Lecture Notes
Lecture Notes
ie rASL worcD
?iLrg-
In1
in rLk ling sp cVl
itc ALu..
buy tes
words
Table 4. Lexical representations for the words sudden and help. The syllable structure of
s
s = syllable, so = onset, r = rime).
each word is schematized at the top (o
r o r r
o o
/ r
s v d e n h ∈ 1 p
o r
Vowel > -i-
. "·· -
Glide
vowel Consonant
+ + +
Stressed +
glide
Reducible
consonant . + + + + +
stressed Continuant
+ +
Sonorant
reducible - +
Strident
continuant +
Lips
sonorant Tongue blade- + +
strident Tongue
+ body
lips Round +
tongue blade Anterior
+ + + +
tongue body Lateral
round High
anterior Low+ + + +
lateral Back +
high
Tense
Spread Glottis
low
back Nasal + +
tense Stiff vocal
folds
spread glottis +
nasal +
Lexical representations
+
stiff vocal folds for the words sudden
+ and help.
The syllable structure of each word is schematized at the
.~~~ _ ._ _{
~i- ~ ,xL~Jat
: oa
av,'z
\n
T.
t n&mTaxks
I I Time (msl I
-- I,-- -
-3-
N N
mr , 0 00
E[L ' ' s'° r 'r- 4n t
'O
~I ts ° < rI r 8
C t-n*E
tI il1 0 o 'o -°,r ) 0 liii
2NCo
iI, 11 ,
tr c:n aE S
'
E
, 9Lf 3M vz
E3r-,,c vn
-
x- C0 ,v
o c o - '0o- o
3
ti 0
V,
wi N- 3 C A r v
_N
- v Er
0
'/-.
0
a,
00
I--
'0
I I
r
/011~1\ (up) dwV -le (gP) WY¥ -1Tl 8o0 CD CD
0D
~o o 0
9,
N N
£ E i m8 vo 0
m
E or'
co~~ £° !t n
v~" r ° i i) Co I II II o1 I II
t- A: v.L v..-Ci0 v))L a= 6 v > ow w m V);
- - - -; X~~~~
/- It
0
0 i ~~~- ' I-kc N N
._
I __ _ ~~~~~~~~~f i
o 0
0) /
_
_ < /
o 2
0
0)
N aa)
0
__ / =_ _
0i -,2,
< I¢ 8
_ 7c
=: 6 t
0
o (CP) od D
(BP) JWV ?9 8 0 00 (Sp) cd'V 1t'I 00 0c
o 0 01 C
0 0 C0
'0 'o
---------------------------------------
e of v' Ai\ i t BV V\ea
< I
look at
c Simple Model
pe-cAI BtiCti,).
Bpr .·c
Phonological
Planning Stage
Representations
s;uhp
I 1 de
Discrete
Representation
Gesture Calculations
Continuous
gestures Discrete
epresentation
Vocal Apparatus
Continuous gestures
Sound Output
A Planning Stage is added to the basic component in the Figure before.
Figure 2. A Planning Stage is added to the basic components in Figure 1.
~
.A, ~~ ~ ~ ~ ~ i iO .S~t
(b)
---
0 200 400 600 0 200 400 -600 800
Time (ms) Time (ms)
_7_ IC__-··-·---·-------1IIIII··I-PC-
cra;~-
r
5
(C)
4 i i i~;
53
0 -
0 OZU 400
Time (ms) Time (ms) Time (ms)
T it+- A
I
(b)
c-
91
f- - 5
5
1--
I) (b)
i
· 4 ..I
4 4
53.
r-
M 31t
ZTs
uL
a,
c -LL
0 n
0 200 400 600 o0 600
Time (ms) Time (ms)
_·rPl______sCC____I___ll__q______ ---
I-,
%
,J
3
u
"-2
n
%/
r~
U zU0 400 600 800
Time (ms)
I I- r------- ~----- r
(b)
&L co·
5 ct
U
-r
v
c
if
a,
2
qPj
LL
UU 800 1000
Time (ms)
I-
!-- 5 l
(a) 'I ¶ I,
A (b)
d)
4 ljfflXl"f E 4
3
Ii;
u
o2
2!
e1
-2
I
0
4VUu oUU U LUU 4UU 6uu rSUU
Time (ms) Time (ms)
b1 ctN bass
---
c- ---- -f-----~------
REVISED
\ryx MODEL Oaa jU
Language Specific
Universal Component
Component
Phonological
Planning Stage Enhancement
Representations
Discrete
Representation
with flags
Enhancement
Gesture Calculations
Gesture Calculations
Continuous
gestures
Vocal Apparatus
,estures
Sound Output
Enhancement and Enhancement Gesture Calculations are added to the 'Simple Model' figure. The Enhancement
Figure 4. Enhancement and Enhancement Gesture Calculations are
added to Figure 2. The Enhancement component augments Planning Stage
component augments Planning Stage representations by flagging
representations by flagging relevant features.relevant features. Enhancement gestures are
Enhancement gestures are
calculated for these flagged features. calculated for these flagged features.
Phonological
Planning Stage Enhancement
Representations
Discrete
Representation
with flags
Enhancement
Gesture Calculations
Discrete Gesture Calculations
Continuous representation
with flags
representation
Vocal Apparatus
Sound Output
_ -·II·-(llliPllgCC-1
E
N E
F
AII
u, *d4 m N -- 0
(zMH)Abuanbaij
$7?
f".
(,aP) cJWV-19
0
m, 41
m A> m
_-T
__CI I sll_ _r___
a
2:
I0
"-2
1
a
200
Figure 10.5 Spectrogram of the sentence I can't go up to Sweden (male speaker), illustrating
modifications of consonants when they occur in a sequence of two or more consonants. The text
discusses the sequences /nt#g/ (400 to 550 ms), /p#t/ (700 to 850 ms), /sw/ (1000 to 1150 ms),
and /dn/ (1200 to 1270 ms).
Courtesy of MIT Press. Used with permission.
Acoustic input
Intended
word Extract acoustic parameters
sequence and cues in the vicinity
of landmarks
Compare
Synthesize
landmarks and parameters Estimate feature bundles and
syllable structure
Hypothesize
Match to Lexicon
word sequences
Lexicon
- --
·c"---------------------~~-~~~