP: spreading [+voi] from any segment that has it (including
vowels)
23
Binary feature analysis (Rubach 1996)
WP
a. /j a k #
[-voi] [-voi]
o
i/
[+voi]
default
b. /j a k
#
[-voi] [-voi]
mo e /
[+voi]
CP
/j a k # o
[-voi]
/j a k
i/
[+voi]
#
[-voi]
mo e /
[+voi]
default
c. /j a k
[-voi]
#
d o b e/
[+voi]
/j a k
[-voi]
#
d o b
e/
[+voi]
24
How about Laryngeal Realism?
Polish is a voicing language (Co vs. CL)
WP works perfectly
Phonology
a. /j a ko # oo
Phonetic interpretation
i/
> [jak o i]
b. /j a ko # mo o e /
> [jak mo e ]
c. /j a ko # d o b e/
> [jag dob e]
L
CP is a nightmare!
25
Variation in laryngeal systems and a hypothesis…
phonetic categories
[b]
Slavic &
Romance
[ph]
[p]
L
H
Icelandic
English
Dutch???
H
H
26
Laryngeal Relativism
phonetic categories
[b]
Warsaw
Polish
Cracow-Poznań
Polish
[p]
[ph]
L
H
Voicing of obstruents is passive in CP, and active in WP
27
Final Devoicing in CP is interpretational
not computational
/ oaboa/ > [ aba] ~
/ oabo/ > [ ap]
Final Devoicing is rather an absence of passive voicing
Textbook question: Are we dealing with FOD or intervocalic voicing in
[Zaba~Zap]?
Textbook answer: FOD, because if there was a rule of intervocalic voicing,
then /mapa/ → *[maba]
Wrong: we do not expect intervocalic delaryngealization
/mapHa/ → /mapoa/ > [*maba] in CP
CP has Neutralization, but it takes place in the contexts {_#, _C}
/mapH/
→
/mapo/
>
[map]
28
Neutralization and Regressive Assimilation
in Laryngeal Relativism
a. liczba
/lj i t
o
- bo a/
>
[ljid ba]
/H/
b. żabka
/
a bo
- k a/
>
[ apka]
/H/
29
What about
Cracow-Poznań Sandhi voicing?
30
Just two more details…
The target of sandhi voicing must be /Co/
- either lexically neutral
- or neutralized
The source of voicing of obstruents:
WP = /L/
CP = phonetically voiced context
CL
Co
31
A reminder of what happens in Warsaw…
Co must be voiceless in an L-system
Phonology
Phonetic interpretation
/j a ko #
oo
/j a ko #
mo o
/j a ko #
d o b
> [jak o i]
i/
e /
> [jak mo e ]
e/
> [jag dob e]
L
32
In Cracow-Poznań, on the other hand…
Phonology
/j a k
ko #
Phonetic interpretation
oo
i/
> [jag o i]
H
/j a k
ko #
mo o
e /
> [jag mo e ]
e/
> [jag dob e]
H
/j a ko #
do o b
H
33
Because in Cracow-Poznań…
/Co/
[+voi]
must be voiced in front of V, S, C
inside words
and
CoVo [dom]
CoS o [brat ]
CoCo [gd ]
=
=
=
between words
Co#Vo [brad-ojtsa]
Co#So [kub-r be]
Co#Co [jag-dob e]
Sandhi phonetics is a very apt term to apply to CP voicing
34
The main pillars of this analysis
 „Reversed” marking of obstruents in CP and WP:
CP system = CH-Co
WP system = Co-CL
Warsaw Co cannot be passively voiced
 CP voicing requires:
 A system with marked voicelessness: CH-Co
 Passive voicing
 Neutralization CH → Co / {_#, _C}
35
Advantages of this analysis
 Sonorants remain unmarked
 Their voicing is only of phonetic nature and importance
 No special phonological rule is required for CP sandhi
voicing
 No rule ordering either
 Sandhi voicing = word-internal voicing in CP
36
Consequences of this analysis
and Laryngeal Relativism
 There is no phonological voicing in CP
 Only spontaneous and passive
 Final Obstruent Devoicing can be:
 Phonological (in Warsaw system)
 Interpretational (in Cracow-Poznań system)
 Assimilations can be:
 Phonological


Spreading of /H/ or /L/
Neutralization (deletion of /H/ or /L/)
 Interpretational (WP /toxou/, CP /jako doob e/)
 Full voicing of obstruents, FOD and RVA are not adequate
criteria for claiming that a given language has [+voi]
 The relation between phonological categories (H,L) and
phonetic categories (b-p-ph) is by and large arbitrary!
37
Between phonology and phonetics…
Sound system (e.g. Laryngeal system)
Phonology
Representation
&
Computation
Phonetics
Phonetic categories
&
Phonetic interpretation
-privative categories
-universal phonetic principles
-(un)licensing
-universal principles of
phonetic interpretation
-(de)composition:
spreading, delinking
-system specific conventions
-sociolinguistic modifications
38
Aim: to understand…
 Phonetic properties of voicing
 Phonological properties of voicing:
 Representation of contrast, e.g. b/p
 Distribution of laryngeal contrast

Processes connected with voicing:
 Neutralization of contrast
 Final Obstruent Devoicing (FOD)
 Regressive Voice Assimilation (RVA)
 Progressive Voice Assimilation
 Role of sonorants as the target, source and barrier
 Relationship between phonology and phonetics
39
C
|
V
|
o
C V
|
No
|
H
C
|
k
|
H
V
You!
40