Speech Science Exam 3 Flashcards
(173 cards)
1
Q
____ are speech sounds characterized by obsrtuction of the vocal tract compared to vowels.
A
consonants
2
Q
Consonants range from vowel-like sounds w/ relatively open vocal tract, like ___ such as [w] and [j] to sounds produced with sever vocal tract constriction like ___ [s], [t], and [f].
A
glides; obstruents
3
Q
The three features of consonants are ___, __, and ___.
A
voicing, place of articulation, manner of articulation
4
Q
What are different places of articulation for consonants? 7
A
bilabial, labiodental, (inter)dental, alveola, palatal, velar, glottal
5
Q
What are the different manners of articulation for consonants? 6
A
stops, affricates, fricatives, nasal, liquids (lateral, rhotic), and glides
6
Q
____ is that each of the distributed acoustic consequences of a gesture has some value as an acoustic cue.
A
Lisker’s Rule
7
Q
___ is another name for stop consonants.
A
Plosive
8
Q
The acoustic correlate for vocal tract closure is ____.
A
stop gap
9
Q
The acoustic correlate for release of the closure is ____.
A
stop burst
10
Q
The acoustic correlate for rapid articulatory movement is ____.
A
relatively fast formant transitions (mostly F1)
11
Q
The acoustic correlate for rapid opening or closing gesture is ____.
A
rapid rise/fall in intensity
12
Q
A stop gap is the acoustic correlate for ____.
A
vocal tract closure
13
Q
A stop burst is the acoustic correlate for ____.
A
release of the closure
14
Q
A relatively fast formant transition (mostly of F1) is the acoustic correlate for ____.
A
rapid articulatory movments
15
Q
A rapid rise/fall in intensity is the acoustic correlate for ____.
A
rapid opening or closing gesture
16
Q
If you have a lag between the burst and the vowel, the stop is _____.
A
aspirated
17
Q
____ is the time between the release of the burst and the beginning of the next sound.
A
Voice Onset Time (VOT)
18
Q
Harmonics during the ___ on a voiced stop consonant is apparent on a FFT on the spectrogram.
A
gap
19
Q
What are the 3 acoustic cues for stop place of articulation?
A
1 energy peak in the spectrum of the burst (mostly in non-final stops)
2 second formant (F2) transition
3 the duration of the VOT for initial stops
20
Q
Energy peak in the spectrum of the burst is an acoustic cue to the ___.
A
stop place of articulation
21
Q
Second formant (F2) transition is an acoustic cue to the ___.
A
stop place of articulation
22
Q
The spectrum of the ____burst has most of the energy under 600 Hz, and it has an overall down tilt.
A
labial ([p], [b])
23
Q
The spectrum of the _______ burst has most of the energy around 3000 to 4000 Hz for a male speaker. (may have a slight uptilt)
A
alveolar ([t], [d])
24
Q
The spectral peak of the ____ burst is linked to the F2 of the following vowel, and it is usually a few hundred Hz higher than the F2 of the following vowel.
A
velar ([k], [g])
25
_____ tend to have narrow spectral peaks linked to the F2 of the vowel.
velar stops
26
A labial stop has its energy spectral peak at the burst under ___
600 Hz (male speaker)
27
A alveolar stop has its energy spectral peak at the burst ___
3000-4000 Hz (male speaker)
28
A velar stop has its energy spectral peak at the burst is ___ (___).
narrow (a few hundred Hz higher than the F2 of the following vowel)
29
The F2 transition for bilabial stop is ____.
an up-sweep.
30
The F2 transition for alveolar stop is ___.
flat
31
The F2 transition for velar stop is ___.
a down-sweep
32
There is an interaction between ___ & ___ in addition to the F2 transition and the energy peak.
Burst & Vowel Formants
33
Technically speaking, the difference between a voiced and voiceless consonant is that the former are associated with ___
vocal fold vibration.
34
What are the 2 main cues for distinguishing voiced and voiceless stops in initial stressed position?
VOT (voice onset time) (the time interval btw the burst and the onset of voicing) & aspiration (puff of air - after the burst)
35
What are 3 other cues (besides VOT and aspiration) for distinguishing voiced from voiceless stops in initial stressed position?
1 low versus high starting position F1
2 relatively larger versus small F1 change (less imp)
3 voicing during the stop gap can also be a cue
36
What is the order of VOT, stop gap and burst?
1 stop gap
2 burst
3 VOT
37
____ is one of the main cues for differentiating initial /b/ and /p/ in English.
VOT values
38
Negative VOTs are only relevant for ______.
absolute initial stops
39
In medial and final stops VOT cannot be ___.
negative
40
A negative VOT is indicative of ____ (in absolute initial voiced stops).
prevoicing
41
In voiced/voiceless stop pairs in initial position based on placement, will have _____ VOT values for the voiced and ____ VOT value for the voiceless.
shorter (possibly negative - prevoiced); longer
42
The average VOT for /d/ is ____ compared to /b/.
longer
43
The average VOT for /p/ is ___ compared to /b/.
longer
44
The average VOT for /b/ is ___ compared to /t/.
shorter
45
How much longer do VOT have for voiced stop compared to voiceless?
40 msec
46
VOT matters more in ___ syllables and in ____ position.
stressed; initial
47
There is a greater difference between voiced and voiceless VOT in ____ syllables.
stressed
48
For unstressed syllables, VOT is the most reliable cue. T/F
false
49
A more reliable cue than VOT for differentiating voiced and voiceless stops is the ____ and ____ for the F1.
starting point; change
50
___ starts lower for voiced stops than for voiceless.
F1
51
F1 change is relatively ___ for voiced than for voiceless stops.
larger
52
F1 starts ___ for voiced stops than for voiceless.
lower
53
___ change is relatively smaller for voiceless than for voiced stops.
F1
54
Voicing during the ___ can also be reliable clue for differentiating voiced stops from voiceless in initial position.
stop gap
55
______ is the most salient cue for stop voicing in medial position
voicing during closure
56
What are four important cues to stop voicing in medial position? 4
*1 voicing during closure* most imp
2 duration during the stop gap
3 length of the preceding vowel
4 F1 transition for voiced stops
57
In medial position, unstressed medial voiced stops have ___ stop gaps than their voiceless counterparts.
shorter
58
Typically, vowels are ___ before a voiced stop in medial position than before a voiceless stop (same speaking rate).
longer
59
___ is the most salient cue for stop
| voicing in final position
voicing during closure
60
What are four important cues to stop voicing in final position? 4
*1 voicing during closure* most imp
2 duration during the stop gap
3 length of the preceding vowel
4 F1 falls at the end of the vocalic portion of a voiced stop
61
Typically, vowels are ___ before a voiced stop in final position than before a voiceless stop (same speaking rate).
longer
62
In final position, final voiced stops have ___ stop gaps than their voiceless counterparts.
longer
63
F1 ___ at the end of the vocalic portion of a voiced stop in final position.
drops or falls
64
____ is a consonant produced w/ a narrow constriction through which air escapes w/ a continuous noise.
Fricative
65
___ is a stop plus fricative combination.
Affricate
66
How many American english fricatives are there?
9 (2 dental, 2 labio-dental, 2 alveolar, 2 palatal, and 1 glottal)
67
How many American English affricates are there?
2 (palatal)
68
Fricatives have aperiodic ('irregular') noise component, this is ___ cue.
manner
69
Fricatives duration (of the noise) is relatively ___ than aspiration noise after voiceless stops or the fricative noise in affricates.
longer
70
What are 2 manner cues for fricatives?
1 aperiodic component/noise
| 2 duration of noise is longer than other sources of aperiodic noise
71
Voiced fricatives also have __ which is not a manner cue for fricatives.
periodic component (voicing; f0)
72
___ are the "noisy" fricatives and affricates - louder. (such as /s/, /sh/, /z/, /genre/)
stridents
73
Voiced fricatives will have a ____ component (As well as ___).
periodic; aperiodic
74
What's the difference between a voiced and voiceless fricative?
voicing!
75
Looking at an FFT, we can tell the difference btw voiced and voiceless fricatives by ___.
a low frequency peak
76
____ of the fricative has relevant place cues. (hint: s has a relatively high peak energy, and sh has a relatively lower peak energy).
The spectrum
77
The /s/ has a relatively ____ peak, and the peak of energy is related to gender too.
high (btw 4000-8000)
78
The /sh/ has a relatively ___ peak (compared to /s/). Related to gender.
lower; btw 2500-4500
79
___, ___, and ___ fricatives do not have narrow spectra.
Labio-dental, dental, glottal
80
A ___ for a male speaker often passes for a ___ for a female speaker (two fricatives).
/s/ to /sh/
81
For a male /s/ the peak energy will be around ___ and for a male /sh/ the peak energy will be around ___.
5kHz (8 for women); 3kHz (4 for women)
82
The ___ of fricatives relative to the vowel (stridents, such as "sh", "s", "z", and "zsh" are ____ than other fricatives).
amplitude; louder
83
For labio-dental, dental and glottal fricatives can be identified by ____ (relative to ___) for place.
amplitude/intensity/volume relative to the neighboring vowel
84
____ is another place cue for fricatives (besides spectrum) , which is relevant for distinguishing labio-dental and dental fricatives.
Formant transition of the vowel (mainly F2)
85
___ is a short stop + extra short fricative
| broadly
Affricate
86
What are relevant characteristics of affricates? 3
– silence (stop gap), burst, fricative
| with its amplitude rising fast
87
Which is the most and least salient characteristic for IDing affricates?
least: burst; most: sharply rising fricative noise
88
An affricate is a ___ stop + a ____ fricative.
short; extra short
89
The characteristic of the fricative portion of an affricate is ___.
its amplitude rises fast
90
The fricative noise in an affricate is ___ than the noise in a pure fricative.
shorter
91
If you have an utterance initial voiceless stop or affricate, the stop gap ___.
cannot be measured
92
The rise time (amplitude) of the fricative vs. fricative portion of affricate is ___ vs. ___ on average
76 ms; 33ms
93
The steady-state duration (length) of fricative vs. fricative portion of an affricate is ___ vs. ___ on average.
100 ms; 48 ms
94
Even though all English affricates are made in the same place, ____ and __ are place cues for affricates.
stop burst and the formant transitions
| especially F2
95
The presence of voicing (f0, periodicity) characterizes voiced affricates; voiceless affricates do not have ___.
voicing
96
What are the american english glides?
/w/ and /j/
97
What are the American English liquids?
/l/ & /r/
98
___ may involve a gliding motion from a partly constricted state to a more open state.
glides
99
___ are called semi-vowels.
liquids, glides
100
The palatal glide is ___.
/j/
101
The labio-velar glide is ___.
/w/
102
Liquids and glides have ____ that change fairly rapidly and are typically voiced.
resonant frequencies (formants)
103
for /ja/ and [wa] are differentiated based on ___.
F1 and F2
104
[le] and [re] are differentiated based on ___.
F3
105
/w/ is noted for a very fast ____.
formant transitions
106
Place cues for glides are usually found on ____.
the first two formants
107
__will have a falling F2 (it's a glide).
/j/
108
___ will have a riding F2 (it's a glide)
/w/
109
In terms of its acoustic, /j/ is similar to a ___
/i/
110
In terms of its acoustics, /w/ is similar to a __
/u/
111
Which two consonants are differentiated by the third formant?
/l/ and /r/
112
___ has a low F3 (which rises). (It's a liquid).
/r/
113
___ has a high F3 that does not vary in most vowel contexts.
/l/
114
Liquids and glides place cues work as ___
pairs.
115
The main difference between an l and an r is __
the third formant
116
Prevoicing is only relevant for ___.
initial stops
117
Prevoicing is relevant for ____ or ___
word initial prevoiced stops or affricates
118
The palatal glide is ___.
/j/
119
The labio-velar glide is ___.
/w/
120
Liquids and glides have ____ that change fairly rapidly and are typically voiced.
resonant frequencies (formants)
121
for /ja/ and [wa] are differentiated based on ___.
F1 and F2
122
[le] and [re] are differentiated based on ___.
F3
123
/w/ is noted for a very fast ____.
formant transitions
124
Place cues for glides are usually found on ____.
the first two formants
125
__will have a falling F2 (it's a glide).
/j/
126
___ will have a riding F2 (it's a glide)
/w/
127
In terms of its acoustic, /j/ is similar to a ___
/i/
128
In terms of its acoustics, /w/ is similar to a __
/o/
129
Which two consonants are differentiated by the third formant?
/l/ and /r/
130
___ has a low F3 (which rises). (It's a liquid).
/r/
131
___ has a high F3 that does not vary in most vowel contexts.
/l/
132
Liquids and glides place cues work as ___
pairs.
133
The main difference between an l and an r is __
the third formant
134
Prevoicing is only relevant for ___.
initial stops
135
(Negative) prevoicing is relevant for ____ or ___
word initial prevoiced stops or affricates
136
If there is a prevoicing, then the VOT will be ___.
negative
137
___ is the strongest cue for nasal identification.
Nasal Murmur or Nasal Resonance/Formant
138
All nasals are ___, __ intensity, have relatively ___ formants, and ____ resonance.
voiced, low intensity, steady-state formant; low-frequency
139
Nasal formant is usually below ____ Hz.
500 Hz (often below 300 Hz)
140
Nasal murmurs are/are not that different from each other.
are not
141
The nasal murmur has ___ amplitude (i.e. it is ___er than) the neighboring vowels.
lower; softer
142
All nasal consonants are tyically voiced/voiceless.
voiced
143
Three nasals are ___, ___, and ___.
/m/ (bilabial); /n/ (alveolar); /ng/ (velar)
144
____ (1956) found that identification of the isolated nasal murmur is not too reliable.
Malecot
145
Malecot tested the identification of nasals based on ___.
nasal murmur
146
The main cue for place of articulation for nasals is ___.
second formant transition
147
____ provide place cues for nasals as demonstrated by Malecot.
formant transitions (mainly F2)
148
What are two hypotheses on infant speech perception?
infants have limited perceptual ability
| infants are able to perceive all speech contrasts
149
What are the methods for determining which infant speech perception is true?
habituation versus dishabituation and measure the non-nutritive sucking rate, head turn, or heart rate
150
What is the stimulus for the infant speech perception research?
1 repeat stimulus (papapapa) and then change stimulus (like bababa or frequency)
151
_____ ran a study on infant speech perception on infants 26 to 30 months old.
Eimas and colleagues (1971)
152
What method did Eimas use?
pa vs. ba stimulatiosn with VOT at -20, 0, 20, 40, 60, 80 msecs measuring non-nutritive sucking rate; difference perceived
153
___ tests infant speech perception more, but testing 8 Spanish and 8 English infants.
Eilers et al (1979)
154
How did Eilers et al. test the Spanish and English infants?
pa vs. ba in English and Spanish contrasts; head turn paradigm; and the English infants did very well on the English contrast and not well on the Spansih, but Spanish did equally well on both.
155
What was the conclusion of Eilers et al?
some contrasts are harder to perceive than others; not all equal across languages (think and this, word initial consonant) (/r/ is acquired later)
156
____ has one of the most famous studies on infant speech perception which used English, Hindi, and Salish dental and retroflex stops using conditioned head turning.
Werker & Tees (1984)
157
Werker and Tees used infants of ___, ___, and ___ backgrounds of 0;6 to 1;0.
English, Hindi and Salish
158
What did Werker and Tees find for English speaking infants finding correct responses?
at first perceive native and non-native differences but around 10 months of age, native categories become much more relevant and cannot differentiate non-native pairs
159
_____ performed an experiment on her theory called the Native Language Magnet (1. infants detect patterns in input -> 2.
exploit statistical properties of input -> 3.
perception altered (warped) by experience.)
Kuhl et al. (2000, 2001)
160
What was Kuhl's Native Language Magnet theory? 3
1. infants detect patterns in input
2. exploit statistical properties of input
3. perception altered (warped) by experience
161
____ states that linguistic experience warps perception.
Native Language Magnet
162
___ are the most typical exemplars of a category, which are powerful acnchors that function as perceptual magnets to strengthen category cohesion. (Kuhl et al.)
Prototypes
163
Kuhl's 1991 study compared humans to ____; which looked at ___.
Rhesus monkeys; % correct ID of stimulus change by prototype effect
164
Prototypes are ___ represented. Monkeys are not affected by them.
mentally
165
What language and phonemes did Kuhl also work with?
Swedish /i/ versus /y/
166
What was the results of Kuhl's Swedish language study?
by 6 months infants show stronger magnet effect for their native language prototypes (American English for /i/ and Swedish for /y/)
167
___ criticized infant speech perceptions studies and said we should reconsider the notion of innate phonetic boundaries.
Nitrouer (2001)
168
Nittrouer hasd groups of young and not slightly older children (2;6 to 3;4) and exposed them to 5 sets of natural stimuli with sets of ysnthesize stimuli with hybrid stimuli with syntehtic noise and conditioned head turning to prove what?
that we should reconsider the notion of innate phonetic boundaries
169
What were Nittrouer's conclusions?
children may be uncooperative due to inability to perceive contrasts; notion of innate phonetic boundaries needs to be reconsidered
170
Infant speech perception aids ___.
phonological and language acquisition
171
Word segmentation occurs around ____ months of age out of input.
6-7 months
172
Initially infants are more able to differentiate ___ contrasts, but they gradually specialize and tune into ___.
non-native language; native language
173
Some aspects of language are perceived earlier than others, for example ___ or __.
prosody or fricative differentiation
